Towards Effective and Interpretable Person-Job Fitting

Le, Ran; Hu, Wenpeng; Song, Yang; Zhang, Tao; Zhao, Dongyan; Yan, Rui

doi:10.1145/3357384.3357949

Cited by 40 publications

(29 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many studies follow a similar approach as language models such as BERT [33] or ELMo [97], and extend on these embeddings by adding additional hierarchical attention networks [91,100]. Also embeddings based on CNNs are common [70], or mixtures of the above approaches [84,13,14,58,125]. Siamese neural networks have also been used for this purpose [107].…”

Section: Model-based Methods On Shallow Embeddingsmentioning

confidence: 99%

“…For Type 2 behavioral feedback, common strategies for defining negatives include using shown but skipped items [54,59,70,79,89,91,94,99,101,120,126,130,133,47], picking negative samples at random (not per se uniform) [13,124,125,73,29], replacing the job (but not the candidate and further context) at random [132], using vacancies of which the vacancy details were shown, but did not lead to an application [101,100], or if the method allows for sparse matrices (such as in some matrix factorization methods): using all possible vacancy-user interactions [74,98,71,68,19,102,4,72,80,22,15,83,105]. Others incorporate negative sampling into the estimation method itself [76,14].…”

Section: Choices In Negative Samplingmentioning

confidence: 99%

See 1 more Smart Citation

Job Recommender Systems: A Review

Ruijt¹,

Bhulai²

2021

Preprint

View full text Add to dashboard Cite

This paper provides a review of the job recommender system (JRS) literature published in the past decade (2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020)(2021). Compared to previous literature reviews, we put more emphasis on contributions that incorporate the temporal and reciprocal nature of job recommendations. Previous studies on JRS suggest that taking such views into account in the design of the JRS can lead to improved model performance. Also, it may lead to a more uniform distribution of candidates over a set of similar jobs. We also consider the literature from the perspective of algorithm fairness. Here we find that this is rarely discussed in the literature, and if it is discussed, many authors wrongly assume that removing the discriminatory feature would be sufficient. With respect to the type of models used in JRS, authors frequently label their method as 'hybrid'. Unfortunately, they thereby obscure what these methods entail. Using existing recommender taxonomies, we split this large class of hybrids into subcategories that are easier to analyse. We further find that data availability, and in particular the availability of click data, has a large impact on the choice of method and validation. Last, although the generalizability of JRS across different datasets is infrequently considered, results suggest that error scores may vary across these datasets.

show abstract

Section: Model-based Methods On Shallow Embeddingsmentioning

confidence: 99%

Section: Choices In Negative Samplingmentioning

confidence: 99%

Job Recommender Systems: A Review

Ruijt¹,

Bhulai²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Memory networks are adopted to embed the preference in the representations of the candidate and the job post. Le et al [17] define the intention of the candidate and the recruiter according to the actions including submitting a resume, accepting a resume and rejecting a resuming. Next, they train the representations of the job post and the resume to predict the intention rates and matching score simultaneously.…”

Section: Person-job Fitmentioning

confidence: 99%

“…The extracted entities are human readable, which help explain the matching results. Our method also exploits the actions as [17] to infer the preference (or intention) of candidates and recruiters; nonetheless, we learn the representations by accumulating all actions of each candidate and each recruiter instead of a single action. Our final representation of a candidate or job post is a fusion of the representations for the explicit and implicit intentions.…”

Section: Person-job Fitmentioning

confidence: 99%

“…Recently, deep learning models have largely improved the performance of natural language processing tasks, including semantic matching [12,22] and question answering. Deep-learning-based methods [4,17,23,32] are consequently introduced for person-job fit, focusing on learning effective representations of the free text of the job post and resume. The learned representations are then compared to generate a matching score.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Learning Effective Representations for Person-Job Fit by Feature Fusion

Jiang¹,

Ye²,

Wang

et al. 2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

Person-job fit is to match candidates and job posts on online recruitment platforms using machine learning algorithms. The effectiveness of matching algorithms heavily depends on the learned representations for the candidates and job posts. In this paper, we propose to learn comprehensive and effective representations of the candidates and job posts via feature fusion. First, in addition to applying deep learning models for processing the free text in resumes and job posts, which is adopted by existing methods, we extract semantic entities from the whole resume (and job post) and then learn features for them. By fusing the features from the free text and the entities, we get a comprehensive representation for the information explicitly stated in the resume and job post. Second, however, some information of a candidate or a job may not be explicitly captured in the resume or job post. Nonetheless, the historical applications including accepted and rejected cases can reveal some implicit intentions of the candidates or recruiters. Therefore, we propose to learn the representations of implicit intentions by processing the historical applications using LSTM. Last, by fusing the representations for the explicit and implicit intentions, we get a more comprehensive and effective representation for person-job fit. Experiments over 10 months real data show that our solution outperforms existing methods with a large margin. Ablation studies confirm the contribution of each component of the fused representation. The extracted semantic entities help interpret the matching results in the case study.

show abstract

Beyond Matching: Modeling Two-Sided Multi-Behavioral Sequences for Dynamic Person-Job Fit

Liu

Zhu

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Towards Effective and Interpretable Person-Job Fitting

Cited by 40 publications

References 20 publications

Job Recommender Systems: A Review

Job Recommender Systems: A Review

Learning Effective Representations for Person-Job Fit by Feature Fusion

Beyond Matching: Modeling Two-Sided Multi-Behavioral Sequences for Dynamic Person-Job Fit

Contact Info

Product

Resources

About