Virtual user approach for group recommender systems using precedence relations

Kagita, Venkateswara Rao; Pujari, Arun K.; Padmanabhan, Vineet

doi:10.1016/j.ins.2014.08.072

Cited by 39 publications

(16 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…13 The second approach constructs a joint user profile for all the group members by merging different individual's profiles or preferences and then generates recommendations for this artificial user standing for the whole group. 3 However, both strategies ignore the social influence between the group members when reaching a consensus. We overcome this limitation in our study by fully considering the role of social influence in 3 group recommender systems.…”

Section: Group Recommendationmentioning

confidence: 99%

“…2 Kagita and his colleagues proposed a novel virtual user strategy using precedence relations to develop a new scheme for group recommender systems. 3 Compared with recommendations made for individual users, numerous new issues arise with the advent of group recommendations. Among differences between individual and group recommender systems, one of the most significant is social influence, 4-6 a process in which people directly or indirectly influence thoughts, feelings, and actions of others.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Social Influence Approach for Group User Modeling in Group Recommendation Systems

Guo

Zhu

et al. 2016

IEEE Intell. Syst.

View full text Add to dashboard Cite

While many studies on typical recommender systems focus on making recommendations to individual users, many social activities involve groups of users. Issues related to group recommendations are increasingly becoming hot research topics. Among differences between individual and group recommender systems, the most significant one is social factors of group users.Social factors, including personality, expertise factor, interpersonal relationships, and preference similarities, widen the gap between group and individual recommendations. Here, a new approach focusing on the impact of social factors on group recommender systems is proposed. A computational model integrating the influences of personality, expertise factor, interpersonal relationships, and preference similarities is described in detail. Comparative experiments are conducted on two datasets.The experimental results show that the proposed approach can provide more accurate and satisfactory group recommendations, especially when social influences are significant.

show abstract

Section: Group Recommendationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Social Influence Approach for Group User Modeling in Group Recommendation Systems

Guo

Zhu

et al. 2016

IEEE Intell. Syst.

View full text Add to dashboard Cite

show abstract

“…After the PLSA [31] was introduced, principal component analysis (PCA) [2], VarSelectSVD [3], and LTSMF [6] were proposed on top of the matrix decomposition approach. For group users, Venkateswara et al proposed a novel strategy using precedence relations and they show the performance of their algorithm is better in terms of precision and recall [33]. However, unlike LainRec, none of these methods considers the user/item influence level and influence dissemination.…”

Section: Robust Rssmentioning

confidence: 99%

Toward trustworthy social network services: A robust design of recommender systems

Noh

Lee

et al. 2015

J. Commun. Netw.

View full text Add to dashboard Cite

In recent years, electronic commerce and online social networks (OSNs) have experienced fast growth, and as a result, recommendation systems (RSs) have become extremely common. Accuracy and robustness are important performance indexes that characterize customized information or suggestions provided by RSs. However, nefarious users may be present, and they can distort information within the RSs by creating fake identities (Sybils). Although prior research has attempted to mitigate the negative impact of Sybils, the presence of these fake identities remains an unsolved problem. In this paper, we introduce a new weighted link analysis and influence level for RSs resistant to Sybil attacks. Our approach is validated through simulations of a broad range of attacks, and it is found to outperform other state-of-the-art recommendation methods in terms of both accuracy and robustness.

show abstract

“…Recommending to group of people and satisfying them is even more complicated than recommending to individuals. There are several ways of extending personal recommendation to a group recommendation [13], [14], [15] and are categorized into merging profiles, merging recommendation and merging results. Merge profiles approach creates the recommendation list from merging the profiles of each group member which in turn is based on individual ratings for content or genre.…”

Section: Introductionmentioning

confidence: 99%

“…Merge profiles approach creates the recommendation list from merging the profiles of each group member which in turn is based on individual ratings for content or genre. For instance, consider two user group [15], [16], [17]. In the merging recommendation approach, GRS first generates a recommendation for each user individually based on personalized profiles.…”

Section: Introductionmentioning

confidence: 99%

A Novel Social-Choice Strategy for Group Modeling in Recommender Systems

Kagita

Meka

Padmanabhan

2015

2015 International Conference on Information Technology (ICIT)

Self Cite

View full text Add to dashboard Cite

Personalized recommender systems are usually designed to provide recommendations adapted to the preferences of a single user. Group recommender systems on the other hand suggest items to a group by combining individual models into a group model. This group model allows to merge the preferences of the individual members of a group and thereby derive a group preference for each item by using different group modeling strategies. In this paper we first critically analyse various group modeling strategies as outlined in the literature and point out the limitations of each of them. Thereafter we propose a group modeling technique called Most Members Merry (MMM) strategy to address those limitations. Our experimental results show that MMMF exhibits high performance in terms of precision and recall as compared to the existing strategies.

show abstract

Virtual user approach for group recommender systems using precedence relations

Cited by 39 publications

References 28 publications

A Social Influence Approach for Group User Modeling in Group Recommendation Systems

A Social Influence Approach for Group User Modeling in Group Recommendation Systems

Toward trustworthy social network services: A robust design of recommender systems

A Novel Social-Choice Strategy for Group Modeling in Recommender Systems

Contact Info

Product

Resources

About