Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AE SemRL: Learning Semantic Association Rules with Autoencoders

Erkan Karabulut, Victoria Degeler, Paul Groth

TL;DR

This study proposes an Autoencoder-based approach to learn and extract association rules from time series data (AE SemRL) and argues that in the presence of semantic information related to time series data sources, semantics can facilitate learning generalizable and explainable association rules.

Abstract

Association Rule Mining (ARM) is the task of learning associations among data features in the form of logical rules. Mining association rules from high-dimensional numerical data, for example, time series data from a large number of sensors in a smart environment, is a computationally intensive task. In this study, we propose an Autoencoder-based approach to learn and extract association rules from time series data (AE SemRL). Moreover, we argue that in the presence of semantic information related to time series data sources, semantics can facilitate learning generalizable and explainable association rules. Despite enriching time series data with additional semantic features, AE SemRL makes learning association rules from high-dimensional data feasible. Our experiments show that semantic association rules can be extracted from a latent representation created by an Autoencoder and this method has in the order of hundreds of times faster execution time than state-of-the-art ARM approaches in many scenarios. We believe that this study advances a new way of extracting associations from representations and has the potential to inspire more research in this field.

AE SemRL: Learning Semantic Association Rules with Autoencoders

TL;DR

This study proposes an Autoencoder-based approach to learn and extract association rules from time series data (AE SemRL) and argues that in the presence of semantic information related to time series data sources, semantics can facilitate learning generalizable and explainable association rules.

Abstract

Association Rule Mining (ARM) is the task of learning associations among data features in the form of logical rules. Mining association rules from high-dimensional numerical data, for example, time series data from a large number of sensors in a smart environment, is a computationally intensive task. In this study, we propose an Autoencoder-based approach to learn and extract association rules from time series data (AE SemRL). Moreover, we argue that in the presence of semantic information related to time series data sources, semantics can facilitate learning generalizable and explainable association rules. Despite enriching time series data with additional semantic features, AE SemRL makes learning association rules from high-dimensional data feasible. Our experiments show that semantic association rules can be extracted from a latent representation created by an Autoencoder and this method has in the order of hundreds of times faster execution time than state-of-the-art ARM approaches in many scenarios. We believe that this study advances a new way of extracting associations from representations and has the potential to inspire more research in this field.
Paper Structure (23 sections, 3 figures, 5 tables, 1 algorithm)

This paper contains 23 sections, 3 figures, 5 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Association Rule Mining and Semantics
  4. Deep Learning-based Association Rule Mining
  5. Problem Definition
  6. Input
  7. Binding
  8. Output
  9. Learning Semantic Association Rules - AE SemRL
  10. Autoencoder Architecture
  11. Pipeline
  12. Rule Extraction from Autoencoders
  13. Evaluation
  14. Setup
  15. Evaluation Metrics
  16. ...and 8 more sections

Figures (3)

  • Figure 1: Full pipeline of operations for AE SemRL.
  • Figure 2: An illustration of association rule extraction from a trained Autoencoder.
  • Figure 3: (Bottom-Left) Execution time of FP-Growth (confidence = 0.8). (Bottom-Right) Effect of the number of time series data sources on execution time (AE SemRL's similarity threshold = 0.5, support=0.25, confidence=0.8 for FP-Growth). (Top-Left) The number of rules learned by AE SemRL. (Top-Right) Execution time of AE SemRL per different number of antecedents.