Two-Level Automatic Adaptation of a Distributed User Profile for Personalized News Content Delivery

Abstract

This paper presents a distributed client-server architecture for the personalized delivery of textual news content to mobile users. The user profile consists of two separate models, that is, the long-term interests are stored in a skeleton profile on the server and the short-term interests in a detailed profile in the handset. The user profile enables a high-level filtering of available news content on the server, followed by matching of detailed user preferences in the handset. The highest rated items are recommended to the user, by employing an efficient ranking process. The paper focuses on a two-level learning process, which is employed on the client side in order to automatically update both user profile models. It involves the use of machine learning algorithms applied to the implicit and explicit user feedback. The system's learning performance has been systematically evaluated based on data collected from regular system users.

Personalization algorithms for portable personality

“…Ref Music [36], [7] Social networks [7] Movie/TV program [26], [ While combining different user information, the UP can be divided into static and dynamic parts ( [22], [39], [25]). On the one hand, the static part includes all fixed user information, which does not change regularly, such as name and address.…”

“…The UP in [22] applies such a division. Moreover, the dynamic part can further be split into short-term and long-term interests as done in [25], where a long-term profile has general information and a short-term profile holds more detailed information.…”

“…Implicit feedback is extracted from browsing manners such as how much time is spent for certain news types or single news as well as the average read time spent on the news which has positive feedback. [25] works on the same topic of personalized news delivery, where explicit feedback is used for user profile initialization and implicit feedback to monitor user interactions with content items such as the number of key clicks in a document, the amount of scrolling through the document, or the amount of time spent reading the document.…”

“…The most common one nowadays is the Internet especially the World Wide Web by using server-side ( [33], [25]) and client-side methods [30]. Server-side acquisitions use i.e.…”

“…The approach of personalized news content in [25] focuses on a two-level learning process, which is employed on the mobile device side in order to automatically update the general and specific profile models. It involves the use of machine learning algorithms applied to the implicit and explicit user feedback.…”