Embedding-based News Recommendation for Millions of Users

Okura, Shumpei; Tagami, Yukihiro; Ono, Shunsuke; Tajima, Akira

doi:10.1145/3097983.3098108

Cited by 420 publications

(317 citation statements)

References 8 publications

Supporting

Mentioning

317

Contrasting

Order By: Relevance

“…Neural Recommender Systems. Neural networks have been extensively applied in recommender systems thanks to their highquality recommendations [9,10,13,16,23,30]. Particularly, deep learning is able to capture non-linear and non-trivial relationships between users and items, which provides in-depth understanding of user demands and item characteristics, as well as the interactions between them.…”

Section: Related Workmentioning

confidence: 99%

Interact and Decide

Tran

Pham

Tay

et al. 2019

Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

This paper proposes Medley of Sub-Attention Networks (MoSAN), a new novel neural architecture for the group recommendation task. Group-level recommendation is known to be a challenging task, in which intricate group dynamics have to be considered. As such, this is to be contrasted with the standard recommendation problem where recommendations are personalized with respect to a single user. Our proposed approach hinges upon the key intuition that the decision making process (in groups) is generally dynamic, i.e., a user's decision is highly dependent on the other group members. All in all, our key motivation manifests in a form of an attentive neural model that captures fine-grained interactions between group members. In our MoSAN model, each sub-attention module is representative of a single member, which models a user's preference with respect to all other group members. Subsequently, a Medley of Sub-Attention modules is then used to collectively make the group's final decision. Overall, our proposed model is both expressive and effective. Via a series of extensive experiments, we show that MoSAN not only achieves state-of-the-art performance but also improves standard baselines by a considerable margin. CCS CONCEPTS• Information systems → Recommender systems; • Computing methodologies → Neural networks.

show abstract

Section: Related Workmentioning

confidence: 99%

Interact and Decide

Tran

Pham

Tay

et al. 2019

Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

show abstract

“…However, pure CF methods usually suffer from the sparsity and the cold-start problems, which are especially significant in news recommendation scenarios [19]. Thus, content-based techniques are usually complementary methods to CF [2,20,21,24,27,29,32,39]. For example, Liu et al [21] proposed to incorporate user interests for news recommendation.…”

Section: Related Work 21 News Recommendationmentioning

confidence: 99%

“…They use a Bayesian model to predict the interests of users based on the distributions of their clicked news articles in different categories. Okura et al [24] proposed to learn news embeddings based on the similarities between news articles in the same and different categories. They use recurrent neural networks to learn user representations from the browsing histories through time to predict the score of news.…”

Section: Related Work 21 News Recommendationmentioning

confidence: 99%

“…Existing news recommendation methods are usually based on collaborative filtering (CF) techniques and news content [20][21][22]24]. For example, Liu et al [21] proposed a CF-based approach for news recommendation based on user interests.…”

Section: Introductionmentioning

confidence: 99%

“…They use a Bayesian model to extract the interest features of users based on the click distributions on news articles in different categories. Okura et al [24] proposed to first learn the distributed representations of news articles based on similarity and then use recurrent neural networks to learn user representations from browsing histories for click prediction. Lian et al [20] proposed a deep fusion model (DMF) to learn representations of users and news using combinations of fully connected networks with different depth.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Npa

et al. 2019

Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

220

View full text Add to dashboard Cite

News recommendation is very important to help users find interested news and alleviate information overload. Different users usually have different interests and the same user may have various interests. Thus, different users may click the same news article with attention on different aspects. In this paper, we propose a neural news recommendation model with personalized attention (NPA). The core of our approach is a news representation model and a user representation model. In the news representation model we use a CNN network to learn hidden representations of news articles based on their titles. In the user representation model we learn the representations of users based on the representations of their clicked news articles. Since different words and different news articles may have different informativeness for representing news and users, we propose to apply both word-and news-level attention mechanism to help our model attend to important words and news articles. In addition, the same news article and the same word may have different informativeness for different users. Thus, we propose a personalized attention network which exploits the embedding of user ID to generate the query vector for the wordand news-level attentions. Extensive experiments are conducted on a real-world news recommendation dataset collected from MSN news, and the results validate the effectiveness of our approach on news recommendation.

show abstract