MONERS: A news recommender for the mobile web
Introduction
The mobile web allows users navigate the Web using wireless devices such as cell phones and personal digital assistants (PDAs). During early days of the mobile web, ring tones and wallpaper were the most commonly accessed content, but in recent years there has been a significant increase in the use of data content such as music, movie information, and news (Kelly, Gray, & Minges, 2003; Kim, Lee, Cho, & Kim, 2004). Because news services on the mobile web are so ubiquitous, there are many users.
Currently, most mobile web news services are sorted by category or by news article attributes, such as recency. An inconvenient user interface may constitute a barrier to browsing between contents; therefore, mobile web news services should provide easy access to the categories or content preferred by users. Researchers have conducted many studies on personalizing news services in order to make recommendations, using the existing web (Billsus and Pazzani, 2000, Goldberg et al., 1992, Konstan et al., 1997, Shepherd et al., 2002). News can be filtered using collaborative filtering, which is based on similarities among user preferences, or by using information filtering, which uses news article keywords and user profiles (Aas, 1997, Belkin and Croft, 1992, Foltz and Dumais, 1992, Konstan et al., 1997, Mostafa et al., 1997).
User interface, user behavior, and browsing between content are different on the mobile web than they are in a web-based service (Brunato and Battiti, 2003, Ho and Kwok, 2003, Kim et al., 2004); therefore, the mobile web requires its own recommendation method, adapted to its features.
This paper presents a mobile web news recommendation system, Mobile News Recommendation Systems (MONERS), which incorporates news article attributes, user preferences with regard to category and news articles, and user segments. MONERS incorporates the news article’s importance, as well as its recency, calculated by the difference between the time it was posted and the present. It also incorporates a user segment that is focused on user profiles, reading patterns of news articles, changes in user interest, and usage patterns. Users of the actual mobile web tested the news recommender’s performance, and its performance was measured and analyzed.
Section 2 summarizes studies related to news article recommendation and the mobile web. Section 3 introduces MONERS’ flow and recommendation algorithms. Section 4 presents the experiment used to analyze the news recommendation method performance, and an analysis of the results. Section 5 discusses this study’s implications and conclusions.
Section snippets
News recommendation
Personalization and recommendation involving news services on the web use the following approaches:
- •
Collaborative filtering: This is a method for calculating expected user preference for a product, using evaluation by, or the preferences of, other users who have experienced the product (Billsus and Pazzani, 1998, Goldberg et al., 1992, Konstan et al., 1997). Collaborative filtering is currently widely applied and is used not only for news but also for various products such as music or movies (
Considerations for mobile web news recommendation
Three points must be considered when designing news content personalization for the mobile web. The first is news content. On the mobile web, news services focus on the distribution of current news rather than on past or related news articles; searching past or related articles and browsing between pages is not easy, so mobile web news services mainly provide current reports or important articles sorted by news category. It is therefore important, when personalizing news, to consider the
Experiments
An experiment on the use of this news recommendation service was carried out on service members who were registered with a Korean mobile service provider’s intelligent wireless service. The news recommendation service’s usage log from October 2003 to the end of April 2004 was analyzed.
When a news service is accessed, screens similar to ones in Fig. 4 appear. The root menu of a news service consists of recommended news, news by category, and current news. Recommended news provides personally
Performance of MONERS recommendation
As the comparison between performances in Section 4.2 indicates, news by category had the most news hits, and current and recommended news had a similar number of hits. News by category enables users to read various articles in one category; they can access news categories interesting to them such as MLB news or entertainment. Users often selected several news articles in one category; however, if a news article consisted of several pages, they tended to read more pages of recommended articles
Conclusion and limitations
This paper presented the MONERS news recommendation system, a system that incorporates the characteristics of users as well as the nature of the mobile web news service and content. MONERS calculates the distance between the present time and the time an article was posted to determine news recency; it also incorporates the importance of news articles. MONERS also considers changes in user preference with regard to news categories. It can identify preference for news category by user segment and
References (24)
- Aas, K. (1997). A survey on personalized information filtering systems for the World Wide Web. Norwegian Computing...
- et al.
Information filtering and information retrieval: two sides of same coin
Communications of the ACM
(1992) - Billsus, D., & Pazzani, M. J. (1998). Learning collaborative information filters. In Proceedings of 15th international...
- et al.
User modeling for adaptive news access
User Modeling and User-Adapted Interaction
(2000) - Billsus, D., Pazzani, M. J., & Chen, J. (2000). A learning agent for wireless news access. In Proceedings of IUI 2000...
- Breese, J. S., Heckerman, D., & Kadie, C. (1998). Empirical analysis of predictive algorithms for collaborative...
- Brunato, M., & Battiti, R. (2003). PILGRIM: A location broker and mobility-aware recommendation system. In Proceedings...
- et al.
Combining content-based and collaborative filters in an online newspaper
- et al.
Personalized information delivery: an analysis of information filtering methods
Communications of the ACM
(1992) - et al.
Using Collaborative filtering to weave an information tapestry
Communication of the ACM
(1992)
The attraction of personalized service for users in mobile commerce: an empirical study
ACM SIGecom Exchanges
A personal news service based on a user model neural network
IEICE Transactions on Information and Systems
Cited by (108)
A hybrid similarity model for mitigating the cold-start problem of collaborative filtering in sparse data
2024, Expert Systems with ApplicationsExploring the effects of different Clustering Methods on a News Recommender System
2021, Expert Systems with ApplicationsCD-CARS: Cross-domain context-aware recommender systems
2019, Expert Systems with ApplicationsCitation Excerpt :CDRS commonly lead to a higher user satisfaction by addressing cold-start, sparsity, and diversity problems, although they may not be necessarily more accurate than traditional CF recommender systems (Cantador et al., 2015). In parallel, Context-Aware Recommender Systems (CARS) takes into account the user’s context (e.g., location, time, mood, etc.) (Adomavicius & Tuzhilin, 2015; Lee & Park, 2007; Unger, Bar, Shapira, & Rokach, 2016; Viktoratos, Tsadiras, & Bassiliades, 2018; Villegas, Sánchez, Díaz-Cely, & Tamura, 2018; Wang, Li, Zhao, & Chen, 2017; Zarka, Cordier, Egyed-Zsigmond, Lamontagne, & Mille, 2016). In many applications, such as recommending a vacation package or a TV program, it may not be sufficient to consider only users and items ratings.
News reading on smartphones: How do mood, modality interactivity, and news story length influence news engagement?
2024, Behaviour and Information Technology