Movies Recommenders Systems: Automation of the Information and Evaluation Phases in a Multi-criteria Decision-Making Process

Abstract

The authors' interest is focused on advanced recommending functionalities proposed by more and more Internet websites w.r.t. the selection of movies, e-business sites, or any e-purchases. These functionalities often rely on the Internet users' opinions and evaluations. A « movie-recommender » application is presented. Recommender websites generally propose an aggregation of the user's evaluations critics according to different relevant criteria w.r.t. the application. The authors propose an Information Processing System (IPS) to collect, process and manage as automatically as possible these opinions or critics to support this multi criteria evaluation for recommendation. The RS (Recommender System) firstly proposes information extraction techniques in order to classify the available users' critics w.r.t. the criteria implied in the evaluation process and to automatically associate numerical scores to these critics. Then multicriteria techniques are introduced to numerically evaluate, compare and rank the competing movies the critics are reported to. Finally the RS is presented as an interactive Decision-Making Support System (DMSS) relying on a sensibility analysis of the movies ranking. A particular attention is paid to the automation of the information phase in the decision-making process: movie comments cartography according to users' evaluation criteria and attribution of a partial score to each critic considered as the expression of a value judgment in natural language.

Analysis and Classification of Multi-Criteria Recommender Systems

“…In many cases, the user model is initially 'Empty,' and then slowly created throughout the users' interactions with the system. Few multi-criteria recommendation systems use some 'Stereotyping' (demographic) technique, some initial 'Training set,' or some other 'Heuristic' method in order to build [4,17,54,109] Value-focused models [5,20,27,28,31,37,42,45,46,48,52,58,63,64,72,73,76,79,80,82,95,96,99,100,101,113,114] Optimization models [83,103] Outranking relations [74,75,112] Type…”

Çok Kriterli Ürün-Tabanlı İşbirlikçi Filtrelemede Ağırlıklandırma Yöntemlerini Kullanarak Tahmin Performansının Arttırılması

“…Çok kriterli öneri sistemlerinin uygulanmaya başlanması 2000'li yılların başlarına dayanır. İlk olarak Plantié, Montmain, ve Dray (2005) İşbirlikçi filtreleme sistemlerinde benzerlik tabanlı yaklaşımlar kullanılırken, kullanıcılar / ürünler arasında ortak derecelendirilen ürünlerin / kullanıcıların sayısı önemli bir faktördür. Öneri sistemlerinin yapısı düşünüldüğünde, çoğu ürünün derecelendirilmediği görülür.…”

Artificial Neural Networks and Particle Swarm Optimization Algorithms for Preference Prediction in Multi-Criteria Recommender Systems

“…Though research on recommender systems has a long history in relation to information filtering and retrieval, the study of recommender systems as an independent research field began with the emergence of papers on collaborative filtering in the mid-1990s [16][17][18][19]. Furthermore, in 2005, Michael et al [20] proposed a decision-making support system that combines text-mining techniques with multi-criteria analysis techniques for recommending movies to users. Their paper was among the studies that marked the beginning of extending traditional collaborative filtering techniques to multi-criteria recommendation problems [21].…”