Mining Social Networks for Recommendation 阅读总结

2012-09-14

Recommender systems

• Web Search:

  • need contents

  • can not feed to users’ different interests

• Data:

  • User action

  • User profile

• Tasks:

  • Rating prodiction

  • Top-N recommendation

  • Link recommendation(only if social network)

• Privacy Issues

  • More data,better recommendation

  • User need be able to set choice

• Collaborate Filtering

  • Nearest Neighbor-based approach

    • User-based

    • Item-based

  • Model-based approch

    • MF(Matrix Factorization)

      • Outperforms the NN-based CF

• Content-based

  • Keywords

    • TF-IDF

• Hybrid recommander System

  • Approach 1: combine separate recommenders Combine results

    • e.g. using linear combination or voting[Pazzani 1999]

  • Approach 2: add aspects of content-based method to CF. [Pazzani 1999]

  • Approach 3: add aspects of CF to content-based method [Soboroff et al., 1999] .

  • Approach 4: Unified recommendation model

    • E.g., combine topic model, i.e. Latent Dirichlet Allocation, with MF.

  • Latent Dirichlet Allocation (LDA) [Blei et al., 2003]

  • Graphical Model

  • Collaborative topic regression [Wang et al., 2011]

• Performance Evaluation

  • Cross-validation on offline dataset

    • Limitations:

      • Measures only accuracy of recommendation

      • Does not measure other aspects such as diversity

      • Does not measure how recommendation change user behavior

      • this is the ultimate goal of a recommender!

  • In industry

    • Want to evaluate user satisfaction and business profit

    • A/B test in online system

    • Evaluate Measures:

      • click-through rate usage

      • return rate of customers profit

• Challenges:

  • Cold start problem

    • for Users

      • Typically,~50% of users cold start

      • CF fails-there are no similar users(User-based) and no Item rating to aggregate

    • for Items

      • CF fails

      • Content based methods works

Recommendation in social networks

• Formation and Evolution are affect by many effectsTrust network

  • Self-interst

  • Social and resource exchange

  • Balance

  • Homophily

  • Proximity

• Social rating network

  • users are associated with item ratings

  • Social influence:

    • ratings are influenced by ratings of friends

    • ratings are influenced by ratings of actorswith similar ratings

  • Benifit:

    • Exploit social influence

      • corelational,influence,transitivity,selection

    • Can deal with code start problem

    • Are more robust to fraud, in particular to profile attacks

  • Challenges

    • Low probability of finding rater at small network distance

    • Noisy ratings at large network distances

    • Social network data is very sensitive

    • Edges in online social networks are of greatly varying reliability / strength?

Mining social networks

• Type

  • Analysis of social influence

  • Models of social rating networks

  • Inference of social networks

Memory based approaches

• Problem defination

  • Input

    • rating matrix

      • real value or binary

    • social matrix

      • weighted or binary

• Datasets for Recommandation in SNsapproaches

  • Epinions-Online product review

    • Users review and rate products in differentcategories

    • Users express trust on other reviewers

    • URL:

      • http://www.trustlet.org/wiki/Epinions_dataset

      • http://alchemy.cs.washington.edu/data/epinions/

  • Flixster

    • Social metworking service for rating movies

    • http://www.sfu.ca/~sja25/datasets/

• approaches

  • Explore the social network for raters

  • Aggregate the ratings to compute prediction

  • Store the social rating network

  • No Learning phase

  • Slow in prediction

  • Most pioneer works for recommendation in SN are memory based approaches.

Model based approaches

• approaches

  • Learn a model

  • Store the model parameters only

  • Extra time for learning

  • Fast in Prediction

  • Most models are based on matrix factorization

Link prediction

• Problem defination

  • Given a user pair (u,v),estimate the probability of creation of the link u->v.

  • Given a user u, recommend a list of top users for u to connect to.

• Finding similariry

  • cosine similarity

  • Pearson corelation

  • Jaccard’s coefficient

Social networks with distrust

• Few works have addressed negative relations

  • [Leskovec et al., 2010]

  • [Kunegis et al., 2009]

  • [Brzozowski et al., 2008]

  • [Guha et al., 2004]

• Prior work shifted the trust to avoid negative values

Summary

• State-of-the-art methods for recommendation in social networksLink Prediction

  • Memory based approaches

    • ModelTrust [Massa 2007], Modified BFS

    • TidalTrust [Golbeck 2005], Modified BFS

    • TrustWalker [Jamali et al., 2009], Random Walk

  • Model based approaches

    • SoRec [Ma et al., 2008], Matrix Factorization

    • FIP [Yan et al., 2011], Matrix Factorization

    • STE [Ma et al., 2009], Matrix Factorization

    • SocialMF[Jamali et al., 2010], Matrix Factorization

    • GSBM[Jamali et al., 2011], Stochastic BlockModel

• Link prodiction

  • Pair-wise profile similarity approaches

    • Information theoretic based definition of similarity

  • Network topology based approaches

    • Common neighbors

  • Path based approaches

    • Katz, Hitting time, RWR, SimRank

• Social Networks with distrust

  • Propagation of distrust

  • Theories behind distrust

  • Recommendation with distrust [Ma et al., 2009.b]

• Future Research Directions

  • Exploring other machine learning models

  • Privacy of recommendation in social networks

    • How to preserve privacy while employing social networks?

  • Improving the diversity of recommendations

    • How to evaluate the diversity?

  • Recommendation of cold-start items

    • They are very important!

  • Recommendation in mobile social networks

    • Distributed algorithm

    • How to exploit the user location?

  • Recommendation in social networks with documents (posts)

    • E.g., Twitter

    • Integration with topic models