Group Modeling: Selecting a Sequence of Television Items to Suit a Group of Viewers

Abstract

Abstract. Watching television tends to be a social activity. So, adaptive television needs to adapt to groups of users rather than to individual users. In this paper, we discuss di¡erent strategies for combining individual user models to adapt to groups, some of which are inspired by Social Choice Theory. In a ¢rst experiment, we explore how humans select a sequence of items for a group to watch, based on data about the individuals' preferences. The results show that humans use some of the strategies such as the Average Strategy (a.k.a. Additive Utilitarian), the Average Without Misery Strategy and the Least Misery Strategy, and care about fairness and avoiding individual misery. In a second experiment, we investigate how satis¢ed people believe they would be with sequences chosen by di¡erent strategies, and how their satisfaction corresponds with that predicted by a number of satisfaction functions. The results show that subjects use normalization, deduct misery, and use the ratings in a non-linear way. One of the satisfaction functions produced reasonable, though not completely correct predictions. According to our subjects, the sequences produced by ¢ve strategies give satisfaction to all individuals in the group. The results also show that subjects put more emphasis than expected on showing the best rated item to each individual (at a cost of misery for another individual), and that the ratings of the ¢rst and last items in the sequence are especially important. In a ¢nal experiment, we explore the in£uence viewing an item can have on the ratings of other items. This is important for deciding the order in which to present items. The results show an e¡ect of both mood and topical relatedness.

Personalization in cultural heritage: the road travelled and the one ahead

“…For instance, see (Jameson and Smyth, 2007;Vassileva and Sun, 2007;Masthoff, 2004;Masthoff and Gatt, 2006;Ardissono et al, 2010). While group modeling has substantially different meanings in the management of synchronous activities, such as the museum visit, or asynchronous ones, such as the participation in a virtual community, an intersection between these two research fields can be found as far as modeling group interests and knowledge is concerned.…”

Probabilistic Score Normalization for Rank Aggregation

“…To name a few examples, rank aggregation takes place in the combination of multiple criteria for document/query similarity assessment in most search engines; in merging the outputs of different engines for meta-search; in the combination of query-based and preference-based relevance for personalized search [1]; or even in the combination of preferences from multiple users for collaborative retrieval [5]. Both rank-based and score-based aggregation techniques have been explored in prior research on this topic [7].…”

Development of a group recommender application in a Social Network

“…The general opinion is that there is a need to adapt the recommendation process to group composition [2,22]. This is backed up by some recent works that have focused their studies on analyzing the effectiveness of group recommendations according to different aspects, such as group size and inner group similarity [3].…”

“…each group member is taken). With the data retrieved in our experiments in simulated environments [10,11] we have performed a conscientious experimentation comparing the recommendation results of the state-of-theart aggregation functions presented by [22] when applying them to what we will next define as standard and social recommendations approaches. The results of these experiments are out of the scope of this paper but can be found in [64].…”

“…For example [22] in the recommendation of TV programs domain or FlyTrap [15] where previous selections are also taken into consideration. Besides, some recommendations tend to be repeatedly detrimental to the same group members.…”