Personalized News Recommendation: Methods and Challenges

Abstract

Personalized news recommendation is an important technique to help users find their interested news information and alleviate their information overload. It has been extensively studied over decades and has achieved notable success in improving users' news reading experience. However, there are still many unsolved problems and challenges that need to be further studied. To help researchers master the advances in personalized news recommendation over the past years, in this paper we present a comprehensive overview of personalized news recommendation. Instead of following the conventional taxonomy of news recommendation methods, in this paper we propose a novel perspective to understand personalized news recommendation based on its core problems and the associated techniques and challenges. We first review the techniques for tackling each core problem in a personalized news recommender system and the challenges they face. Next, we introduce the public datasets and evaluation metrics used for personalized news recommendation. We then discuss the key points on improving the responsibility of personalized news recommender systems. Finally, we raise several research directions that are worth investigating in future. This paper can provide up-to-date and comprehensive views to help readers understand the personalized news recommendation field. We hope this paper can facilitate research on personalized news recommendation and as well as related fields in natural language processing and data mining.

Using user's local context to support local news

“…There are two main evaluation methods: objective measures, such as accuracy and diversity, and subjective measures, such as reader satisfaction. This study focused on objective measures since we were using offline experiments and did not have access to the • HR@K: Among the top K items (articles in our experiments) HR finds the rate of times in which relevant items are retrieved [16,26]. • NDCG@K: is a standard ranking metric that uses the graded relevance to rank each item (which commonly, an article is viewed as either relevant or not relevant by the other ranking metrics) [16,26].…”

Quality-aware News Recommendation

“…News recommendation techniques have been extensively studied over years (Wu et al, 2021). Most existing news recommendation methods rely on click signals to model user interest and train recommendation models (Okura et al, 2017;Wu et al, 2019c;Wang et al, 2020;Liu et al, 2020;Hu et al, 2020a,b;Qi et al, 2021a;Zhang et al, 2021b,a).…”

“…News recommendation is important for online news platforms to provide users with personalized news reading services to alleviate their information overload (Wu et al, 2020c). Predicting whether a user will click on a candidate news is a core task in news recommendation (Wu et al, 2021). For example, Okura et al (2017) proposed to use a GRU network to model user interest from clicked news, and match it with candidate news for click prediction.…”

FUM: Fine-grained and Fast User Modeling for News Recommendation

“…A critical step of news recommendation is to accurately model the interest of a target user [12,23,33]. Existing methods usually first independently encode user's clicked news into news embeddings and then aggregate them to build user embedding [13,14,16,26,31,34]. For example, Wu et al [25] first employ the self-attention mechanism to learn news embeddings for user's clicked news from their titles, independently.…”