Abstract:Job recommender systems are crucial for aligning job opportunities with job-seekers in online job-seeking. However, users tend to adjust their job preferences to secure employment opportunities continually, which limits the performance of job recommendations. The inherent frequency of preference drift poses a challenge to promptly and precisely capture user preferences. To address this issue, we propose a novel session-based framework, BISTRO, to timely model user preference through fusion learning of semantic and behavioral information. Specifically, BISTRO is composed of three stages: 1) coarse-grained semantic clustering, 2) fine-grained job preference extraction, and 3) personalized top-$k$ job recommendation. Initially, BISTRO segments the user interaction sequence into sessions and leverages session-based semantic clustering to achieve broad identification of person-job matching. Subsequently, we design a hypergraph wavelet learning method to capture the nuanced job preference drift. To mitigate the effect of noise in interactions caused by frequent preference drift, we innovatively propose an adaptive wavelet filtering technique to remove noisy interaction. Finally, a recurrent neural network is utilized to analyze session-based interaction for inferring personalized preferences. Extensive experiments on three real-world offline recruitment datasets demonstrate the significant performances of our framework. Significantly, BISTRO also excels in online experiments, affirming its effectiveness in live recruitment settings. This dual success underscores the robustness and adaptability of BISTRO. The source code is available at <a class="link-external link-https" href="https://github.com/Applied-Machine-Learning-Lab/BISTRO" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

### What problem does this paper attempt to solve? This paper primarily aims to address the issue of job seekers' preference drift in online recruitment platforms. Specifically: 1. **Job Seekers' Preference Drift**: - Job seekers' interests continuously adjust over time and with experience, making it difficult for traditional recommendation systems to accurately capture their latest preferences. - For example, job seekers might adjust their resumes based on the reasons for not getting a job, thereby changing their job search direction. 2. **Noise Interaction Impact**: - During the job search process, users' short-term behaviors might be influenced by noise data such as accidental clicks, further complicating the capture of true preferences. To tackle these issues, the authors propose a new framework called **BISTRO (BehavIoral-Semantic fusion for job RecOmmendation)**. This framework is implemented through the following three stages: 1. **Coarse-Grained Semantic Clustering**: - Grouping users or jobs based on semantics to more broadly identify matches between people and jobs. 2. **Fine-Grained Job Preference Extraction**: - Constructing a multi-granularity interaction hypergraph and employing a novel spectral hypergraph wavelet learning method to capture preference drift and denoise. 3. **Personalized Top-k Job Recommendation**: - Using recurrent neural networks to analyze short-term interaction sequences to infer personalized preferences and generate Top-k job recommendations. Through these technical means, BISTRO aims to improve the accuracy and effectiveness of job recommendations on online recruitment platforms, particularly excelling in handling frequent preference drift and noise data.

Adapting Job Recommendations to User Preference Drift with Behavioral-Semantic Fusion Learning

Feature-Level Attentive Neural Model for Session-Based Recommendation

Exploiting Structural and Temporal Influence for Dynamic Social-Aware Recommendation

Job Recommendation through Progression of Job Selection

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

Learn over Past, Evolve for Future: Search-based Time-aware Recommendation with Sequential Behavior Data

STAN: Stage-Adaptive Network for Multi-Task Recommendation by Learning User Lifecycle-Based Representation

Fusing temporal and semantic dependencies for session-based recommendation

Algorithmic Drift: A Simulation Framework to Study the Effects of Recommender Systems on User Preferences

Behavior Pattern Mining-based Multi-Behavior Recommendation

JobFormer: Skill-Aware Job Recommendation with Semantic-Enhanced Transformer

A bilateral heterogeneous graph model for interpretable job recommendation considering both reciprocity and competition

A Two-Sided Matching and Diversity-Enhanced Method for Job Recommendation with Employer Behavioral Data

Long Short-Term Temporal Meta-learning in Online Recommendation

A Session-based Job Recommendation System Combining Area Knowledge and Interest Graph Neural Networks

Job and Candidate Recommendation with Big Data Support: A Contextual Online Learning Approach

Iterative framework based on multi-task learning for service recommendation

Learning to Retrieve for Job Matching

Leveraging Search History for Improving Person-Job Fit

Parallel Knowledge Enhancement Based Framework for Multi-behavior Recommendation

Learning User Preferences in Non-Stationary Environments