Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Causal Discovery in Recommender Systems: Example and Discussion

Emanuele Cavenaghi, Fabio Stella, Markus Zanker

TL;DR

The resulting causal graph shows that only a few variables effectively influence the analysed feedback signals, which contrasts with the recent trend in the machine learning community to include more and more variables in massive models, such as neural networks.

Abstract

Causality is receiving increasing attention by the artificial intelligence and machine learning communities. This paper gives an example of modelling a recommender system problem using causal graphs. Specifically, we approached the causal discovery task to learn a causal graph by combining observational data from an open-source dataset with prior knowledge. The resulting causal graph shows that only a few variables effectively influence the analysed feedback signals. This contrasts with the recent trend in the machine learning community to include more and more variables in massive models, such as neural networks.

Causal Discovery in Recommender Systems: Example and Discussion

TL;DR

The resulting causal graph shows that only a few variables effectively influence the analysed feedback signals, which contrasts with the recent trend in the machine learning community to include more and more variables in massive models, such as neural networks.

Abstract

Causality is receiving increasing attention by the artificial intelligence and machine learning communities. This paper gives an example of modelling a recommender system problem using causal graphs. Specifically, we approached the causal discovery task to learn a causal graph by combining observational data from an open-source dataset with prior knowledge. The resulting causal graph shows that only a few variables effectively influence the analysed feedback signals. This contrasts with the recent trend in the machine learning community to include more and more variables in massive models, such as neural networks.
Paper Structure (10 sections, 2 figures)

This paper contains 10 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Causal Discovery Process
  3. Remove features
  4. Discretize features
  5. Build Prior-Knowledge
  6. Structure Learning Phase
  7. Average Causal Graph
  8. Learned Causal Graph
  9. Conclusions
  10. Full Learned Causal Graph

Figures (2)

  • Figure 1: Markov blanket of the feedback signals in the CG. Nodes are coloured by their semantics: users, items and context features are in blue, red and green, respectively.
  • Figure 2: Full SCM learned from data and expert knowledge.