An Explainable Person-Job Fit Model Incorporating Structured Information

Zhang, Yunchong; Liu, Baisong; Qian, Jiangbo; Qin, Jiangcheng; Zhang, Xueyuan; Jiang, Xueyong

doi:10.1109/bigdata52589.2021.9672057

Cited by 4 publications

(2 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Natural language sentences consist of words and phrases that follow grammatical rules and convey complex semantic information, involving both sequential and hierarchical elements that are crucial for understanding them. Existing research on person-job fit predominantly employs the Word2Vec method for sentence representations [27][28][29][30][31] but overlooked word positional information within sentences. Furthermore, some studies employ Doc2Vec to capture paragraph semantics and syntax but encounter quality issues due to limited data [32].…”

Section: Text Representation Learningmentioning

confidence: 99%

Your Profile Reveals Your Traits in Talent Market: An Enhanced Person-Job Fit Representation Learning

Jian,

Rao,

2024

Preprint

View full text Add to dashboard Cite

Person-job fit lies at the heart of online recruitment, measuring the compatibility between job seekers and vacancies. Current researchers mainly focus on job-resume matching between job requirements and work experiences, presenting two notable limitations. Free text representation is constrained by word-level polysemy and sentence-level anisotropy. Also, overreliance on final representations hampers the exploration of inter-feature relationships. Towards this end, we proposed a novel Attentive Person-Job Fit Multifaceted feature Fusion model (APJFMF), aiming at obtaining more precise and comprehensive interactive person-job fit feature representations. The main contributions are: (1) Introduction of a method for multifaceted feature extraction and fusion from multi-source heterogeneous person-job data; (2) Enhanced the person-job free text representations through unsupervised fine-tuning of BERT; (3) Investigated global feature interactions by integrating diverse attention mechanisms. Extensive experiments on a real-world recruitment dataset confirmed the effectiveness of APJFMF and its individual components. The code is released at https://github.com/raochongzhi/APJFMF.

show abstract

Section: Text Representation Learningmentioning

confidence: 99%

Your Profile Reveals Your Traits in Talent Market: An Enhanced Person-Job Fit Representation Learning

Jian,

Rao,

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…However, the path to personalized recommendation often entails a trade-off between precision and privacy. Recurrently, user interaction history datasets, prediction models, and recommendation results exploited by RSs may inadvertently leak user privacy [18]. Moreover, such leakage can be exacerbated in scenarios necessitating the exchange of these datasets between users and servers.…”

Section: Related Workmentioning

confidence: 99%

Differentially Private Knowledge Graph Neural Network for Recommeder System

胡,

刘

2023

Preprint

View full text Add to dashboard Cite

Knowledge Graphs (KGs) have been instrumental in mitigating the challenges of cold start and data scarcity in recommender systems, serving as a kind of auxiliary information. However, recent studies reveal that KGs, while beneficial, may also inadvertently furnish malicious attackers with more comprehensive background knowledge, escalating the risk of user privacy data breaches within these recommender systems. Current frameworks, which amalgamate KGs with privacy-preserving techniques, predominantly concentrate on enhancing privacy preservation, often at the expense of recommendation efficacy. In this research, we introduce a framework, named Differential Privacy Knowledge graph Neural network for Recommender systems (DPKNRec), designed to harmonize the trade-off between differential privacy noise and recommendation performance. Firstly, we formulate an integrated KG convolutional neural network adept at mining interrelationships between KG entities and item correlations effectively. Secondly, we incorporate a differential privacy technique to impose uniform Laplacian noise on feature embeddings and obfuscate genuine user-item interactions through pseudo-interaction item sampling, thereby augmenting user data security. Lastly, we evaluate our model using three distinct datasets-movies, music, and books. Empirical results from three benchmarks indicate that DPKNRec markedly supersedes contemporary benchmarks.

show abstract