ISoTrustSeq: a social recommender system based on implicit interest, trust and sequential behaviors of users using matrix factorization.
Vahideh NobahariMehrdad JalaliSeyyed Javad Seyyed MahdaviPublished in: J. Intell. Inf. Syst. (2019)
Keyphrases
- recommender systems
- matrix factorization
- trust aware
- collaborative filtering
- rating prediction
- cold start problem
- data sparsity
- implicit feedback
- factorization methods
- negative matrix factorization
- trust propagation
- low rank
- item recommendation
- user generated content
- user profiles
- user model
- factor analysis
- information overload
- recommendation algorithms
- personalized ranking
- user interests
- recommendation quality
- nonnegative matrix factorization
- recommendation systems
- user preferences
- social influence
- cold start
- social network sites
- product recommendation
- social interaction
- social networks
- binary matrix
- stochastic gradient descent
- social media
- user feedback
- factor matrices
- missing data
- personalized recommendation