A Content-boosted matrix factorization technique via user-item subgroups

Download
2014
Aslan Oğuz, Evin
This thesis mainly focuses on improving the recommendation accuracy of collaborative filtering (CF) algorithm via merging two successful approaches. Since CF algorithmsgrouplike-mindedusers,atechniquecalledMulticlassCo-Clustering(MCoC) is used in order to group like-minded users more effectively. Since, CF approaches lack incorporating content information, a content-boosted CF approach that embeds content information into recommendation process is used. In the MCoC, a user or an item can belong to zero, one or more subgroups. Thus, it is possible to predict the rating scores of users to items present in the same subgroup. However the prediction resultsfor all users anditems arenot obtainedby MCoC,since auser oran itemmay belong to zero subgroups. Therefore, content-boosted CF algorithm is applied to the whole set of users and items besides subgroups and finally the results are merged. The content-boosted approach, on the other hand, considers content information in the recommendation process. As content, the genres of movies are embedded into the item latent factor vector in the matrix factorization technique. To sum up, the content-boosted algorithm is applied to the subgroups and the whole set, and the obtained results are merged. Hence the recommendation accuracy is improved.

Suggestions

A Hypergraph based framework for representing aggregated user profiles, employing it for a recommender system and personalized search through a hypernetwork method
Tarakçı, Hilal; Manguoğlu, Murat; Çiçekli, Fehime Nihan; Department of Computer Engineering (2017)
In this thesis, we present a hypergraph based user modeling framework to aggregate partial profiles of the individual and obtain a complete, semantically enriched, multi-domain user model. We also show that the constructed user model can be used to support different personalization services including recommendation. We evaluated the user model against datasets consisting of user's social accounts including Facebook, Twitter, LinkedIn and Stack Overflow. The evaluation results confirmed that the proposed use...
A Hybrid geo-activity recommendation system using advanced feature combination and semantic activity similarity
Sattari, Masoud; Toroslu, İsmail Hakkı; Department of Computer Engineering (2013)
In this study, a new method for analyzing and representing the discriminative information, distributed in functional Magnetic Resonance Imaging (fMRI) data, is proposed. For this purpose, a local mesh with varying size is formed around each voxel, called the seed voxel. The relationships among each seed voxel and its neighbors are estimated using a linear regression equation by minimizing the expectation of the squared error. This squared error coming from linear regression is used to calculate various info...
A Graph-based core model and a hybrid recommender system for TV users
Taşcı, Arda; Çiçekli, Fehime Nihan; Department of Computer Engineering (2015)
This thesis proposes a core model to represent user profiles in a graph-based environment which can be the base of different recommender system approaches as well as other cutting edge applications for TV domain. The proposed graph-based core model is explained in detail with node types, properties and edge weight metrics. The capabilities of this core model are described in detail. Moreover, in this thesis, a hybrid recommender system based on this core model is presented with its design, development and e...
Frequency-driven late fusion-based word decomposition approach on the phrase-based statistical machine translation systems
Önem, İsmail Melih; Alpaslan, Ferda Nur; Department of Computer Engineering (2013)
Recommender systems are very popular in information systems and in the research community, where many different approaches geared towards giving better recommendations have been proposed. In this thesis, we propose a methodology that uses social network information to improve the performance of recommender systems. Our proposed methodology heuristically improves the success rate and performance of recommendation algorithms using social distance measures on a dataset that comprises people in professional occ...
A recommendation framework using ontological user
Yaman, Çağla; Çiçekli, Fehime Nihan; Department of Computer Engineering (2011)
In this thesis, a content recommendation system has been developed. The system makes recommendations based on the preferences of the users on some aspects of the content and also preferences of similar users. The preferences of a user are extracted from the choices of that user made in the past. Similarities between users are defined by the similarities of their preferences. Such a system requires both qualified content and user information. The proposed system uses semantic user and content profiles to mor...
Citation Formats
E. Aslan Oğuz, “A Content-boosted matrix factorization technique via user-item subgroups,” M.S. - Master of Science, Middle East Technical University, 2014.