DBAIDec 16, 2021

Unsupervised Matching of Data and Text

arXiv:2112.08776v127 citations
Originality Highly original
AI Analysis

This addresses the challenge of integrating unstructured text and structured data for applications like question answering and search, offering a flexible solution that is particularly useful in domain-specific contexts.

The paper tackles the problem of matching textual content with structured data in an unsupervised setting, introducing a framework that builds a fine-grained graph and learns word embeddings to achieve effective matching across different granularities, with experiments showing it outperforms word embeddings and fine-tuned language models in quality and execution times.

Entity resolution is a widely studied problem with several proposals to match records across relations. Matching textual content is a widespread task in many applications, such as question answering and search. While recent methods achieve promising results for these two tasks, there is no clear solution for the more general problem of matching textual content and structured data. We introduce a framework that supports this new task in an unsupervised setting for any pair of corpora, being relational tables or text documents. Our method builds a fine-grained graph over the content of the corpora and derives word embeddings to represent the objects to match in a low dimensional space. The learned representation enables effective and efficient matching at different granularity, from relational tuples to text sentences and paragraphs. Our flexible framework can exploit pre-trained resources, but it does not depends on their existence and achieves better quality performance in matching content when the vocabulary is domain specific. We also introduce optimizations in the graph creation process with an "expand and compress" approach that first identifies new valid relationships across elements, to improve matching, and then prunes nodes and edges, to reduce the graph size. Experiments on real use cases and public datasets show that our framework produces embeddings that outperform word embeddings and fine-tuned language models both in results' quality and in execution times.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes