Publication Type

Journal Article

Version

publishedVersion

Publication Date

3-2024

Abstract

Knowledge graphs can be used to enhance text search and access by augmenting textual content with relevant background knowledge. While many large knowledge graphs are available, using them to make semantic connections between entities mentioned in the textual content remains to be a difficult task. In this work, we therefore introduce contextual path generation (CPG) which refers to the task of generating knowledge paths, contextual path, to explain the semantic connections between entities mentioned in textual documents with given knowledge graph. To perform CPG task well, one has to address its three challenges, namely path relevance, incomplete knowledge graph, and path well-formedness. This paper designs a two-stage framework the comprising of the following: (1) a knowledge-enabled embedding matching and learning-to-rank with multi-head self attention context extractor to determine a set of context entities relevant to both the query entities and context document, and (2) a non-monotonic path generation method with pretrained transformer to generate high quality contextual paths. Our experiment results on two real-world datasets show that our best performing CPG model successfully recovers 84.13% of ground truth contextual paths, outperforming the context window baselines. Finally, we demonstrate that non-monotonic model generates more well-formed paths compared to the monotonic counterpart.

Keywords

information retrieval, knowledge graph, contextual path generation, generation model

Discipline

Databases and Information Systems | Management Information Systems | Numerical Analysis and Scientific Computing

Research Areas

Data Science and Engineering

Publication

ACM Transactions on Management Information Systems

Volume

15

Issue

1

First Page

1

Last Page

28

ISSN

2158-656X

Identifier

10.1145/3627994

Publisher

Association for Computing Machinery (ACM)

Copyright Owner and License

Authors

Creative Commons License

Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License.

Additional URL

https://doi.org/10.1145/3627994

Share

COinS