This workshop builds upon the success of the two previous editions held in conjunction with the 3rd and 4th ACM Conferences on Recommender Systems in 2009 and 2010.
The importance of contextual information has been recognized by researchers and practitioners in many disciplines, including e-commerce personalization, information retrieval, ubiquitous and mobile computing, data mining, marketing, and management. While a substantial amount of research has already been performed in the area of recommender systems, the vast majority of existing approaches focuses on recommending the most relevant items to users and does not take into account any additional contextual information, such as time, location, weather, or the company of other people.
In the last year a growing number of researchers have tried to address this limitation, but still the vast majority of the current solutions reduce the context model to a fixed set of contextual features that are used either to pre-filter the data used for building the prediction model or to post-filter the recommendations using the contextual conditions of the user. Therefore, this workshop aims to bring together researchers with wide- ranging backgrounds to identify important research questions, to exchange ideas from different research disciplines, and, more generally, to facilitate discussion and innovation in the area of context-aware recommender systems (CARS).
In the last decade digital music has transformed the landscape of music experience and distribution. It is not uncommon to find personal music collections that exceed thousands of tracks, and thanks to the Internet finding and accessing music has become simpler than ever. As a result, music recommendation systems have become an increasingly important way for music listeners to discover and navigate music.
Music is inherently different than other types of media. The space of recommended items is extremely large -compared to other domains- with a typical online music store offering 10 million titles to chose from. People interact with music differently than they do with other types of media. A new song can be auditioned in a matter of minutes whereas a movie may take a couple of hours to watch, and a book may take a dozen hours to read. People enjoy listening to music over and over, but it is the rare book that is read more than once. Listeners vary their music preference based upon context and activities. A playlist for jogging is likely to be very different than a playlist created by the same user for relaxing. Listeners enjoy listening to sequences of songs often getting as much enjoyment from the song transitions as from the songs themselves. The uniqueness of music as recommendation domain present challenges not seen in other recommender domains. It is important to consider the special nature of music when building recommenders for music.
We see this workshop as a platform where the RecSys, Music Information Retrieval, User Modeling, Music Cognition, and Music Psychology communities could meet, exchange ideas and collaborate.
The exponential growth of the social web continues to pose challenges and opportunities for recommender systems. The social web has turned information consumers into active contributors creating massive amounts of information. Finding relevant and interesting content at the right time and in the right context is challenging for existing recommender approaches. At the same time, social systems by their definition encourage interaction between users and both online content and other users, thus generating new sources of knowledge for recommender systems. Web 2.0 users explicitly provide personal information and implicitly express preferences through their interactions with others and the system (e.g. commenting, friending, rating, etc.). These various new sources of knowledge can be leveraged to improve recommendation techniques and develop new strategies which focus on social recommendation. The Social Web provides huge opportunities for recommender technology and in turn recommender technologies can play a part in fuelling the success of the Social Web phenomenon. The goal of this workshop, the third in the series, is to bring together researcher and practitioners to explore, discuss, and understand challenges and new opportunities for recommender systems and the Social Web.
Following the success of CAMRa2010, we are pleased to announce this year’s Challenge on Context-Aware Movie Recommendation.
The majority of existing recommendation approaches does not take into account contextual information, such as time, location, or weather. This challenge aims to tackle the practical issue of context-aware movie recommendation. A new movie rating dataset from Moviepilot will be released for the challenge. The dataset contains a number of contextual features, typically not found in standard collaborative filtering datasets. The participating teams are requested to use the additional contextual features to generate context-aware recommendations.
The challenge focuses on classification accuracy metrics. The participants are invited to submit papers focusing on the challenge to the workshop, which will be held in conjunction with RecSys-2011.
Increasing attention is given to finding ways for combining, integrating and mediating heterogeneous sources of information for the purpose of providing better personalized services in many information seeking and e-commerce applications. Information heterogeneity can indeed be identified in any of the pillars of a recommender system: the modeling of user preferences, the description of resource contents, the modeling and exploitation of the context in which recommendations are made, and the characteristics of the suggested resource lists.
Thus, in recommender systems, among other issues: a) user models could be based on different types of explicit and implicit personal preferences, such as ratings, tags, textual reviews, records of views, queries, and purchases; b) recommended resources may belong to several domains and media, and may be described with multilingual metadata; c) context could be modeled and exploited in multi-dimensional feature spaces; d) and ranked recommendation lists could be diverse according to particular user preferences and resource attributes, oriented to groups of users, and driven by multiple user evaluation criteria.
In the HetRec workshop, we would like to raise awareness of the potential of using multiple sources of information, and look for sharing expertise and suitable models and techniques. Another dire need is for strong datasets, and one of our aims is to establish benchmarks and standard datasets on which the problems could be investigated. In this edition, we make available on-line datasets with heterogeneous information from several social systems. These datasets can be used by participants to experiment and evaluate their recommendation approaches, and be enriched with additional data, which may be published at the workshop website for future use.
Interacting with a recommender system means to take different decisions such as selecting a song/movie from a recommendation list, selecting specific feature values (e.g., camera’s size, zoom) as criteria, selecting feedback features to be critiqued in a critiquing based recommendation session, or selecting a repair proposal for inconsistent user preferences when interacting with a knowledge-based recommender. In all these scenarios, users have to solve a decision task.
The complexity of decision tasks, limited cognitive resources of users, and the tendency to keep the overall decision effort as low as possible leads to the phenomenon of bounded rationality, i.e., users are exploiting decision heuristics rather than trying to take an optimal decision. Furthermore, preferences of users will likely change throughout a recommendation sessions, i.e., preferences are constructed in a specific decision environment and users do not know their preferences beforehand.
Decision making under bounded rationality is a door opener for different types of non‐conscious influences on the decision behavior of a user. Theories from decision psychology and cognitive psychology are trying to explain these influences, for example, decoy effects and defaults can trigger significant shifts in item selection probabilities; in group decision scenarios, the visibility of the preferences of other group members can have a significant impact on the final group decision.
The major goal of this workshop is to establish a platform for industry and academia to present and discuss new ideas and research results that are related to the topic of human decision making in recommender systems.
The rise of location-enabled mobile phones and location based services offers a great opportunity to apply personalization and recommender system technology to people's everyday lives. A variety of digital traces can now be used to infer how people move about their city and extract their context and habits. Personalization and recommender systems, potentially merged with the data that people store online (e.g., social networks, web ratings), can then not only be used to recommend new places and events that they may find interesting to attend, but, more broadly, personalize and enhance any service that people find themselves using.
The research questions that emerge from this setting currently span a variety of fields: pervasive and persuasive technology, smart city/ubiquitous systems, personalization and recommender systems, mobility, social networking, and human-computer interaction. This workshop aims to bring together a cross section of researchers and practitioners to discuss practical challenges, solutions, and opportunities that come from this setting.
October 23 (morning)
Most research and development efforts in the Recommender Systems field have been focused on accuracy in predicting and matching user interests. However there is a growing realization that there is more than accuracy to the practical effectiveness and added-value of recommendation. In particular, novelty and diversity have been identified as key dimensions of recommendation utility in real scenarios, and a fundamental research direction to keep making progress in the field.
Novelty is indeed essential to recommendation: in many, if not most scenarios, the whole point of recommendation is inherently linked to a notion of discovery, as recommendation makes most sense when it exposes the user to a relevant experience that she would not have found, or thought of by herself. Not only does a varied recommendation provide in itself for a richer user experience. Given the inherent uncertainty in user interest prediction, avoiding a too narrow array of choice is generally a good approach to enhance the chances that the user is pleased by at least some recommended item. Sales diversity may enhance businesses as well, leveraging revenues from market niches. The challenge is to enhance these aspects while still delivering quality and achieving a fair match of the user's interests.
DiveRS 2011 aims to gather researchers and practitioners interested in the role of novelty and diversity in recommender systems, in a forum where participants can discuss problems, exchange ideas, and find opportunities for collaboration. The workshop seeks to advance towards a better understanding of what novelty and diversity are, how they can improve the effectiveness of recommendation methods and the utility of their outputs. We aim to identify open problems, relevant research directions, and opportunities for innovation in the recommendation business in this area.
October 23 (afternoon)
Research on "human-recommender interaction" is scarce. Algorithm optimization and off-line testing using measures like RMSE are dominant topics in the RecSys community, but theorizing about consumer decision processes and measuring user satisfaction in online tests is less common. Researchers in marketing and decision-making have been investigating consumer choice processes in great detail, but only sparingly put this knowledge to use in technological applications. Human-computer interaction has been focusing on the usability of interfaces for ages, but does not seem to link research on consumer choice and recommender system interfaces.
During RecSys 2010, we organized the first UCERSTI workshop to bridge these gaps. Two keynote speeches, 7 accepted papers and a lively panel discussion introduced the visitors of RecSys 2010 to the field of human-recommender interaction. By means of UCERSTI 2 we hope to further strengthen the bonds between these researchers, to exchange new experiences, and meet and interest other new researchers working on user-centric research in recommender systems. The proposed format will be a half-day workshop with paper and poster presentations, an invited talk and a plenary discussion.