VLSNR:Vision-Linguistics Coordination Time Sequence-aware News Recommendation
This addresses the problem of improving news recommendation accuracy for users by integrating multimodal semantics, though it appears incremental as it builds on existing methods with added visual components.
The paper tackles news recommendation by incorporating both visual and textual information to model users' dynamic interests, achieving state-of-the-art performance on a newly constructed multimodal dataset.
News representation and user-oriented modeling are both essential for news recommendation. Most existing methods are based on textual information but ignore the visual information and users' dynamic interests. However, compared to textual only content, multimodal semantics is beneficial for enhancing the comprehension of users' temporal and long-lasting interests. In our work, we propose a vision-linguistics coordinate time sequence news recommendation. Firstly, a pretrained multimodal encoder is applied to embed images and texts into the same feature space. Then the self-attention network is used to learn the chronological sequence. Additionally, an attentional GRU network is proposed to model user preference in terms of time adequately. Finally, the click history and user representation are embedded to calculate the ranking scores for candidate news. Furthermore, we also construct a large scale multimodal news recommendation dataset V-MIND. Experimental results show that our model outperforms baselines and achieves SOTA on our independently constructed dataset.