Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

xLP: Explainable Link Prediction for Master Data Management

Balaji Ganesan, Matheen Ahmed Pasha, Srinivasa Parkala, Neeraj R Singh, Gayatri Mishra, Sumit Bhatia, Hima Patel, Somashekar Naganna, Sameep Mehta

TL;DR

This demo presents explanations for link prediction in a creative way, to allow users to choose explanations they are more comfortable with and to allow users to choose explanations they are more comfortable with.

Abstract

Explaining neural model predictions to users requires creativity. Especially in enterprise applications, where there are costs associated with users' time, and their trust in the model predictions is critical for adoption. For link prediction in master data management, we have built a number of explainability solutions drawing from research in interpretability, fact verification, path ranking, neuro-symbolic reasoning and self-explaining AI. In this demo, we present explanations for link prediction in a creative way, to allow users to choose explanations they are more comfortable with.

xLP: Explainable Link Prediction for Master Data Management

TL;DR

This demo presents explanations for link prediction in a creative way, to allow users to choose explanations they are more comfortable with and to allow users to choose explanations they are more comfortable with.

Abstract

Explaining neural model predictions to users requires creativity. Especially in enterprise applications, where there are costs associated with users' time, and their trust in the model predictions is critical for adoption. For link prediction in master data management, we have built a number of explainability solutions drawing from research in interpretability, fact verification, path ranking, neuro-symbolic reasoning and self-explaining AI. In this demo, we present explanations for link prediction in a creative way, to allow users to choose explanations they are more comfortable with.
Paper Structure (10 sections, 4 figures, 2 tables)

This paper contains 10 sections, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Link Prediction
  4. Explainability
  5. Implementation
  6. Link Verification with External Information
  7. Anchors for Link Explanation
  8. Path Ranking based Link Explanation
  9. Evaluation
  10. Conclusion

Figures (4)

  • Figure 1: Predicting links to watch-list nodes. A list of COVID19 persons can be uploaded as watch-list and their links to people in a master data could be predicted.
  • Figure 2: xLP: Explainable Link Prediction. The above image shows a subgraph with links predicted. On the right we see verification text for the predicted link and a comparison of the nodes involved in the link.
  • Figure 3: Path Ranking based explanation for Link Prediction. Data Stewards can explore the already existing paths (if any) and understand the predicted link. They can also provide feedback.
  • Figure 4: Anchors Explanation