| Abstrakt: |
An ontology is a formal representation of domain knowledge as a set of concepts and their relationships in structured format. Besides various applications, ontology has been frequently used for the development of movie recommender systems that aim to predict the rating a user would give to a movie. Though recommender system is a well-studied research field, issues like cold-start, data sparsity, limited-content, and long tail are still open challenges. Context-aware recommender systems have been considered by many researchers to deal with some of these issues. However, non-existence of domain ontology containing contextual information is the major hindrance for the development of such systems. This is mainly due to the fact that datasets used in most of the existing context-aware recommender systems are either synthetic or generated through surveys. In this paper, we present the development of a context-aware movie ontology (CAMO) that contains movie concepts, relationships, and various representational and interactional contextual features. Since existing movie databases do not contain all contextual features, we have generated a real-world movie dataset from Linked Open Data (LOD) and movie databases like IMDB and Rotten Tomatoes that contain complete context-based movie profiles. CAMO contains the conceptualization of total 1103 movies. Since users generally express their opinion over the interactional features like story, direction, music, visual effects, etc. of the movies, CAMO also contains aspect-based sentiment polarity identified from the reviews generated by 78056 users and 235 critics. The usefulness of CAMO is empirically evaluated using two similarity measures and two state-of-the-art recommendation methods, and it seems a useful movie knowledgebase for the development of context-aware movie recommendation systems. The source code of CAMO is uploaded at https://github.com/vineet-sejwal/CAMO for research and academic reference purposes. The generated movie dataset containing movie details, user ratings, and reviews generated by both users and critics is also uploaded, and it can be used as a benchmark dataset to evaluate movie recommender systems. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink