TrNews: Heterogeneous User-Interest Transfer Learning for News Recommendation

Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics, pages 734–744 April 19 - 23, 2021. ©2021 Association for Computational Linguistics 734 TrNews: Heterogeneous User-Interest Transfer Learning for News Recommendation Guangneng Hu HKUST / Hong Kong, China njuhgn@gmail.com Qiang Yang HKUST / Hong Kong, China qyang@cse.ust.hk Abstract We investigate how to solve the cross-corpus news recommendation for unseen users in the future. This is a problem where traditional content-based recommendation techniques of- ten fail. Luckily, in real-world recommen- dation services, some publisher (e.g., Daily news) may have accumulated a large corpus with lots of consumers which can be used for a newly deployed publisher (e.g., Politi- cal news). To take advantage of the existing corpus, we propose a transfer learning mod- el (dubbed as TrNews) for news recommenda- tion to transfer the knowledge from a source corpus to a target corpus. To tackle the hetero- geneity of different user interests and of differ- ent word distributions across corpora, we de- sign a translator-based transfer-learning strate- gy to learn a representation mapping between source and target corpora. The learned trans- lator can be used to generate representation- s for unseen users in the future. We show through experiments on real-world datasets that TrNews is better than various baselines in terms of four metrics. We also show that our translator is effective among existing transfer strategies. 1 Introduction News recommendation is key to satisfying users’ information need for online services. Some news articles, such as breaking news, are manually se- lected by publishers and displayed for all users. A huge number of news articles generated everyday make it impossible for editors and users to read through all of them, raising the issue of informa- tion overload. Online news platforms provide a service of personalized news recommendation by learning from the past reading history of users, e.g., Google (Das et al., 2007; Liu et al., 2010), Ya- hoo (Trevisiol et al., 2014; Okura et al., 2017), and Bing news (Lu et al., 2015; Wang et al., 2018). When a new user uses the system (cold-start users) or a new article is just created (cold-start items), there are too few observations for them to train a reliable recommender system. Content- based techniques exploit the content information of news (e.g., words and tags) and hence new articles can be recommended to existing users (Pazzani and Billsus, 2007). Content-based recommendation, however, suffers from the issue of data sparsity since there is no reading history for them to be used to build a profile (Park and Chu, 2009). Transfer learning is a common technique for alle- viating the issues of data sparsity (Pan et al., 2010; Cantador et al., 2015; Liu et al., 2018). A user may have access to many websites such as Twitter.com and Youtube.com (Roy et al., 2012; Huang and Lin, 2016), and consume different categories of prod- ucts such as movies and books (Li et al., 2009). In this case, transfer learning approaches can recom- mend articles …

Content truncated to 500 words