The international workshop IFUP 2016 will be co-located with ACM UMAP 2016: the 24th Conference on User Modeling, Adaptation and Personalization, held in Halifax, Canada on 13-17 July 2016.
> News: The IFUP proceeding is available online Now. Thanks for all your contributions!
Abstract: In this talk, Prof. Parra will talk about different research works where the traditional recommendation problem (predicting unobserved ratings or implicit feedback) is enhanced by fusing additional contexts in domains such as article recommendation, POI recommendation, web marketplace recommendation, and music recommendation. During the talk, Dr. Parra will survey several works on recommendation with contextual data from article suggestion using social bookmarking data to more recent works using weather data for POI recommendation. After the survey, Prof. Parra will summarise lessons based on this previous work and will describe potential challenges and application domains to further develop recommender systems in the aforementioned and other domains.
Recently, auxiliary information (e.g., social friends, item content) has been incorporated in many recommendation models to enhance the performance of both rating prediction and item ranking. However, the used auxiliary data is often referred to as single-dimensional information, such as social trust or item category. Many existing studies focus on how to make the best use of a single facet, such as temporal factors or geo-locations to improve recommendations. However, with the advent of context-aware recommender systems, it gets more and more important to incorporate multiple kinds of auxiliary information the case of which is closer to and more prevalent in practice. The information may be either homogenous or heterogeneous. For example, it may be necessary to consider multiple social relationships (e.g., social trust, friendship, membership, followship) simultaneously to make recommendations rather than merely one of them. Another example is that product recommendation may take into account all kinds of users’ historical data, including purchase, click, browse and wanted list. On one hand, information in different dimensions reflects various views of user modeling and preferences. On the other hand, information from different dimensions is often co-related and dependent in some manner. In this regard, it is necessary to consider all these kinds of information as a whole for user modeling and for further improving recommendation performance. Therefore, how to effectively leverage multi-dimensional information and how these dimensions interacting with each other influence recommendations are the two challenging questions the research community need to resolve.
The international workshop IFUP 2016 aims to provide a dedicated forum for discussing open problems, challenges and innovative research approaches in fusing multi-dimensional information for user modeling and recommender systems. The major goal of this workshop is to promote advanced recommendation solutions that can be easily and readily deployed to meet industrial demands for personalized recommendations.
The scope of the workshop includes (but is not limited to):
All workshop submissions must be formatted according to ACM SIG Proceedings template, and the submissions can be made in either long or short format:
Authors should submit original papers in PDF format through the Easychair system at: https://easychair.org/conferences/?conf=ifup2016
All the accepted manuscripts will be included in the UMAP supplemental proceedings published with CEUR. For the authors of accepted papers, at least one of your paper authors must attend and present your paper at the workshop to ensure the paper appearance in the proceedings of the workshop.
Authors retain copyright of their papers. The editors hold copyright for the proceedings volume. The following copyright statement will be included in the proceedings volume itself:
Copyright © 2016 for the individual papers by the papers' authors. Copying permitted for private and academic purposes. This volume is published and copyrighted by its editors.