Description-Based Text Similarity

Shauli Ravfogel, Valentina Pyatkin, Amir David Nissan Cohen, Avshalom Manevich, Yoav Goldberg

Research Area: Data, LMs on diverse modalities and novel applications

Keywords: semantics, similarity, description based similarity, retrieval

TL;DR: We define the notion of description-based similarity, use LLMs to create a tailored dataset for this kind of similarity, and train a corresponding retrieval model.

Abstract:

Identifying texts with a given semantics is central for many information seeking scenarios. Similarity search over vector embeddings appear to be central to this ability, yet the similarity reflected in current text embeddings is corpus-driven, and is inconsistent and sub-optimal for many use cases. What, then, is a good notion of similarity for effective retrieval of text?

We identify the need to search for texts based on abstract descriptions of their content, and the corresponding notion of \emph{description based similarity}. We demonstrate the inadequacy of current text embeddings and propose an alternative model that significantly improves when used in standard nearest neighbor search. The model is trained using positive and negative pairs sourced through prompting a LLM, demonstrating how data from LLMs can be used for creating new capabilities not immediately possible using the original model.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the COLM Code of Ethics on https://colmweb.org/CoE.html

Author Guide: I certify that this submission complies with the submission instructions as described on https://colmweb.org/AuthorGuide.html

Submission Number: 96