Using Recommender Systems to Help Revitalize Local News Payam Pourashraf DePaul University Chicago, IL, USA ppourash@depaul.edu Bamshad Mobasher DePaul University Chicago, IL, USA mobasher@cs.depaul.edu ABSTRACT American local newspapers have been experiencing a large loss of reader retention and business within the past 15 years due to the proliferation of online news sources and social media. This has led to a disturbing trend where local journalism and local news outlets are being forced out of business often leaving whole communities without a key source of credible information. This trend has a poten- tially broad societal impact as these key anchors of local trust and democracy are slowly becoming extinct. Local media companies are starting to shift from an advertising-supported business model to one based on subscriptions to mitigate their fnancial crises. But with strong competition from a variety of online news sources, these companies need to increase user engagement by providing signifcant added value. Providing more personalized content in the local context may be one way that these companies can succeed in this efort. Recommender system technologies are the primary enabling mechanisms for delivering such personalized content. However, using standard machine learning models that focus on users’ global preferences is not appropriate in this context because the local preferences of users exhibit some specifc characteristics which do not necessarily match their long-term or global prefer- ences in the news. The overall goal of this research is to develop predictive models that more efectively derive user engagement through automatic personalization. Efective recommender systems may be among the tools that can help reverse the current decline of interest in local news sources. Our research explores approaches to learning localized models from user interaction data with news articles, particularly in news categories where there is intense local interest and there is a signifcant diference between users’ global and local news preferences. Specifcally, we propose using such localized models in a session-based recommender system where the system can switch between users’ global and local preference models automatically when warranted. We report experiments per- formed on a news dataset from a local newspaper show that these local models, particularly the Life-and-Culture news category, do indeed provide more accuracy and efectiveness for personalization which, in turn, may lead to more user engagement with local news content. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for proft or commercial advantage and that copies bear this notice and the full citation on the frst page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). UMAP ’22 Adjunct, July 04ś07, 2022, Barcelona, Spain © 2022 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-9232-7/22/07. https://doi.org/10.1145/3511047.3536411 CCS CONCEPTS · Information Systems → Personalization; Recommender Sys- tems. KEYWORDS News recommender system, Session-based recommendation, Local news ACM Reference Format: Payam Pourashraf and Bamshad Mobasher. 2022. Using Recommender Systems to Help Revitalize Local News. In Adjunct Proceedings of the 30th ACM Conference on User Modeling, Adaptation and Personalization (UMAP ’22 Adjunct), July 04ś07, 2022, Barcelona, Spain. ACM, New York, NY, USA, 5 pages. https://doi.org/10.1145/3511047.3536411 1 INTRODUCTION Intelligent personalized applications such as recommender sys- tems have become essential online tools in many domains such as e-commerce, music and video streaming, online news, and so- cial media marketing. These systems help alleviate information overload and assist users in decision making by tailoring their recommendations users’ preferences. One of the key emerging ap- plication domains is the recommendation of online news [16, 17]. We believe that news recommender systems, with their ability to personalize and tailor news delivery to users’ personal interests and preferences, provide an enabling technology to help local news outlets to better engage users and provide added value that would help these outlets remain relevant and develop customer loyalty. Efective local news recommender systems can be one tool to help revitalize local journalism and to reverse the decline of local news outlets. However, current machine learning and predictive modeling approaches used in personalized news recommendation tend to prioritize news items based on topics in global user preferences across all news categories and geographical locations, as well as based on popularity and recency of news items. This focus on long- term high-level preferences may, in fact, result in taking attention away from local stories that help provide the distinguishing added value for local news outlets. For example, a local newsreader may not generally be a sports fan, and hence sports-related features may not play a major role in generating personalized content for that user. However, the user, like many other local readers, might have an intense interest in the local high school team. Similarly, local readers may be interested in reading about local crime, even though for most users, this may not be part of their global news preferences. In such situations, global user preferences cannot be used to generate recommendations to users interested in local news content. At the same time, purely localized models will not be able to represent the users’ broader 80