• Corpus ID: 14373328

SAR: A Semantic Analysis Approach for Recommendation

  title={SAR: A Semantic Analysis Approach for Recommendation},
  author={Han Xiao and Minlie Huang and Xiaoyan Zhu},
Recommendation system is a common demand in daily life and matrix completion is a widely adopted technique for this task. However, most matrix completion methods lack semantic interpretation and usually result in weak-semantic recommendations. To this end, this paper proposes a Semantic Analysis approach for Recommendation systems (SAR), which applies a two-level hierarchical generative process that assigns semantic properties and categories for user and item. SAR learns semantic… 
1 Citations

Figures and Tables from this paper

Content-based Re-ranking Method for Recommendation

The proposed content-based re-ranking method can make full use of the data set, and effectively supplement and re-rank the recommended results produced by some deep learning algorithms based on the edge information to effectively improve the diversity while preserving the accuracy of recommendation results.



A Neighborhood-based Matrix Factorization Technique for Recommendation

This paper incorporates the coupling relationship analysis to capture the under-discovered relationships between items and proposes a neighborhood-based matrix factorization model, which considers both the explicit and implicit correlations between items, to suggest the more reasonable items to user.

Learning Optimal Social Dependency for Recommendation

A novel recommendation method, named probabilistic relational matrix factorization (PRMF), which aims to learn the optimal social dependency between users to improve the recommendation accuracy, with or without users' social relationships.

Multi-Aspect Collaborative Filtering based on Linked Data for Personalized Recommendation

This paper proposes an approach of utilizing semantic information of items to extract essential aspects to perform multi-aspect collaborative filtering to recommend users with items in a personalized manner.

LLORMA: Local Low-Rank Matrix Approximation

This paper proposes, analyzes, and experiment with two procedures, one parallel and the other global, for constructing local matrix approximations, which approximate the observed matrix as a weighted sum of low-rank matrices.

Personalized Geo-Specific Tag Recommendation for Photos on Social Websites

This work focuses on the personalized tag recommendation task and tries to identify user-preferred, geo-location-specific as well as semantically relevant tags for a photo by leveraging rich contexts of the freely available community-contributed photos.

Localized matrix factorization for recommendation based on matrix block diagonal forms

It is shown formally that the LMF framework is suitable for matrix factorsization and that any decomposable matrix factorization algorithm can be integrated into this framework, which has the potential to improve prediction accuracy by factorizing smaller and denser submatrices independently.

Probabilistic Matrix Factorization

The Probabilistic Matrix Factorization (PMF) model is presented, which scales linearly with the number of observations and performs well on the large, sparse, and very imbalanced Netflix dataset and is extended to include an adaptive prior on the model parameters.

Optimal Discrete Matrix Completion

A novel optimal discrete matrix completion model, able to learn optimal thresholds automatically and also guarantees an exact low-rank structure of the target matrix is proposed, which is able to handle online data and easy to parallelize.

An experimental study on implicit social recommendation

  • Hao Ma
  • Computer Science
  • 2013
This study conducts comprehensive experimental analysis on three recommendation datasets and indicates that implicit user and item social information, including similar and dissimilar relationships, can be employed to improve traditional recommendation methods.

Collaborative Filtering with Graph Information: Consistency and Scalable Methods

This work formulate and derive a highly efficient, conjugate gradient based alternating minimization scheme that solves optimizations with over 55 million observations up to 2 orders of magnitude faster than state-of-the-art (stochastic) gradient-descent based methods.