Automatic Feature Recommendation System for Job Postings Using Unsupervised Clustering and Association Rule Mining

Abstract

All sectors of business use online employment platforms to publish job postings to find the qualified personnel they need. A job posting text should be neither long enough to make it difficult to read, nor short enough to create a feeling of carelessness, and should contain clear information about the position. However, the preparation of a proper job posting text that complies with the qualifications needed by the job requires expertise and experience. Here, we propose a feature recommendation system to assist recruiters who prepare job posting text. The proposed system mines the relations between the features on a job posting dataset obtained from Kariyer.Net. We used text-mining and natural language processing techniques and employed the DBSCAN clustering algorithm to obtain feature clusters. In addition, we used the Apriori algorithm for finding association rules and frequent item sets to discover relations between feature clusters. Furthermore, we developed a user interface to experiment with feature suggestion results on selected job titles. Our results demonstrate that using such a recommender system makes the job posting preparation process fast and easy and increases the quality of the job postings created.

Keywords

• Artificial Intelligence
• Association Rule Mining
• Clustering Algorithms
• Natural Language Processing
• Text Mining

Denetimsiz Kümeleme ve İlişkisel Kural Analizi Kullanılarak İş İlanları için Otomatik Özellik Öneri Sistemi

Abstract

Tüm sektörlerdeki işletmeler, ihtiyaç duydukları nitelikli personeli bulmak için iş ilanları yayınlamak üzere çevrimiçi istihdam platformlarını kullanmaktadır. Bir iş ilanı metni, okunmasını zorlaştıracak kadar uzun veya özensizlik hissi uyandıracak kadar kısa olmamalı ve pozisyon hakkında net bilgiler içermelidir. Ancak, işin gerektirdiği niteliklere uygun, doğru bir iş ilanı metni hazırlamak uzmanlık ve deneyim gerektirir. Bu çalışmada, iş ilanı metni hazırlayan insan kaynakları uzmanlarına yardımcı olmak için bir özellik önerme sistemi geliştirilmesi hedeflenmiştir. Geliştirilen sistem ile, Kariyer.Net'ten alınan iş ilanı verisetindeki özellikler kümelenmiş ve aralarındaki ilişkiler incelenmiştir. Özellik kümelerinin oluşturulması için metin madenciliği ve doğal dil işleme tekniklerinden yararlanılmış ve DBSCAN kümeleme algoritması kullanılmıştır. Ayrıca, özellik kümeleri arasındaki ilişkileri keşfetmek, ilişki kurallarını ve birlikte sık kullanılan özellikleri belirlemek amacıyla Apriori algoritması tercih edilmiştir. Belirtilen iş alanına uygun özellik önerilerinin sunulması amacıyla bir kullanıcı arayüzü geliştirilmiştir. Sonuçlarımız, yapay zeka destekli bir öneri sisteminin kullanılmasının, iş ilanı hazırlama sürecini hızlı ve kolay hale getirdiğini ve oluşturulan iş ilanlarının kalitesini artırdığını göstermektedir.

Keywords

• Yapay Zeka
• İlişkisel Kural Analizi; Kümeleme Algoritmaları
• Doğal Dil İşleme
• Metin Madenciliği

References

1. Califf ME, Mooney RJ. Bottom-up relational learning of pattern matching rules for information extraction. J Mach Learn Res 2003;4:177-210. doi:10.1162/153244304322972685.
2. Muludi K, Widyantoro DH, Santoso OS. Multi-inductive learning approach for information extraction. In: Proceedings of the International Conference on Electrical Engineering and Informatics, IEEE, Bandung, Indonesia; 2011, p. 1-6. doi:10.1109/ICEEI.2011.6021680.
3. Lavelli A, Califf ME, Ciravegna F, Freitag D, Giuliano C, Kushmerick N, et al. Evaluation of machine learning-based information extraction algorithms: criticisms and recommendations. Lang Resour Eval 2008;42(4):361-393. doi:10.1007/s10579-008-9079-3.
4. Elgammal Z, Barmu A, Hassan H, Elgammal K, Özyer T, Alhajj R. Matching applicants with positions for better allocation of employees in the job market. In: Proceedings of the 22nd International Arab Conference on Information Technology (ACIT), IEEE, Muscat, Oman; 2021, p. 1-5. doi:10.1109/ACIT53391.2021.9677374.
5. Sandanayake TC, Limesha GAI, Madhumali TSS, Mihirani WPI, Peiris MSA. Automated CV analyzing and ranking tool to select candidates for job positions. In: Proceedings of the 6th International Conference on Information Technology: IOT and Smart City, Osaka, Japan; 2018, p. 13-18. doi:10.1145/3301551.3301579.
6. Mhamdi D, Azzouazi M, El Ghoumari MY, Moulouki R, Rachik Z. Enhancing recruitment process using semantic matching. In: Proceedings of the International Conference on Advanced Intelligent Systems for Sustainable Development, Springer, Tangier, Morocco; 2020, p. 370-378. doi:10.1007/978-3-030-90639-9_29.
7. Menacer MA, Hamda FB, Mighri G, Hamidene SB, Cariou M. An interpretable person-job fitting approach based on classification and ranking. In: Proceedings of the 4th International Conference on Natural Language and Speech Processing (ICNLSP), Trento, Italy; 2021, p. 130-138.
8. Ahmed Awan MN, Khan S, Latif K, Khattak AM. A new approach to information extraction in user-centric e-recruitment systems. Appl Sci 2019;9(14):2852. doi:10.3390/app9142852.

9. Mahbub S, Pardede E. Using contextual features for online recruitment fraud detection. In: Proceedings of the 27th International Conference on Information Systems Development, Lund, Sweden; 2018.
10. Nasser IM, Alzaanin AH, Maghari AY. Online recruitment fraud detection using ANN. In: Proceedings of the Palestinian International Conference on Information and Communication Technology (PICICT), IEEE, Gaza, Palestine; 2021, p. 13-17. doi:10.1109/PICICT53635.2021.00015.
11. Mehboob A, Malik MSI. Smart fraud detection framework for job recruitments. Arab J Sci Eng 2021;46(4):3067-3078. doi:10.1007/s13369-020-04998-2.
12. Chern A, Liu Q, Chao J, Goindani M, Javed F. Automatically detecting errors in employer industry classification using job postings. Data Science and Eng 2018;3(3):221-231. doi:10.1007/s41019-018-0071-7.
13. Wu X, Xu T, Zhu H, Zhang L, Chen E, Xiong H. Trend-aware tensor factorization for job skill demand analysis. In: Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI), Macao, China; 2019, p. 3891-3897. doi:10.24963/ijcai.2019/540.
14. Xu T, Zhu H, Zhu C, Li P, Xiong H. Measuring the popularity of job skills in recruitment market: A multi-criteria approach. In: Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, Louisiana, USA; 2018, p. 2572-2579. doi:10.1609/aaai.v32i1.11847.
15. Li S, Shi B, Yang J, Yan J, Wang S, Chen F, et al. Deep job understanding at LinkedIn. In: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, Xian, China; 2020, p. 2145-2148. doi:10.1145/3397271.3401403.
16. Cao L, Zhang J, Ge X, Chen J. Occupational profiling driven by online job advertisements: Taking the data analysis and processing engineering technicians as an example. Plos One 2021;16(6):e0253308. doi:10.1371/journal.pone.0253308.
17. Xu G. Human resource recommendation based on recurrent convolutional neural network. In: Proceedings of the 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE), IEEE, Guangzhou, China; 2022, p. 54-58. doi:10.1109/ICCECE54139.2022.9712799.
18. Wild S, Parlar S, Hanne T, Dornberger R. Naïve Bayes and named entity recognition for requirements mining in job postings. In: Proceedings of the 3rd International Conference on Natural Language Processing (ICNLP), IEEE, Beijing, China; 2021, p. 155-161. doi:10.1109/ICNLP52887.2021.00032.
19. Guo S, Alamudun F, Hammond T. RésuMatcher: A personalized résumé-job matching system. Expert Syst Appl 2016;60:169-182. doi:10.1016/j.eswa.2016.04.013.
20. Çelik D. Towards a semantic-based information extraction system for matching résumés to job openings. Turk J Electr Eng Comput Sci 2016;24(1):141-159. doi:10.3906/elk-1304-130.
21. Tuncer I, Kara KC, Karakaş A. Improving search relevance with word embedding based clusters. In: Proceedings of the International Conference on Artificial Intelligence and Applied Mathematics in Engineering, Springer, Antalya, Turkey; 2020, p. 15-24. doi:10.1007/978-3-030-79357-9_3.
22. Kara KC, Esen S, Kahyalar N, Karakaş AA, Aytekin T. Design and implementation of a job recommender system. In: Proceedings of the International Conference on Computer Science and Engineering (UBMK), IEEE, Antalya, Turkey; 2017, p. 729-733. doi:10.1109/UBMK.2017.8093515.
23. Akın AA, Akın MD. Zemberek, An open source NLP framework for Turkic languages. Structure 2007;10:1-5.
24. Ester M, Kriegel HP, Sander J, Xu X. A density-based algorithm for discovering clusters in large spatial databases with noise. In: Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining, AAAI Press, Portland, Oregon, USA; 1996, p. 226-231.
25. Agrawal R, Imielinski T, Swami A. Mining association rules between sets of items in large databases. ACM SIGMOD Record 1993;22(2):207-216. doi:10.1145/170035.170072.
26. Agrawal R, Srikant R. Fast algorithms for mining association rules in large databases. In: Proceedings of the 20th International Conference on Very Large Data Bases, Santiago de Chile, Chile; 1994, p. 487-499.
27. Ilayaraja M, Meyyappan T. Mining medical data to identify frequent diseases using Apriori algorithm. In: Proceedings of the International Conference on Pattern Recognition, Informatics and Mobile Engineering, Salem, India; 2013, p. 194-199. doi:10.1109/ICPRIME.2013.6496471.
28. Yang Y, Yuan ZZ, Sun DY, Wen XL. Analysis of the factors influencing highway crash risk in different regional types based on improved Apriori algorithm. Adv in Transportation Studies 2019;49:165-178.
29. Dalkılıç F, Aydın Ö. The factors affecting the non-attendance behaviors of the students in the Faculty of Economics and Administrative Sciences in Dokuz Eylül University. Journal of Higher Education and Science 2017;7(3):546-553. doi:10.5961/jhes.2017.231.
30. Manning CD, Raghavan P, Schutze H. Introduction to information retrieval. New York: Cambridge; 2008, p. 121-126.