Information Retrieval and Recommender Systems: Two Sides of the Same Coin

Here, in this Research survey paper, we will learn RS techniques and draw commonalities between IR and RS, and try to adapt and leverage different IR models to RS

Abstract

IR deals with the study of information retrieval techniques based on the inputs given by the user. There is another field called IF, which also revolves around information processing and presents with information to the user which may be of interest to them. One of the widely used IF techniques is the Recommender Systems which provides personalized suggestions to users based on their interests. If we broadly look at IR and IF (RS), both are quite similar in processing the vast available information and share the most relevant ones based on various retrieval techniques.

Table of Contents

1 Introduction

2 What are recommender systems?

• Model approaches
  • Content-based Filtering Recommender
    • Collaborative Filtering Recommender
      • Model-based Recommender
      • Neighborhood-based Recommender
    • Hybrid Filtering Recommender
  • Matrix Factorization to build recommendation algorithm

3 Recommender system evaluation

3.1 Study of rank accuracy metrics for recommender systems

• Robustness to incompleteness:
  • Sparcity bias
  • Popularity bias
  • Discriminative power

3.2 Implications

4 Pseudo-relevance feedback models for top-N recommendation

4.1 Rocchio framework

• Clustering algorithms
  • Hard-Clustering
  • Soft-Clustering

4.2 Improving neighborhoods

4.3 Improving cosine with an oracle

4.4 Language models for computing neighborhoods

5 Other recommendation tasks

5.1 Long tail liquidation

5.2 User-item group formation

6 Recommender system models for Pseudo-relevance feedback

• Linear Models
  • DLime: learns inter-document similarities
  • TLime: learns inter-term similarities

7 Conclusion

7.1 Future Directions

Authors

Archit Bansal

Department of Textile & Fibre Engineering
Indian Institute of Technology Delhi
Hauz Khas, New Delhi, India
tt1180924@iitd.ac.in

Saurav Mittal

Department of Chemical Engineering
Indian Institute of Technology Delhi
Hauz Khas, New Delhi, India
ch1180243@iitd.ac.in