|
S.no | Publication | Dataset | Similarity measure | Environment application | Social network | Ontology used | Linked Open Data | Technique used |
|
(1) | Addressing the user cold start with cross-domain collaborative filtering: exploiting item metadata in matrix factorization [5] | Facebook | Other | Book, movies, music (cross domain) | Facebook | No | DBpedia | Cross-domain collaborative filtering using matrix factorizartion models |
(2) | Latent factor representations for cold-start video recommendation [6] | Video Emotion, Amazon Product | No | Videos (domain independent) | No | No | No | Learning latent factor representation for videos based on modeling the emotional connection between user and item |
(3) | Using linked data to build open, collaborative recommender systems [12] | Smart Radio | Binary cosine similarity | Music | No | No | DBpedia, Myspace | Used linked data about items for collaborative filtering algorithm |
(4) | An effective recommender algorithm for cold-start problem in academic social networks [7] | Created MyExpert app | Others | Academic item | Academic social network | No | No | Enhanced content-based algorithm using social networking |
(5) | A method to solve cold-start problem in recommendation system based on social network sub-community and ontology decision model [8] | MovieLens | Pearson similarity | Videos | No information | Taxanomy | No | Combining social subcommunity division and ontology decision model |
(6) | Exploring social network information for solving cold start in product recommendation [9] | Douban Website | Other | Books | Douban Website | No | No | Used social network textual information to model user interest and item |
(7) | Using semantic web to reduce the cold-start problems in recommendation systems [47] | MovieLens | Composed Similarity, Property Similarity, Jaccard, Jaro Winkler | Movie | No | Yes | No | Used semantic web structure and ontology |
(8) | Proposed approach | MovieLens, Yahoo Webscope | New similarity measure and Euclidean distance | Domain independent | Facebook | Yes | DBpedia | Recommendation based on user and item clustering, improved ontology similarity, and taking into account the social network and Linked Open Data features |
(9) | Social network data to alleviate cold start in recommender system: A systematic review [32] | Review paper: Systematic combining papers published between 2011 and 2017 on dealing cold-start problems using social network data |
|