1. Exploiting Group Recommendation Functions for Flexible Preferences

2. Introduction  This paper target an issue about group recommendation system  Group recommendation system is a system that recommend set of items to the users that belong in a group (to make it clear, each person in the group receive the same recommendations)  This group recommendation is based on the individual preferences that were set by each user in the group  In the previous work, it stated that reaching consensus or agreement between group members is an important step in group recommendations  Also in the previous work, group recommendation system usually based on the user’s past preference

3. Background  In this paper, they want to focused in a specific scenario where users are provided with a way to update their preferences during recommendation time  This flexibility give the users to choose items they would like or not to see during recommendation time and the system take into account the newly provided preferences to update the recommendations  Using this scenario, they believe that this new feature is useful in a number of practical applications where users are likely to be in different mindset and do not want the system to solely based on their past reference  In this case they are using travelling scenario to cope with the problem  They realize that in such condition/scenario, users will much likely to update their preferences consciously in an effort to maximize their individual satisfaction

4. Proposal  They introduce a flexible feedback model in the form of a vector of preferences that any group member could provide at recommendation time  This model contain a feedback box that outputs a feedback vector for each group member such that her satisfaction is maximized in the generated recommendation  The feedback box will replace the preference vectors provided by the users by a potentially different feedback vector that is better suited for helping users enforce their new preferences  In addition, they also realize it can be disconcerting for existing users to experience drastic changes to recommendation due to users update  They introduce recommendation robustness to ensure that generated recommendations after preference updates overlap with the ones generated before

5. Scenario  In order to evaluate the usefulness of the proposed system, they evaluate using two kinds of measurement and group recommendation consensus function  The group recommendation consensus function is a function which is used to identify group’s agreement from several individual data  They use Aggregated Voting and Least Misery group function  To measure user satisfaction they use a simple form of Jaccard Index, Overlap Similarity to calculate the similarity between the user preference/feedback vector and the recommendation  They also use Hamming Distance measurement to calculate the dissimilarity (distance) between them.

6. Interaction Models  To elicit preference, user input Boolean vector data where value 1 corresponds to the items user prefers to consume and value 0 otherwise  After that feedback generation process take place to generate output through the feedback box  During this phase, the system through the feedback box computes for the user a suggested Boolean feedback vector

7. Data model  User Preference Vector : A preference vector under Boolean model defines a subset of items that a user would like to consume  User Feedback Vector : This vector is the generated feedback for each user that the recommendation function uses to recommend items to the group  Robustness : The robustness quantify the similarity between the current recommendation (after user update) and the previous recommendation using similarity measures  User Satisfaction : The inverse of distance, or proportional to the similarity between user preference and recommended item vector

8. Data Model  Aggregated Voting : The Aggregated Voting Consensus generates a set of items recommendation such that an overall aggregated user satisfaction is maximized  Least Misery : The Least Misery Consensus generates a set of item recommendation such that the minimum user satisfaction is maximized

9. Problem Definitions

11. Aggregating Voting Consensus (Overlap Similarity)

12.  In the Generating New Feedback Vector task, they realized that feedback box is not useful  They find that if the aggregated consensus function is used, and satisfaction measure is Overlap Similarity, the feedback(u) will always be subset of pref(u)  This make the feedback(u) will be at most as useful as pref(u)

13. Aggregating Voting Consensus (Hamming Distance)

15. Least Misery Consensus (Overlap Similarity)

17. Least Misery Consensus (Hamming Distance)  In Generating Recommendations Task, they are interested to compute recommendation such that the maximum Hamming Distance between the generated recommendation and individual feedback vector/pref vector is minimized  They use algorithm R-LMD  This algorithm designed as an optimization problem where the quadratic equation is solved by a general purpose solver to obtain recommendation vector where each value is fractional value between 0 and 1  After the optimization process, they perform deterministic rounding

18. Least Misery Consensus (Hamming Distance)

19. Recommendation Robustness  Robustness is considered as a soft constraint that could further be tuned  The primary idea is to add the previously generated recommendation vector as a new feedback vector (as a pseudo-user) to the recommendation function  They believe that this should achieve higher degree of robustness

20. Experiments  Using real world data sets from  Lonely Planet  Flickr  MovieLens  Implemented in C++ using IBM CPLEX for formulations  All experiments conducted on AMD machine quad-core 2.0 GHz CPUS, 8GB Memory, and 1TB HDD, running Ubuntu 12.10

21. Performance Experiments (Recommendation Generation)

24. Performance Experiments (Feedback Box)

26. Quality Experiments  User Satisfaction with feedback box : The objective is to verify if user satisfaction is improved in the presence of a feedback box  The usefulness of a feedback box is evaluated by varying pairwise similarity between users in a group

27. Quality Experiments  Group Size Vs User Satisfaction

28. Recommendation Robustness : Effectiveness and Runtime  Recommendation robustness is a set as a soft constraint which could further be tuned  For example to achieve a robustness weight of 20% after the preference update of a user with k=20, m=125, n=75, the previous recommendation need to be added 15 times

29. Conclusion  They motivate the need for flexible user preferences in group recommendation  They develop a feedback box that computes for each user with evolving preferences the best feedback to provide to maximize user satisfaction in the generated recommendation  They also present robustness to counterbalance the effect of feedback box and ensure group satisfaction  They present a rigorous theoretical and empirical study that corroborates the usefulness of this box