F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.172383.1 Research Article Articles Application of K-Means Clustering for Job Applicant Analysis in Construction Firms Using R

[version 1; peer review: 2 approved with reservations]

 Jaya Daniel Jesayanto Conceptualization Data Curation Formal Analysis Funding Acquisition Investigation Methodology Project Administration Resources Software Supervision Validation Visualization Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0003-1940-6302 a 1 2 Ramdhani Wahyu Muhammad Formal Analysis Funding Acquisition Methodology Resources Validation Writing – Review & Editing https://orcid.org/0009-0000-3816-6100 3 Wati Endang Data Curation Funding Acquisition Software Validation Writing – Review & Editing https://orcid.org/0009-0008-4264-7984 3 Nandes Yogi Novario Funding Acquisition Investigation Writing – Review & Editing https://orcid.org/0009-0000-4028-3672 4 Nadhiroh Ilma Zahriyatun Funding Acquisition Resources Writing – Review & Editing https://orcid.org/0009-0002-0352-2691 4 Pade Reza Bakhrun Fidianto Funding Acquisition Software Writing – Review & Editing https://orcid.org/0009-0009-6281-8418 3 Technology and Vocational Education and Training, Universitas Negeri Yogyakarta, Yogyakarta, Special Region of Yogyakarta, 55282, Indonesia Building Engineering Education, Universitas Negeri Jakarta, East Jakarta, Special Capital Region of Jakarta, Indonesia Educational Research and Evaluation, Universitas Negeri Yogyakarta, Yogyakarta, Special Region of Yogyakarta, 55282, Indonesia English Language Education, Universitas Negeri Yogyakarta, Yogyakarta, Special Region of Yogyakarta, 55282, Indonesia danieljesayanto.2023@student.uny.ac.id

No competing interests were disclosed.

 10 12 2025 2025 14 1388 5 12 2025 Copyright: © 2025 Jaya DJ et al. 2025 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This study applies the K-Means Clustering algorithm to analyse job applicant data at a construction consulting firm. Considering three main variables—AutoCAD drawing skills, planning and supervision report writing skills, and adaptability—this study aims to categorize job applicants into three groups: “Rejected,” “Under Consideration,” and “Accepted.” The clustering process was conducted using the R program with an initial centroid-based approach and visualized through 2D and 3D scatter plots to map the distribution patterns of applicants based on their attributes. The results of the study show that this clustering not only provides a deep understanding of the characteristics of applicants but also supports the company in optimizing data-driven recruitment strategies. Applicants with high scores in all three variables tend to fall into the “Accepted” category, while those with moderate scores are categorized as “Under Consideration.” The visualization of results offers sharper insights into applicant distribution, which can be used to identify training needs or improvements to the selection process. This research contributes to the development of an efficient and objective data-driven recruitment system in the construction sector.

 K-Means Clustering; data-driven recruitment; workforce selection; cluster visualization; construction competencies Lembaga Pengelola Dana Pendidikan 202406111204346 202407111205431 202404111201727 202312211239164 202407112005435 202406111202933 This research was funded by scholarships awarded by the Indonesian Endowment Fund for Education (LPDP), with Grant Numbers 202312211239164 (Daniel Jesayanto Jaya), 202407112005435 (Wahyu Muhammad Ramdhani), 202406111202933 (Endang Wati), 202406111204346 (Yogi Novario Nandes), 202407111205431 (Ilma Zahriyatun Nadhiroh), and 202404111201727 (Reza Bakhrun Fidianto Pade). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. 1. Introduction 1.1 Research background

In the modern workplace, an effective workforce selection process is one of the key factors in supporting human resource development. Career transformation, as part of this development, is influenced not only by technical skills but also by an individual’s adaptability and interpersonal competencies. As explained by Pala (2021), data-driven workforce characteristic grouping can provide deep insights into how workers adapt to various work environments. The recruitment process not only involves searching for candidates from external sources but also requires effective and efficient decision-making, considering recruitment sources, techniques used, and the potential of available local labor ( Widodo, 2018). Recruitment begins with job analysis, which helps describe the tasks, responsibilities, and qualification requirements for each position, thereby facilitating the selection process and ensuring the suitability of applicants for the positions offered ( Widodo, 2018).

In the context of job searching, individuals need to understand their personal assets, study job opportunities, develop career plans, and build networks to obtain jobs that match their skills and needs ( London, 1973). This process reflects the importance of self-development to compete in an increasingly complex job market. Gangl (2003) emphasizes that job skills are the primary factor determining an individual’s chances of securing employment, particularly in sectors requiring specific technical expertise. In the construction industry, technical skills such as the ability to use AutoCAD for drafting, prepare planning reports, and adapt to collaborate with various stakeholders are highly valued attributes.

The increase in infrastructure development in Indonesia, such as the Nusantara Capital City ( Ibu Kota Negara, here-after IKN ) project in East Kalimantan, is driving the need for competent construction workers. This national strategic project aims not only to relocate the country’s administrative functions from Jakarta but also to promote economic equality and development outside Java ( Irmawan et al., 2023). The IKN project requires over 260,000 construction workers by 2024, spanning various educational levels and backgrounds, to meet the demands of technical work and social adaptation on-site ( Supriyanti et al., 2023). However, the labor selection process for large projects like this often faces challenges in managing applicant data efficiently, given the large and diverse volume of data.

The cluster analysis-based approach offers a solution to help companies understand the characteristics of job applicants. Clustering is a statistical technique used to divide data into several groups based on the similarity of certain attributes, thereby facilitating interpretation and decision-making ( Jain et al., 1999). One of the most commonly used clustering algorithms is K-Means Clustering, which divides data into several clusters based on the distance to the centroid, which is the center of the data. This algorithm works through iterations to minimize the distance between the data and the cluster centroid until an optimal result is achieved. In the context of workforce selection, K-Means allows companies to group applicants into specific categories such as Rejected, Under Consideration, and Accepted, based on their specific characteristics.

This process also opens up opportunities to identify applicants with the highest potential for further development, as outlined in the literature on career transformation that emphasizes mapping individual competency strengths ( El Achmar & Bhagat, 2023). Through this mapping, companies can not only screen candidates more objectively but also guide future workforce development. The 2D and 3D scatter plots generated from this clustering provide an intuitive visual representation, making it easier to make data-driven decisions. This research is unique because it applies the K-Means algorithm directly to job applicant data obtained from a construction consulting company. The data includes three main variables: ability to draw using AutoCAD, ability to create planning and supervision reports, and social adaptability. 2D and 3D scatter plots are used to visualize the clustering results, providing deeper insights into the distribution patterns of job applicants in the construction sector.

Previous studies have demonstrated the effectiveness of the K-Means algorithm in various applications. Chen Yu (2020) explains that K-Means can help identify relevant patterns and trends in various fields, including education and industry. Wiharto and Suryani (2020) compared the K-Means algorithm with Fuzzy C-Means and found that K-Means is more efficient in terms of computation time. Additionally, Agustina and Prihandoko (2018) applied K-Means to group employees based on their level of discipline and performance, providing a basis for decision-making in human resource management.

Through this research, it is hoped that the clustering results will not only help companies screen job applicants more objectively but also serve as a basis for developing data-driven recruitment policies. This research also provides insights that can be utilized by educational institutions such as vocational schools and universities to align their curricula with the specific needs of the construction sector. Thus, this research contributes to enhancing the efficiency and accuracy of workforce selection while strengthening the connection between education and the construction industry, particularly in addressing the significant challenges of national development.

 1.2 Literature review

Clustering is a technique used to group objects into specific clusters based on attribute similarities. Objects grouped into one cluster have a high degree of similarity compared to objects in other clusters. This process enables the separation of data regions so that objects with similar characteristics are in the same cluster region ( Jain et al., 1999). With this method, data variation within a group can be minimized, while differences between groups can be maximized, thereby facilitating data analysis and interpretation. Clustering is a statistical technique used to group objects into specific groups based on attribute similarity. Objects within a cluster have a higher degree of relatedness compared to objects in other clusters ( Manikandan et al., 2018). By grouping data or objects, variation within a group can be minimized, while differences between groups can be maximized. This process facilitates the interpretation of complex data, provides deeper insights, and aids in data-driven decision-making ( Darmi & Setiawan, 2016).

One of the most commonly used clustering methods is K-Means Clustering, which works by dividing data into several clusters based on attribute similarities. This method uses centroids as the center of data in each cluster. The position of the centroid will continue to be updated through iterations until stability is achieved, to minimize the distance between each data point and the cluster centroid ( Jain et al., 1999). Fadhli (2017) explains that K-Means is very effective in clustering high-dimensional data, making it a powerful tool for understanding and visualizing complex data. In the context of education, Firza and Sarjono (2020) show that the K-Means algorithm can be applied to analyze student data, such as major preferences and learning achievement evaluations.

In previous research, Wiharto and Suryani (2020) compared the K-Means algorithm with Fuzzy C-Means. The results showed that K-Means was more efficient in terms of computation time, although Fuzzy C-Means could produce higher accuracy in certain cases. Additionally, Agustina and Prihandoko (2018) applied K-Means to cluster employees based on discipline levels and performance, providing a basis for decision-making in human resource management. These studies demonstrate the flexibility of the K-Means algorithm across various applications, including in the fields of education and industry.

1.2.1 K-Means algorithm

K-Means is one of the partitioning clustering methods used in data analysis to divide data into several groups based on similar characteristics. This clustering process is performed iteratively, to minimize the average distance between each data point and its cluster center (centroid). K-Means is highly effective in grouping data because it can create clusters with a high degree of similarity within the group while distinguishing them from other clusters ( Widiyaningtyas et al., 2017).

The K-Means algorithm process begins by determining the desired number of clusters (k). Next, the cluster center or initial centroid is selected randomly to start the iteration. After the first iteration, the centroid value is updated based on the average position of the members in the cluster until the centroid position is stable. The stages of the K-Means algorithm process are generally as follows ( Purba et al., 2018):

Determine the number of clusters

The first step in the K-Means algorithm is to determine the number of clusters to be formed based on the analysis requirements. The value of k is the main parameter that must be set at the beginning of the process.

Determining the initial centroid value

The centroid, or cluster center point, for the initial iteration is selected randomly as the first step in clustering. Once the iteration begins, the centroid position will be recalculated using the following formula. v ij ¯ = 1 N i ∑ i = 0 N i x ij

Explanation:

v ij ¯ = centroid or average for cluster i on the variable j

N i = number of data points belonging to a cluster

i = cluster index

j = variable index

x ij = data value i in the cluster for variable j

Calculate the distance between the centroid and the object

This stage calculates the distance between each data point and the cluster centroid using Euclidean Distance, which is formulated as: D e = ( x i − s i ) 2 + ( y i − t i ) 2

Explanation:

D e = distance between objects and centroid

x , y = object coordinates

s , t = centroid coordinates

i = the number of objects

Grouping objects into clusters

Object data will be allocated to clusters based on the minimum distance to the centroid. Each object will have a membership value in the distance matrix, with a value of 1 if the data is allocated to a specific cluster, and 0 if it is allocated to another cluster. After the objects are grouped, the algorithm will return to the second stage to recalculate the centroid position. This process is repeated until the centroid position no longer changes and the data does not move between clusters, indicating that the grouping process is complete.

Repeat steps 3 and 4

The purpose of this repetition is to achieve a centroid position that no longer changes or to find that the centroid position difference is below the specified threshold.

1.2.2 Worker recruitment

Recruitment is a strategic process carried out by companies to obtain quality workers in line with organizational needs. This process involves searching for candidates from internal and external sources and making effective decisions based on job analysis. Job analysis is an important first step in describing the tasks, responsibilities, and qualifications required for each position, thereby helping to match applicants with company needs ( Widodo, 2018).

According to Smith and Todaro (2015), labor absorption is greatly influenced by the availability of jobs and the quality of human resources. Technical competencies such as the ability to draw using AutoCAD, prepare planning reports, and adaptability are highly valued attributes in the construction sector ( Gangl, 2003). Additionally, Green et al. (2011) state that the job search process aims to match job seekers with suitable job opportunities, which can be facilitated through technology or data-driven methods.

In facing the challenges of managing large and diverse applicant data, algorithms such as K-Means Clustering can be a solution to simplify the analysis of job applicant data. This method divides data into clusters based on attribute similarities, enabling companies to categorize applicants into groups such as Rejected, Under Consideration, and Accepted ( Jain et al., 1999). With this data-driven approach, recruitment can be conducted more efficiently and objectively, supporting companies in selecting the best candidates for their needs.

 2. Methods

The research methodology consists of several stages designed to collect and analyze data on job applicants for the Construction Consulting Company.

The overall workflow for data collection, preprocessing, clustering, and evaluation is summarized in Figure 1. This research methodology is designed to analyse job applicant data at a construction consulting firm using the K-Means Clustering algorithm. The research process begins with the collection of information and a literature review on K-Means Clustering, the use of the R program, and an analysis of relevant skills in the construction sector. The literature used includes reliable sources such as Google Scholar, research journals, and reference books.

Workflow research diagram.

This figure illustrates the overall research workflow, including data collection, literature review, data selection, centroid initialization, K-Means clustering (Excel), and subsequent evaluation and visualization using R software.

Data collection was conducted directly from construction consulting companies, covering applicants’ ability to draw using AutoCAD, their ability to compile planning reports, and their social adaptability. Irrelevant or incomplete data was filtered through a data cleaning process to ensure the accuracy of the analysis. According to Kassambara (2017), the data cleaning step is very important in cluster analysis to produce valid and reliable results.

In this study, the K-Means Clustering-based clustering method was applied to analyze job applicant data. This algorithm is highly relevant due to its ability to group applicants based on characteristics such as drawing skills, planning reports, and adaptability. As explained by Lynn et al. (2023), this type of clustering approach is often used in the context of career development to map individuals into groups based on similar attributes.

The K-Means algorithm was applied using the R program. The number of clusters was determined using the elbow method, which is a common technique for determining the optimal number of clusters in K-Means analysis. The clustering process involved calculating the Euclidean distance between the data and the cluster centroid, with iterations continuing until the cluster center stabilized. The clustering results are then evaluated using the Davies-Bouldin index to assess the validity and quality of the clusters ( Gie & Jollyta, 2020). This evaluation aims to ensure that the resulting grouping is relevant and appropriate for the analysis needs.

The clustering results are visualized using 2D and 3D scatter plots to provide a clear picture of the distribution of applicants in each cluster. Further analysis is conducted to understand the characteristics of each cluster, including the dominant abilities and work preferences of applicants. With this approach, construction consulting companies can be more effective in developing data-driven recruitment strategies.

The final stage of this research is the preparation of a report containing findings, conclusions, and recommendations. This report is expected to make a significant contribution to companies in improving the efficiency of workforce selection, as well as supporting the construction sector in facing national development challenges.

 3. Results and discussion

Data were obtained from CV Ardantama Putra Perkasa as part of their internal recruitment records. The company had posted a job vacancy on JobStreet Indonesia as the external recruitment channel; however, all applicant data used in this study originated from the company’s internal testing and selection process, not from the JobStreet platform.

In total, 161 applicants applied to the vacancy, and 30 candidates who met the minimum requirements were invited for in-person testing. The assessment consisted of three evaluation components: AutoCAD Drawing Skills (X), Ability to Prepare Planning and Supervision Reports (Y), and Adaptability (Z). Descriptive information for the 30 tested applicants is provided in Table 1.

Applicant demographic data.
Respondent code Gender AutoCAD drawing skills (X) Ability to prepare planning and monitoring reports (Y) Adaptability (Z)
Resp1 Female 92 75 68
Resp2 Male 68 65 66
Resp3 Male 73 86 87
Resp4 Male 69 74 73
Resp5 Male 78 72 91
Resp6 Female 84 90 92
Resp7 Male 69 76 87
Resp8 Female 95 73 76
Resp9 Female 90 80 85
Resp10 Male 68 82 68
Resp11 Male 63 75 71
Resp12 Male 75 93 77
Resp13 Female 62 72 68
Resp14 Male 90 61 72
Resp15 Female 84 63 90
Resp16 Female 94 70 89
Resp17 Female 73 87 80
Resp18 Female 71 73 95
Resp19 Female 93 62 70
Resp20 Male 90 68 89
Resp21 Female 87 94 87
Resp22 Male 60 90 64
Resp23 Female 65 64 93
Resp24 Male 69 84 75
Resp25 Male 66 63 72
Resp26 Male 95 85 93
Resp27 Male 75 80 83
Resp28 Male 92 85 93
Resp29 Male 71 71 85
Resp30 Male 92 61 88

With this data, the objects were grouped into three clusters with the attributes Rejected, Under Consideration, and Accepted. Next, determine the Centroid Value. The centroid value is chosen randomly, as follows: Rejected = (60,75,85), Considered = (62,77,88), Accepted = (70,84,92). The next step is to calculate the distance based on the determined Centroid point. Initial centroid selection and the corresponding Euclidean distances for each applicant are reported in Table 2.

Initial centroid distances.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 60 75 85 62 77 88 70 84 92
Resp1 92 75 68 36,24 36,11 33,78
Resp2 68 65 66 22,91 25,77 32,26
Resp3 73 86 87 17,15 14,25 6,16
Resp4 69 74 73 15,03 16,82 21,49
Resp5 78 72 91 19,21 17,03 14,46
Resp6 84 90 92 29,15 25,87 15,23
Resp7 69 76 87 9,27 7,14 9,49
Resp8 95 73 76 36,19 35,34 31,65
Resp9 90 80 85 30,41 28,32 21,56
Resp10 68 82 68 20,05 21,47 24,17
Resp11 63 75 71 14,32 17,15 23,90
Resp12 75 93 77 24,76 23,37 18,19
Resp13 62 72 68 17,38 20,62 28,00
Resp14 90 61 72 35,57 36,00 36,46
Resp15 84 63 90 27,29 26,15 25,32
Resp16 94 70 89 34,60 32,77 27,95
Resp17 73 87 80 18,38 16,88 12,73
Resp18 71 73 95 15,00 12,08 11,45
Resp19 93 62 70 38,51 38,86 38,69
Resp20 90 68 89 31,06 29,43 25,79
Resp21 87 94 87 33,08 30,25 20,35
Resp22 60 90 64 25,81 27,37 30,33
Resp23 65 64 93 14,49 14,25 20,64
Resp24 69 84 75 16,19 16,34 17,03
Resp25 66 63 72 18,68 21,63 29,27
Resp26 95 85 93 37,27 34,32 25,04
Resp27 75 80 83 15,94 14,25 11,05
Resp28 92 85 93 34,47 31,45 22,05
Resp29 71 71 85 11,70 11,22 14,80
Resp30 92 61 88 35,06 34,00 32,08

After calculating the distance matrix by entering the formula: C 1 = ( 60 + 92 ) 2 + ( 75 + 75 ) 2 + ( 85 + 68 ) 2 = 36 , 24 C 2 = ( 62 + 92 ) 2 + ( 77 + 75 ) 2 + ( 88 + 68 ) 2 = 36 , 11 C 3 = ( 70 + 92 ) 2 + ( 84 + 75 ) 2 + ( 92 + 68 ) 2 = 33 , 78

This formula is repeated or iterated until the 30th data point is filled, then the cluster members are determined according to the minimum distance from the centroid. Referring to the distance matrix, this can be seen in the values obtained in Table 3 above, which are colored red for Cluster 1, yellow for Cluster 2, and green for Cluster 3. Next, each cluster is combined or grouped into one cluster, and the second centroid point and its distance are calculated. Updated cluster memberships and second-stage centroid distances are summarized in Table 3.

Second-stage centroid values and distances.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 70,80 72,80 69,90 68,33 70,33 88,33 84,71 78,53 86,06
Resp2 68 65 66 9,16 22,96 29,40
Resp4 69 74 73 3,78 15,78 20,92
Resp10 68 82 68 9,80 23,44 24,84
Resp11 63 75 71 8,18 18,73 26,65
Resp13 62 72 68 9,04 21,36 29,74
Resp14 90 61 72 22,63 28,69 23,09
Resp19 93 62 70 24,69 31,84 24,49
Resp22 60 90 64 21,15 32,38 35,05
Resp24 69 84 75 12,44 19,10 19,97
Resp25 66 63 72 11,11 18,06 28,08
Resp7 69 76 87 17,49 5,86 15,94
Resp23 65 64 93 25,39 8,54 25,45
Resp29 71 71 85 15,21 4,32 15,67
Resp1 92 75 68 21,40 31,55 19,79
Resp3 73 86 87 21,71 16,40 13,92
Resp5 78 72 91 22,31 10,17 10,58
Resp6 84 90 92 30,96 25,41 12,94
Resp8 95 73 76 24,96 29,50 15,42
Resp9 90 80 85 25,47 23,96 5,60
Resp12 75 93 77 21,82 26,20 19,64
Resp15 84 63 90 25,97 17,38 16,04
Resp16 94 70 89 30,18 25,68 12,95
Resp17 73 87 80 17,56 19,21 15,67
Resp18 71 73 95 25,10 7,66 17,27
Resp20 90 68 89 27,50 21,80 12,15
Resp21 87 94 87 31,69 30,17 15,67
Resp26 95 85 93 35,61 30,79 14,00
Resp27 75 80 83 15,53 12,90 10,28
Resp28 92 85 93 33,64 28,23 11,97
Resp30 92 61 88 30,27 25,44 19,09

Based on the table above, the results of the calculation of the centroid point and distance point for the second stage have been obtained. It turns out that there is still a shift in cluster data, so it is necessary to calculate the centroid point and distance point for the third stage. The third-stage centroid updates and distances are presented in Table 4.

Third-stage centroid values.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 68,33 74,00 69,89 70,80 71,20 90,20 86,50 78,25 84,19
Resp2 68 65 66 9,81 25,14 29,13
Resp4 69 74 73 3,18 17,52 21,20
Resp10 68 82 68 8,23 24,85 24,87
Resp11 63 75 71 5,54 21,07 27,14
Resp13 62 72 68 6,90 23,89 30,02
Resp14 90 61 72 25,36 28,35 21,41
Resp22 60 90 64 18,98 34,01 35,32
Resp24 69 84 75 11,25 19,95 20,58
Resp25 66 63 72 11,44 20,53 28,31
Resp7 69 76 87 17,24 6,04 17,87
Resp23 65 64 93 25,40 9,66 27,26
Resp29 71 71 85 15,64 5,21 17,13
Resp5 78 72 91 23,31 7,29 12,56
Resp18 71 73 95 25,27 5,13 19,61
Resp19 93 62 70 27,43 31,39 22,53
Resp1 92 75 68 23,76 30,93 17,40
Resp3 73 86 87 21,41 15,30 15,82
Resp6 84 90 92 31,47 23,04 14,33
Resp8 95 73 76 27,38 28,12 12,92
Resp9 90 80 85 27,09 21,75 4,00
Resp12 75 93 77 21,35 25,83 20,04
Resp15 84 63 90 27,77 15,54 16,51
Resp16 94 70 89 32,25 23,26 12,14
Resp17 73 87 80 17,12 18,93 16,62
Resp20 90 68 89 29,51 19,50 11,85
Resp21 87 94 87 32,27 28,15 16,01
Resp26 95 85 93 36,96 28,00 13,98
Resp27 75 80 83 15,89 12,12 11,69
Resp28 92 85 93 34,86 25,45 12,39
Resp30 92 61 88 32,51 23,63 18,50

Based on the table above, the results of the calculation of the centroid point and distance point for the third stage have been obtained, but it turns out that there is still a shift in cluster data, so it is necessary to calculate the centroid point and distance point for the fourth stage. The fourth-stage centroid updates and distances are presented in Table 5.

Fourth-stage centroid values.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 65,63 75,63 69,63 73,00 72,14 89,71 87,80 77,60 82,80
Resp2 68 65 66 11,47 25,27 28,86
Resp4 69 74 73 5,04 17,29 21,50
Resp10 68 82 68 6,99 24,37 25,11
Resp11 63 75 71 3,03 21,41 27,59
Resp13 62 72 68 5,38 24,34 30,27
Resp22 60 90 64 16,43 33,90 35,78
Resp24 69 84 75 10,51 19,32 21,34
Resp25 66 63 72 12,85 21,13 28,37
Resp7 69 76 87 17,70 6,18 19,33
Resp23 65 64 93 26,11 11,88 28,44
Resp29 71 71 85 16,93 5,25 18,18
Resp5 78 72 91 24,96 5,16 13,95
Resp18 71 73 95 26,07 5,72 21,27
Resp3 73 86 87 21,54 14,12 17,53
Resp15 84 63 90 30,20 14,31 16,72
Resp14 90 61 72 28,52 26,96 19,93
Resp19 93 62 70 30,58 29,86 20,84
Resp1 92 75 68 26,43 28,99 15,60
Resp6 84 90 92 32,33 21,10 15,90
Resp8 95 73 76 30,17 25,94 10,92
Resp9 90 80 85 29,15 19,31 3,93
Resp12 75 93 77 21,08 24,51 20,85
Resp16 94 70 89 34,82 21,12 11,60
Resp17 73 87 80 17,07 17,75 17,75
Resp20 90 68 89 32,06 17,51 11,64
Resp21 87 94 87 33,11 26,10 16,95
Resp26 95 85 93 38,69 25,69 14,51
Resp27 75 80 83 16,91 10,53 13,02
Resp28 92 85 93 36,47 23,18 13,28
Resp30 92 61 88 35,32 22,09 17,90

Based on the table above, the results of the calculation of the centroid point and distance point for the fourth stage were obtained, but it was found that there was still a shift in the cluster data, so it was necessary to calculate the centroid point and distance point for the fifth stage. The fifth-stage centroid updates and distances are presented in Table 6.

Fifth-stage centroid values.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 66,44 76,89 70,78 73,25 73,13 88,88 89,92 76,69 83,00
Resp2 68 65 66 12,91 24,84 30,11
Resp4 69 74 73 4,45 16,46 23,35
Resp10 68 82 68 6,02 23,28 27,09
Resp11 63 75 71 3,93 20,69 29,52
Resp13 62 72 68 7,17 23,74 32,04
Resp22 60 90 64 16,10 32,85 37,86
Resp24 69 84 75 8,66 18,13 23,56
Resp25 66 63 72 13,95 20,97 29,68
Resp17 73 87 80 15,17 16,47 20,04
Resp7 69 76 87 16,45 5,46 21,31
Resp23 65 64 93 25,73 12,97 29,70
Resp29 71 71 85 16,05 4,96 19,86
Resp5 78 72 91 23,80 5,32 15,11
Resp18 71 73 95 24,95 6,53 22,71
Resp3 73 86 87 19,73 13,01 19,72
Resp15 84 63 90 29,51 14,81 16,48
Resp27 75 80 83 15,24 9,21 15,29
Resp14 90 61 72 28,44 26,69 19,16
Resp19 93 62 70 30,45 29,50 19,86
Resp1 92 75 68 25,78 28,12 15,24
Resp6 84 90 92 30,50 20,25 17,12
Resp8 95 73 76 29,29 25,28 9,40
Resp9 90 80 85 27,69 18,52 3,87
Resp12 75 93 77 19,27 23,22 22,91
Resp16 94 70 89 33,75 20,98 9,87
Resp20 90 68 89 31,08 17,52 10,56
Resp21 87 94 87 31,28 25,07 18,00
Resp26 95 85 93 37,08 25,12 13,96
Resp28 92 85 93 34,82 22,57 13,17
Resp30 92 61 88 34,67 22,35 16,60

Based on the table above, the results of the centroid point and distance point calculations for the fifth stage were obtained, but it was found that there was still a shift in the cluster data, so it was necessary to calculate the centroid point and distance point for the sixth stage. The sixth-stage centroid updates and distances are presented in Table 7.

Sixth-stage centroid values.
Respondent data Rejected Under consideration Accepted
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 67,30 78,50 71,40 73,25 73,13 88,88 91,17 75,33 83,50
Resp2 68 65 66 14,56 24,84 30,82
Resp4 69 74 73 5,07 16,46 24,56
Resp10 68 82 68 4,93 23,28 28,66
Resp11 63 75 71 5,56 20,69 30,82
Resp13 62 72 68 9,05 23,74 33,20
Resp22 60 90 64 15,50 32,85 39,58
Resp24 69 84 75 6,79 18,13 25,27
Resp25 66 63 72 15,57 20,97 30,29
Resp17 73 87 80 13,37 16,47 21,87
Resp12 75 93 77 17,35 23,22 24,81
Resp7 69 76 87 15,89 5,46 22,45
Resp23 65 64 93 26,12 12,97 30,06
Resp29 71 71 85 15,97 4,96 20,68
Resp5 78 72 91 23,26 5,32 15,52
Resp18 71 73 95 24,51 6,53 23,33
Resp3 73 86 87 18,22 13,01 21,36
Resp15 84 63 90 29,41 14,81 15,68
Resp27 75 80 83 14,00 9,21 16,83
Resp14 90 61 72 28,67 26,69 18,41
Resp19 93 62 70 30,57 29,50 19,06
Resp1 92 75 68 25,18 28,12 15,53
Resp6 84 90 92 28,91 20,25 18,40
Resp8 95 73 76 28,61 25,28 8,74
Resp9 90 80 85 26,50 18,52 5,04
Resp16 94 70 89 33,09 20,98 8,17
Resp20 90 68 89 30,58 17,52 9,24
Resp21 87 94 87 29,52 25,07 19,44
Resp26 95 85 93 35,72 25,12 14,09
Resp28 92 85 93 33,45 22,57 13,58
Resp30 92 61 88 34,52 22,35 15,05

Based on the table above, red represents Cluster 1, which represents the group of rejected individuals, yellow represents Cluster 2, which represents the group of individuals under consideration, and green represents Cluster 3, which represents the group of accepted individuals. Final cluster assignments for all applicants are listed in Table 8.

Final clustering results.
Respondent data Rejected Under consideration Accepted Clustering
C1(x 1,y 1,z 1) C2(x 2,y 2,z 2) C3(x 3,y 3,z 3)
Name MA LPP KA 67,30 78,50 71,40 73,25 73,13 88,88 91,17 75,33 83,50
Resp2 68 65 66 14,56 24,84 30,82 Cluster 1 (Rejected)
Resp4 69 74 73 5,07 16,46 24,56
Resp10 68 82 68 4,93 23,28 28,66
Resp11 63 75 71 5,56 20,69 30,82
Resp13 62 72 68 9,05 23,74 33,20
Resp22 60 90 64 15,50 32,85 39,58
Resp24 69 84 75 6,79 18,13 25,27
Resp25 66 63 72 15,57 20,97 30,29
Resp17 73 87 80 13,37 16,47 21,87
Resp12 75 93 77 17,35 23,22 24,81
Resp7 69 76 87 15,89 5,46 22,45 Cluster 2 (Under Consideration)
Resp23 65 64 93 26,12 12,97 30,06
Resp29 71 71 85 15,97 4,96 20,68
Resp5 78 72 91 23,26 5,32 15,52
Resp18 71 73 95 24,51 6,53 23,33
Resp3 73 86 87 18,22 13,01 21,36
Resp15 84 63 90 29,41 14,81 15,68
Resp27 75 80 83 14,00 9,21 16,83
Resp14 90 61 72 28,67 26,69 18,41 Cluster 3 (Accepted)
Resp19 93 62 70 30,57 29,50 19,06
Resp1 92 75 68 25,18 28,12 15,53
Resp6 84 90 92 28,91 20,25 18,40
Resp8 95 73 76 28,61 25,28 8,74
Resp9 90 80 85 26,50 18,52 5,04
Resp16 94 70 89 33,09 20,98 8,17
Resp20 90 68 89 30,58 17,52 9,24
Resp21 87 94 87 29,52 25,07 19,44
Resp26 95 85 93 35,72 25,12 14,09
Resp28 92 85 93 33,45 22,57 13,58
Resp30 92 61 88 34,52 22,35 15,05

The results of this study indicate that the application of K-Means Clustering to the test results of job applicants at a construction consulting company can be identified into several groups with different characteristics. After conducting the analysis, it was found that there were three (3) main clusters formed based on the test scores of the applicants. The first cluster consisted of applicants who exceeded the requirements and were declared Accepted, while the second cluster included applicants with test scores that met the criteria and were declared Under Consideration, and the third cluster was dominated by applicants who did not meet the criteria.

After obtaining Cluster data using K-Means in Excel, the next step is to code it in the R programming language to see a picture of multivariate statistics. The analysis process begins with processing the data of participants in the selection test for a construction consulting company. The dataset consists of three main variables, namely AutoCAD Drawing, Supervision Planning Report, and Adaptability, each of which is assessed on a scale of 0 to 100. These three variables describe the technical competencies and soft skills relevant to job selection. The next stage is the initiation of the initial centroid. The initial centroids were determined based on prior calculations taken from the analysis in Microsoft Excel. Three initial centroids were set to begin the clustering process. These initial centroid values represent the initial average values for the three planned cluster groups, namely clusters with Rejected, Under Consideration, and Accepted statuses. These initial centroids will be used as a reference to determine the membership of each data point in the cluster.

To group data into clusters, the k-means_manual function is used. This function works iteratively to calculate the Euclidean distance between each data point and the predetermined initial centroid. Based on the distance calculation results, each data point will be assigned to the cluster with the minimum distance from the centroid. Once cluster membership is determined, a new centroid is calculated as the average value of the data points included in that cluster. This process is repeated until the centroid stabilizes, i.e., when the difference between the new centroid and the previous centroid becomes very small (converges), or until the iteration reaches the specified maximum limit. This process does not yet produce a final output in the form of cluster labels or visualizations, but it has completed an important step in the K-Means algorithm, namely determining the optimal centroid. Thus, the distribution of data in each cluster can better reflect the patterns or structures that exist in the dataset.

At this stage, the clustering process of the data of participants in the entrance test for construction consulting companies is continued by adding the clustering results to the dataset. The clustering results, in the form of cluster numbers, are stored in a new column called Cluster. In addition, to facilitate interpretation of the results, categorization labels are added based on these cluster numbers. These labels are defined using the factor() function with three categories, namely “Rejected,” “Under Consideration,” and “Accepted,” according to the predetermined cluster order. Next, the serial number or respondent identity is added to the dataset in a new column named No, which is generated using the 1:nrow (participant_data) function. This is done to ensure that each row of data has a unique identity that facilitates the tracking of clustering results. The data is then rearranged to be more structured by only displaying relevant columns, namely the respondent number (No), variable values (AutoCAD Drawing, Planning and Supervision Report, Adaptability), cluster number (Cluster), and category label (Category). This rearrangement process is done using the select() function from the dplyr library. As a final result of this stage, the updated dataset with additional columns and a neat layout is printed to the console using the print() function. This dataset provides a clear overview of each respondent along with variable values and clustering results in a simple table format. With this format, the data can be used for further analysis or visualized as needed. The tidy R output showing the variables, cluster IDs, and category labels is shown in Table 9.

R-generated data table.
No AutoCAD_Drafting Planning_Supervision_ Reports Adaptability Cluster Category
1 68 65 66 1 Rejected
2 69 74 73 1 Rejected
3 68 82 68 1 Rejected
4 63 75 71 1 Rejected
5 62 72 68 1 Rejected
6 60 90 64 1 Rejected
7 69 84 75 2 Under Consideration
8 66 63 72 1 Rejected
9 73 87 80 2 Under Consideration
10 75 93 77 2 Under Consideration
11 69 76 87 3 Accepted
12 65 64 93 1 Rejected
13 71 71 85 2 Under Consideration
14 78 72 91 3 Accepted
15 71 73 95 2 Under Consideration
16 73 86 87 2 Under Consideration
17 84 63 90 3 Accepted
18 75 80 83 2 Under Consideration
19 90 61 72 2 Under Consideration
20 93 62 70 2 Under Consideration
21 92 75 68 2 Under Consideration
22 84 90 92 3 Accepted
23 95 73 76 3 Accepted
24 90 80 85 3 Accepted
25 94 70 89 3 Accepted
26 90 68 94 3 Accepted
27 87 84 87 3 Accepted
28 95 85 93 3 Accepted
29 92 85 93 3 Accepted
30 92 61 88 2 Under Consideration

The image above shows the results of running the R program after initializing the data and adding several instructions to generate clusters according to the K-Means algorithm. A two-dimensional visualization of the clustering based on AutoCAD Drawing and Planning/Supervision Reports is shown in Figure 2. Then, visualize the K-Means clusters with a 2D scatter plot. This stage visualizes the clustering results in a 2D scatter plot using ggplot2. The X-axis represents AutoCAD Drawing and the Y-axis represents Supervision Planning Report, with the color of the points indicating the clustering category: Rejected, Under Consideration, or Accepted. The plot title and axis labels were added for clarity, and a minimalist theme was applied for a clean look. This visualization helps to understand the distribution of data between clusters visually. Below is an image of the visualization.

K-means clustering visualization in a 2D scatter plot.

This plot displays job applicants based on AutoCAD Drawing scores (X-axis) and Planning/Supervision Report scores (Y-axis). Data points are assigned to three clusters—Rejected, Under Consideration, and Accepted—highlighting the separation of applicant competency profiles.

This image is a scatter plot of the clustering results using the K-Means method. The horizontal axis (X) shows the AutoCAD Drawing score, while the vertical axis (Y) represents the Supervision Planning Report score. The points on the graph are grouped into three cluster categories:

Red (Rejected): Indicates participants whose cluster scores are in the lowest criteria group.

Yellow (Under Consideration): Indicates participants with intermediate scores who are under further consideration.

Green (Accepted): Indicates participants with the highest scores who are in the highest cluster.

This distribution provides a visual representation of the distribution and relationship between the two main variables based on the clustering results categories. After 2D visualization, 3D visualization is performed. This visualization utilizes three main variables, namely AutoCAD Drawing as the x-axis, Supervision Planning Report as the y-axis, and Adaptability as the z-axis. The color of each data point is determined based on the cluster category that has been generated, with red for the “Rejected” category, orange for “Under Consideration,” and green for “Accepted.”

A three-dimensional visualization using AutoCAD Drawing (X), Planning/Supervision Reports (Y), and Adaptability (Z) is displayed in Figure 3. This 3D scatter plot depicts the clustering results of selection participants based on three main variables: AutoCAD Drawing (MA), Supervision Planning Report (LPP), and Adaptability (KA). Each variable is represented on the X, Y, and Z axes, reflecting participant scores in the range of 60 to 100. This visualization maps participants into a multidimensional space to reveal distribution patterns and relationships between variables. The points in the plot are colored according to the clusterization result category: red for participants in the “Rejected” category, orange for “Under Consideration,” and green for “Accepted.” These colors facilitate interpretation by providing a quick overview of how participants are grouped based on similar characteristics.

K-means clustering visualization in a 3D scatter plot.

This three-dimensional plot represents AutoCAD Drawing (AD), Planning/Supervision Reports (PSR), and Adaptability (A) simultaneously. The spatial separation of the three clusters demonstrates how multivariate skill combinations distinguish applicants across the Rejected, Under Consideration, and Accepted groups.

The distribution in the graph shows that participants with high scores on all three variables tend to be in the “Accepted” category, while lower scores are more often associated with the ‘Rejected’ category. The “Under Consideration” category is in the middle, reflecting participants with moderate scores who have the possibility of moving to either category depending on additional criteria. The separation between clusters is relatively clear, indicating that clustering has successfully grouped participants with similar characteristics. Some points that appear far from the center of their cluster can be identified as outliers, providing additional information about participants with unique characteristics. This visualization provides deep insights into data structures, patterns of relationships between variables, and clustering results relevant to decision making. Next, the participants’ clusters are sorted in order. The dataset sorted by cluster and category is provided in Table 10.

Sorted dataset by cluster categories.
No AutoCAD_Drafting Planning_Supervision_ Report Adaptability Cluster Category
1 68 65 66 1 Rejected
2 69 74 73 1 Rejected
3 68 82 68 1 Rejected
4 63 75 71 1 Rejected
5 62 72 68 1 Rejected
6 60 90 64 1 Rejected
7 66 63 72 1 Rejected
8 65 64 93 1 Rejected
9 69 84 75 2 Under Consideration
10 73 87 80 2 Under Consideration
11 75 93 77 2 Under Consideration
12 71 71 85 2 Under Consideration
13 71 73 95 2 Under Consideration
14 73 86 87 2 Under Consideration
15 75 80 83 2 Under Consideration
16 90 61 72 2 Under Consideration
17 93 62 70 2 Under Consideration
18 92 75 68 2 Under Consideration
19 92 61 88 2 Under Consideration
20 69 76 87 3 Accepted
21 78 72 91 3 Accepted
22 84 63 90 3 Accepted
23 84 90 92 3 Accepted
24 95 73 76 3 Accepted
25 90 80 85 3 Accepted
26 94 70 89 3 Accepted
27 90 68 89 3 Accepted
28 87 94 87 3 Accepted
29 95 85 93 3 Accepted
30 92 85 93 3 Accepted

The hierarchical clustering (PCA biplot with convex hulls) summarizing the separation of the three groups is depicted in Figure 4. After the data cluster sequence table appears, the hierarchical clustering results are displayed visually. Hierarchical clustering visualization is important because it helps to understand the hierarchical structure of the data, allowing the identification of relationships between groups based on distance or similarity. The resulting dendrogram facilitates analysis, such as determining the optimal number of clusters by cutting the tree at a certain level. In addition, this approach is useful for describing data in an intuitive way, especially when the relationships between data are not linear, thus providing deeper insights into the patterns or distribution of data in the dataset.

Hierarchical clustering visualization using PCA-projected dimensions.

This figure maps hierarchical clustering results onto principal component space, with convex hulls outlining the boundaries of the three clusters. The grouping pattern confirms consistency between hierarchical clustering and the K-Means classification of job applicants.

This hierarchical clustering visualization provides an overview of the grouping of construction consultant job applicants based on three main dimensions: AutoCAD Drawing (MA), Planning and Supervision Reports (LPP), and Adaptability (KA). These results illustrate the differences in characteristics among applicants in three clusters, which are relevant for determining job acceptance.

Cluster 1 (Red):

This cluster has characteristics that are concentrated on the left side of the graph, with scores that tend to be lower than the other clusters.

Participants in this cluster are likely to have suboptimal performance on the main criteria tested.

It can be interpreted that these participants fall into the “Rejected” category, as they do not meet the minimum standards for acceptance.

Cluster 2 (Yellow):

Located in the middle of the graph, this cluster shows participants with average performance on the main criteria.

Participants in this group have potential, but may require further evaluation or additional training to meet admission standards.

These participants fall into the “Under Consideration” category, meaning they still have a chance of being accepted.

Cluster 3 (Green):

Located at the bottom right of the graph, this cluster consists of participants with higher scores across all dimensions.

These participants demonstrate the best and most consistent performance on the main criteria, thus meeting the admission requirements.

Participants in this cluster are categorized as “Accepted.”

The results of this study show that the K-Means Clustering algorithm is an effective tool for grouping job applicants based on three main competencies: AutoCAD Drawing, Supervision Planning Reports, and Adaptability. This clustering produces three categories: “Rejected,” “Considered,” and “Accepted.” Applicants with high adaptability scores tend to fall into the ‘Considered’ or “Accepted” categories, which is in line with the findings of Brown & Hesketh (2005), who emphasize the importance of adaptability in supporting successful career transitions. Groups with strong technical skills in AutoCAD drawing and planning reports also show a higher tendency to be accepted, in line with the needs of the modern skills-based labor market.

Cluster visualization through 2D and 3D scatter plots provides intuitive insights into the distribution patterns of applicants based on their main characteristics. This graph shows that applicants with high scores on technical and adaptability variables are more likely to be in the “Accepted” cluster. As explained by Zhang et al. (2024), data visualization allows for a deeper understanding of the relationships between variables, thereby supporting evidence-based decision making. Additionally, the distribution in this graph provides a clear separation between clusters, allowing companies to focus on the group of applicants with the greatest potential. The “Under Consideration” cluster offers a unique opportunity for companies to identify candidates with great potential who require further evaluation or additional training. As highlighted by Rawat et al. (2024), targeted training can improve applicants’ abilities, ensuring that they can meet company standards in the long term. Thus, this approach not only helps companies understand the technical strengths of applicants but also provides insight into how these individuals can sustainably contribute to the organization, as also explained by Hurbean, et al. (2023).

The data-driven approach applied in this study aligns with the views of Akkermans et al. (2024), who emphasize that data-driven analysis provides a more objective and strategic basis for decision-making in the field of human resources. By basing decisions on clustering data, companies can improve the efficiency of the selection process and focus on the most suitable candidates to meet the needs of construction projects.

Overall, this study confirms that technical and adaptive abilities evaluated through clustering contribute significantly to individual and organizational success. As stated by Donald et al. (2024), this approach not only identifies applicants’ strengths but also provides strategic guidance on how to integrate those skills into organizational needs. Therefore, this clustering is relevant not only for the selection process but also for data-driven workforce development.

 4. Conclusions

The results of this study confirm that the K-Means Clustering algorithm is an effective tool for grouping job applicants at construction consulting firms based on three key competencies: AutoCAD Drawing, Supervision Planning Reports, and Adaptability. These three variables reflect the primary needs of the construction industry, where technical skills and soft skills are important indicators in determining the suitability of applicants.

The clustering process resulted in three groups: “Rejected,” “Considered,” and “Accepted.” The “Rejected” cluster includes applicants with overall scores below the expected standard, indicating significant deficiencies in one or more of the core competencies evaluated. The “Under Consideration” cluster identifies applicants with average scores, indicating potential but requiring further evaluation or additional training to meet professional standards. Meanwhile, the “Accepted” cluster includes applicants with the highest scores, indicating readiness for immediate placement in a professional work environment, as they have fully met competency expectations.

This clustering process has direct implications for decision-making in recruitment. By understanding the distribution of applicants across these three clusters, management can focus more on candidates with the best potential to contribute effectively to construction projects. Additionally, visualizing the clustering results in the form of 2D and 3D scatter plots provides deep insights into the relationships between variables, enabling companies to identify priority competencies that need to be improved through targeted training programs.

The application of the K-Means Clustering method to job applicant data offers strategic insights into the workforce selection process. These results align with Diván’s (2017) perspective that data-driven analysis can strengthen the recruitment process, particularly in identifying groups of individuals most suited to the company’s needs. Additionally, as emphasized by Rawat et al. (2024), structured training programs can play a crucial role in enhancing applicants’ capabilities within the “Considered” cluster, ensuring their alignment with the organization’s long-term needs. These findings also reflect the importance of adaptability in facing job insecurity, as outlined by Van der Heijden et al. (2024), who show that this ability contributes significantly to career sustainability and reduced labor instability.

Although this clustering approach has proven effective in identifying potential applicants, its success is highly dependent on the accuracy and completeness of the data used. Further research could explore additional variables or integrate other clustering algorithms to improve the categorization process.

The unique contribution of this research lies in the application of K-Means Clustering to job applicant data in the construction sector—an industry that heavily relies on a balance between technical skills and adaptability. This focused approach provides a framework that can be adapted by other sectors requiring alignment between recruitment processes and specific competency needs. Beyond its technical contributions, this research introduces a data-driven approach that enables companies to enhance hiring accuracy, selection efficiency, and the development of competencies aligned with industry needs. As highlighted by Donald et al. (2024) and Akkermans et al. (2024), a data-driven approach provides a solid foundation for aligning individual potential with organizational goals, supporting both short-term recruitment needs and long-term sustainable workforce management. Therefore, this approach not only addresses challenges in evaluating applicants but also builds a foundation for sustainable talent management in the construction industry.

 Ethical approval

Ethical review and approval were not required for this study because the researchers analyzed fully anonymized secondary data that had been lawfully transferred by CV Ardantama Putra Perkasa under a formal Data Usage Agreement (No. 12/X/S-K/APP/2024). According to Indonesian national research ethics regulations (Permenkes RI No. 74/2016, Article 11) and the general principles of the Declaration of Helsinki, research involving secondary anonymized non-clinical data that cannot identify individuals is exempt from institutional ethical review. Therefore, this study qualifies for an ethics exemption.

 Informed consent

Informed consent for data use was not obtained directly by the researchers, as all data were collected by CV Ardantama Putra Perkasa under standard recruitment procedures.

The company confirmed, through the Data Usage Agreement (No. 12/X/S-K/APP/2024), that job applicants had authorized the use of their anonymized recruitment test results for evaluation and administrative purposes in accordance with Indonesian data protection regulations (UU ITE and PP 71/2019). Because the researchers received only anonymized secondary data and had no access to identifiable information, this study meets the criteria for consent exemption.

 Clinical trial registration

Not applicable.

 Data availability statement Underlying data

The anonymized job applicant dataset is not publicly available due to confidentiality agreements with CV Ardantama Putra Perkasa. Researchers may request access by contacting the corresponding author at danieljesayanto.2023@student.uny.ac.id.

Access may be granted for legitimate academic research and is subject to:

signing a Data Use Agreement;

agreeing not to attempt re-identification;

compliance with Indonesian data protection regulations (UU ITE and PP 71/2019);

approval from CV Ardantama Putra Perkasa as the data owner.

Extended data

Supplementary figures, R scripts, and documentation are openly available in Zenodo at https://doi.org/10.5281/zenodo.17736891 ( Jaya, 2025) under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.

 Acknowledgement

The authors gratefully acknowledge the financial support provided by the Indonesian Endowment Fund for Education (LPDP) as the official sponsor of the scholarships that supported this publication. The authors also thank CV Ardantama Putra Perkasa for granting formal permission to use anonymized job applicant data for research and academic purposes.

 References Akkermans J Donald WE Jackson D : Are we talking about the same thing? The case for stronger connections between graduate and worker employability research. Career Dev. Int. 2024;29(1):80–92. 10.1108/CDI-08-2023-0278 Agustina N Prihandoko P : Perbandingan algoritma K-Means dengan Fuzzy C-Means untuk clustering tingkat kedisiplinan kinerja karyawan. Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi). 2018;2(3):621–626. 10.29207/resti.v2i3.492 Brown P Hesketh A : The mismanagement of talent: Employability and jobs in the knowledge economy. Ind. Labor Relat. Rev. 2005. 10.2189/asqu.2005.50.2.306 Chen Yu : K-Means clustering. Indiana University;2020. Darmi Y Setiawan A : Penerapan metode clustering K-Means dalam pengelompokan penjualan produk. Jurnal Media Infotama. 2016;12(2):148–157. Diván M : Data-driven decision making. 2017 IEEE International Conference on Technological Innovations in ICT for Agriculture and Rural Development (TIAR). IEEE;2017; pp.50–56. 10.1109/ICTUS.2017.8285973 Donald WE Van der Heijden BIJM Manville G : (Re) Framing sustainable careers: Toward a conceptual model and future research agenda. Career Dev. Int. 2024;29(5):513–526. 10.1108/CDI-02-2024-0073 El Achmar D Bhagat R : The conceptual relation between human resource management (HRM) and competency mapping. International Journal of Teaching & Education. 2023. Fadhli M : Manajemen peningkatan mutu pendidikan. Tadbir: Jurnal Studi Manajemen Pendidikan. 2017;1(2):215–240. 10.29240/jsmp.v1i2.295 Firza F Sarjono S : Penerapan algoritma K-Means dalam metode clustering untuk peminatan jurusan bagi siswa Swasta Pelita Raya Kota Jambi. Jurnal Manajemen Sistem Informasi. 2020;5(3):371–382. Gangl M : Labor market structure and re-employment rates: Unemployment dynamics in West Germany and the United States. Research in Social Stratification and Mobility. 2003;20:185–224. 10.1016/S0276-5624(03)20004-4 Gie W Jollyta D : Perbandingan Euclidean dan Manhattan untuk optimasi cluster menggunakan Davies-Bouldin Index: Status COVID-19 wilayah Riau. Prosiding Seminar Nasional Riset Information Science (SENARIS). 2020, July;2:187–191. Green AE Hoyos M Li Y : Job search study: Literature review and analysis of the Labour Force Survey. London: Department for Work and Pensions;2011. Hurbean L Miliaru F Muntean M : The impact of business intelligence and analytics adoption on decision-making effectiveness and managerial work performance. Scientific Annals of Economics and Business. 2023;70:43–54. 10.47743/saeb-2023-0012 Irmawan I Sagharmata FA Ruthriana F : Analisis dampak pembangunan Kota Hutan (Forest City) (Studi kasus: Ibu Kota Nusantara (IKN), Kalimantan). Prosiding Seminar Rekayasa Teknologi (SemResTek). 2023;299–304. Jain AK Murty MN Flynn PJ : Data clustering: A review. ACM Computing Surveys (CSUR). 1999;31(3):264–323. 10.1145/331499.331504 Jaya DJ : Supplementary Materials for “Application of K-Means Clustering for Job Applicant Analysis in Construction Firms Using R”.[Data set]. Zenodo. 2025. 10.5281/zenodo.17736891 Kassambara A : Practical guide to cluster analysis in R. STHDA; 1st ed. 2017. Reference Source London HH : Principles and techniques of vocational guidance. Ohio: Charles E. Merrill Publishing Company;1973. Manikandan S Caroline AL Kanniamma D : The study on clustering analysis in data mining. International Journal of Data Mining Techniques and Applications. 2018;7(1):46–49. Pala SK : Use and applications of data analytics in human resource management and talent acquisition. International Journal of Enhanced Research in Science, Technology & Engineering. 2021;10:2319–7463. Purba W Tamba S Saragih J : The effect of mining data K-Means clustering toward students profile model drop out potential. IOP Conference Series: Journal of Physics. 2018;1007(1):012046–012049. 10.1088/1742-6596/1007/1/012049 Rawat A Nadavulakere S Isenhour L : Career enhancement strategies, supportive work relationships and subjective career success: The moderating role of family–work conflict. Career Dev. Int. 2024;29(4):421–433. 10.1108/CDI-06-2023-0160 Smith SC Todaro MP : Economic development. Boston: Pearson Education; 12th ed. 2015. Supriyanti SS Kusmayanti JD Paluseri ARA : Pemberdayaan masyarakat sekitar di wilayah Ibu Kota Nusantara. Masyarakat Indonesia. 2023;49(1):93–102. Van der Heijden BIJM Hofer A Semeijn J : “Don’t you worry’bout a thing” – The moderating role of age in the relationship between qualitative job insecurity and career sustainability. Career Dev. Int. 2024;29(5):527–543. 10.1108/CDI-08-2023-0280 Widiyaningtyas T Prabowo MIW Pratama MAM : Implementation of K-Means clustering to distribution of high school teachers. Proceeding EECSI, Yogyakarta, 19–21 September. 2017, September;49–54. Widodo SE : Manajemen pengembangan sumber daya manusia. Yogyakarta: Pustaka Pelajar;2018. Wiharto W Suryani E : The comparison of clustering algorithms K-Means and Fuzzy C-Means for segmentation retinal blood vessels. Acta Informatica Medica. 2020;28(1):42–46. 10.5455/aim.2020.28.42-47 Zhang M Zhou S Wu Y : Pressure from social media: Influence of social media usage on career exploration. Career Dev. Int. 2024;29(1):93–112. 10.1108/CDI-01-2023-0016 10.5256/f1000research.190104.r446627 Reviewer response for version 1 Gupta Deepak 1 Referee https://orcid.org/0000-0003-0069-7488 Penn State University, University Park, PA, USA

Competing interests: No competing interests were disclosed.

 3 2 2026 Copyright: © 2026 Gupta D 2026 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

The manuscript addresses a relevant and practically important topic of using K‑Means clustering to support recruitment decisions in a construction consulting firm. The use of real organizational data, clear cluster descriptions, and intuitive 2D/3D visualizations makes the work accessible and potentially useful for practitioners interested in HR analytics and R-based implementations. However, the current version has substantial methodological and presentation limitations that need to be addressed before the article can be considered for indexing.

The study starts from 161 applicants but analyzes only 30 shortlisted candidates who “met the minimum requirements,” with no clear justification of the selection criteria or their implications. This design means the clustering is applied to a highly pre-filtered subset rather than the full applicant pool, which severely limits generalizability to “recruitment strategies” or “workforce selection” more broadly. The authors should explicitly describe the screening rules used to select the 30 candidates (cut‑offs, qualitative judgments, etc.). The authors should re-run the analysis on the full set of 161 applicants (if possible and ethically permissible) or reframe the paper as a methodological demonstration on a small, filtered sample and tone down claims about recruitment optimization.

Initial centroids are chosen manually as “Rejected = (60,75,85); Considered = (62,77,88); Accepted = (70,84,92)” with no methodological justification beyond being “chosen randomly,” which they are not. These values embed prior expectations about the three groups and risk steering the solution toward a desired structure, rather than letting the algorithm discover patterns in the data. The authors should justify the chosen centroids and make that explicit. In addition, run at least one standard initialization strategy (e.g., built‑in k-means in R with multiple random starts, or k-means++). The authors should also report whether the manually initialized solution is stable across different initializations; and if not, the conclusions about three clusters and their interpretation need to be more cautious.

The methods mention Euclidean distance and cite general clustering references, but there is no clear statement on whether the three variables were standardized or used in raw form. There is also no discussion of data cleaning beyond a vague reference to removing “irrelevant or incomplete data.” How many records, if any, were excluded and why? Whether any outliers were identified and how they were handled. The authors should explicitly state whether they standardized the variables before clustering; if not, justify why not and show at least descriptive ranges/variances. They should also provide a proper description of data-cleaning procedure. If outliers exist (some points in the 3D plot appear distant from cluster centers), please clarify whether they influence centroid placement and consider sensitivity analyses.

The discussion and conclusions repeatedly assert that K‑Means is “effective” and “improves recruitment decision-making,” yet there is no external validation. There is no comparison with actual hiring decisions made by the firm. No HR-expert assessment of whether clusters align with real performance or selection outcomes. If such data are not available, please qualify all claims about decision-making benefits as exploratory.

The author states in the “Underlying data” section that the dataset is not publicly available due to confidentiality and must be requested via the corresponding author. At the same time, the text refers to R scripts and supplementary materials in a Zenodo repository as extended data. The authors should clarify precisely what is in the Zenodo archive (R code, example data, figures, documentation) and referencing it consistently in the methods and data availability sections. If contractual and legal constraints allow, please consider publishing an anonymized dataset (at least the 30 shortlisted candidates; ideally all 161) with no identifying information. If this is impossible, explicitly justify the restriction and explain how others can still verify the methodology (e.g., via synthetic data included in the Zenodo package).

The conclusions and parts of the discussion are phrased quite strongly (e.g., “confirm that the K‑Means Clustering algorithm is an effective tool” and “this approach not only addresses challenges in evaluating applicants but also builds a foundation for sustainable talent management”), which is not fully justified given the small, preselected sample and lack of external validation.

Minor comments: The authors are suggested to modify the text in the abstract, discussion, and conclusions to reflect that findings are preliminary and illustrative for one firm and one small dataset.

The literature review is generally coherent but heavily focused on clustering and Indonesian/sector-specific references. There is relatively little engagement with the broader international work on AI in recruitment, algorithmic decision-making, or HR analytics.

Tables 2-7 list centroid updates and distances in excessive detail. It is pedagogically useful but makes the manuscript long and difficult to navigate. Some of the tables can be moved to the supplementary information with their summary in the main text.

There are some grammatical errors, please perform a proofread of the manuscript carefully.

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

Healthcare, 3D Printing, Artificial Intelligence

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Ramdhani Wahyu Muhammad Educational Research and Evalu, Universitas Negeri Yogyakarta, Yogyakarta, Special Region of Yogyakarta, Indonesia

Competing interests: The authors declare that they have no competing interests.

 6 2 2026

Response to Reviewer 2

Deepak Gupta (Penn State University, USA)

We thank the reviewer for the detailed and technically informed evaluation of our manuscript. The feedback has been instrumental in improving methodological rigor, transparency, and interpretive clarity. Our responses are provided below.

Comment 1: Pre-filtered sample and limited generalizability

Response:

We fully agree. The manuscript has been explicitly reframed to clarify that the clustering analysis was applied to a pre-screened subset of applicants who passed document screening. This design choice is now clearly justified in the Methods section, and all claims have been moderated accordingly. The study is presented as an exploratory analytical illustration rather than a full recruitment optimization model.

Comment 2: Manual centroid bias and need for robustness checks

Response:

This concern has been directly addressed. In addition to clarifying the rationale for initial centroid selection, we now report results from multiple random initializations (nstart = 50) using R’s kmeans() function. The three-cluster structure remained largely stable across runs, with only minor variation among borderline profiles. These findings are now reported and discussed in the revised manuscript.

Comment 3: Data normalization, outliers, and cleaning

Response:

The Methods section now explicitly states that all variables were measured on the same 0–100 scale, and therefore raw scores were used without normalization. We also added an outlier and sensitivity analysis, including a leave-one-out test removing the most distant observation. This test did not materially change the cluster structure, and validation metrics showed only marginal changes. These results are now reported to support robustness.

Comment 4: Lack of validation against real hiring outcomes

Response:

We explicitly acknowledge this limitation. The revised manuscript clearly states that no comparison with actual hiring decisions or post-employment performance was possible. All statements regarding recruitment support are now framed as illustrative and exploratory, not empirically validated improvements.

Comment 5: Data availability and Zenodo clarification

Response:

This issue has been resolved by clearly distinguishing:

Confidential underlying data (not publicly shareable);

Publicly available extended data hosted on Zenodo, including R scripts, figures, clustering outputs, and documentation.

 The Data availability section has been rewritten for clarity and compliance with journal policy.

Comment 6: Excessive tables and manuscript length

Response:

In line with the suggestion, detailed centroid iteration tables have been moved to extended data. The main text now focuses on final clustering outcomes, visual interpretation, and discussion, substantially improving readability.

Comment 7: Language and proofreading

Response:

The manuscript has been carefully proofread and revised to improve clarity, reduce repetition, and ensure consistent terminology throughout.

Final note:

We thank the reviewer for the rigorous and constructive critique. The revisions have significantly strengthened the manuscript’s technical soundness, transparency, and responsible positioning within data-driven recruitment research.

 10.5256/f1000research.190104.r446632 Reviewer response for version 1 Shaikh Sonia Najam 1 Referee https://orcid.org/0000-0001-7365-1064 Jiangsu University, Zhenjiang, China

Competing interests: No competing interests were disclosed.

 8 1 2026 Copyright: © 2026 Shaikh SN 2026 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

The paper presents an interesting and practically valuable idea by using K-Means clustering to support recruitment decision-making in a construction firm. The study groups applicants into rejected, under consideration, and accepted categories based on AutoCAD skills, planning/supervision report ability, and adaptability, and it makes good use of R and visualization techniques to demonstrate the results. The topic is timely and relevant to HR analytics and industry needs, and overall the manuscript is readable, logically structured, and supported by real organizational data, which gives it strong applied significance.

However, while the work is clearly presented, it is only partly sufficient in terms of scientific rigor and presentation quality. The manuscript sometimes becomes repetitive and uses very long tables that make it difficult to follow the narrative smoothly. The literature review is decent, but it leans heavily on regional or context-specific studies, and it would benefit greatly from integrating more current international literature on AI-driven recruitment, algorithmic hiring, fairness, and HR analytics trends. Adding such perspectives would make the study more globally relevant and intellectually grounded.

In terms of study design, the concept is appropriate, but there are some important weaknesses that affect technical soundness. The biggest issue is that the study only analyzed 30 shortlisted candidates out of 161 applicants, without providing a clear justification. This introduces bias and limits how generalizable the findings are. Another concern is the selection of initial centroids, which appears subjective rather than determined using a standard procedure such as random initialization or k-means++. There is also no explanation of whether data normalization was applied, which is important since K-Means relies on Euclidean distance and differences in scale can significantly affect results. These issues really need to be addressed for the paper to be scientifically strong.

The paper partly allows replication because the clustering process is explained carefully and tables clearly trace centroid iterations and cluster formation. However, critical methodological details are missing. There is no detailed explanation of data cleaning, whether any applicants were excluded and why, how outliers were treated, whether ethical permission or company approval was obtained, and what settings were used in R. To make the study convincingly reproducible, the authors should provide anonymized raw data, the R script, and a clearer explanation of all methodological choices.

The analysis and interpretation are also only partly sufficient. Although clustering has been correctly applied and visualizations are very good, the study does not present objective validation of the clusters. There is no numerical reporting of validation metrics such as silhouette score, Davies–Bouldin index, or any form of comparison with real hiring outcomes or expert HR evaluation. Without such validation, the findings are mostly descriptive, and the claim that clustering improves recruitment decision-making is not empirically proven. Adding cluster validity results and some form of comparison or evaluation would significantly strengthen the credibility of the conclusions.

Data availability is partly fulfilled. While applicant scores are shown in tables, there is no formal dataset sharing, no supplementary files, and no data availability statement. Reproducibility is an important expectation, so providing anonymized data and the analysis code will help readers trust and build upon the work.

The conclusions are relevant and aligned with the results, but they are currently expressed too strongly considering the methodological limitations and lack of validation. The study needs to acknowledge its limitations, including the small selective sample, potential bias, lack of external validation, and ethical concerns such as fairness and transparency in automated hiring. The tone of the conclusions should be slightly moderated to reflect that the findings are promising but not definitively proven.

Overall, this is a promising and meaningful study with clear practical relevance, strong visualization, and a good foundation. However, to make it scientifically sound and suitable for indexing, the authors must justify or expand the sample, apply standard centroid initialization or explain theirs clearly, normalize or justify scale handling, provide quantitative cluster validation, share data and code, strengthen literature support, and include a more explicit discussion of limitations and ethical considerations. 

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

Human Resource Management, Artificial intelligence, Big data analytics capability

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Ramdhani Wahyu Muhammad Educational Research and Evalu, Universitas Negeri Yogyakarta, Yogyakarta, Special Region of Yogyakarta, Indonesia

Competing interests: The authors declare that they have no competing interests.

 6 2 2026

Response to Reviewer 1

Sonia Najam Shaikh (Jiangsu University, China)

We sincerely thank the reviewer for the careful reading of our manuscript and for the constructive and detailed feedback. We have revised the manuscript substantially to address all major concerns raised. Our point-by-point responses are provided below.

Comment 1: Limited international literature and lack of broader AI/HR analytics perspective

Response:

Thank you for this important suggestion. We have expanded the literature review and discussion to better situate the study within international debates on data-driven recruitment, employability, and HR analytics. Specifically, we incorporated recent global literature addressing analytical decision support, adaptability, and career sustainability (e.g., Akkermans et al., 2024; Donald et al., 2024; Rawat et al., 2024; Zhang et al., 2024; Van der Heijden et al., 2024).

In addition, we explicitly positioned clustering as an exploratory decision-support tool, aligned with international guidance on responsible and transparent use of analytics in employment contexts.

Comment 2: Analysis limited to 30 candidates out of 161 applicants

Response:

We fully agree that this required clearer justification. The revised manuscript now explicitly explains that the clustering analysis was conducted on 30 candidates who passed formal document screening conducted by the company as part of standard recruitment procedures. We clearly state that the study analyzes a pre-screened assessment sample, not the full applicant pool.

Accordingly, the study has been reframed as an exploratory methodological illustration rather than a comprehensive recruitment optimization model. Limitations related to selection bias and generalizability are now explicitly acknowledged in the Methods, Results, and Conclusions sections.

Comment 3: Subjective centroid selection and lack of standard initialization

Response:

Thank you for highlighting this critical methodological issue. We have clarified that the manually specified initial centroids were derived from preliminary inspection of score distributions and used for transparency and pedagogical traceability.

To address concerns regarding subjectivity and robustness, we have now:

Added a dedicated subsection describing multiple random initializations using the built-in kmeans() function in R (nstart = 50);

Reported stability checks, showing that the three-cluster structure was largely consistent across runs, with only minor variations among borderline cases;

Moderated all interpretations to emphasize that clusters reflect similarity patterns, not predefined hiring categories.

Comment 4: Missing details on data cleaning, ethics, and reproducibility

Response:

This concern has been fully addressed. The revised manuscript now includes:

A clearer description of document screening and data preprocessing;

Explicit confirmation that all analyzed records were complete and measured on the same 0–100 scale;

A detailed Ethical approval and informed consent section clarifying the use of fully anonymized secondary data under a formal Data Usage Agreement;

An updated Data availability statement;

An updated Zenodo repository containing R scripts, clustering outputs, and documentation to support reproducibility.

Comment 5: Lack of objective cluster validation

Response:

We agree with this observation. To strengthen technical rigor, we have added a dedicated cluster validity assessment section reporting internal validation metrics. Specifically, we report:

Mean silhouette coefficient (0.16);

Davies–Bouldin Index (DBI = 2.05).

 We explicitly interpret these values as diagnostic indicators of modest separation, appropriate for exploratory analysis, and avoid presenting them as evidence of predictive effectiveness.

Comment 6: Overly strong conclusions

Response:

The Conclusions section has been revised to adopt a more cautious and balanced tone. Claims regarding effectiveness and decision-making improvement are now framed as exploratory and context-specific. Limitations related to sample size, pre-screening, lack of external validation, and ethical considerations are explicitly acknowledged.

Final note:

We sincerely appreciate the reviewer’s thoughtful comments, which have significantly strengthened the manuscript’s methodological transparency, analytical caution, and international relevance.