Description-Based Text Similarity
This addresses the need for more effective text retrieval based on abstract descriptions, offering a novel approach that could benefit information seeking systems.
The paper tackled the problem of text similarity for retrieval by proposing description-based similarity, showing that current embeddings are inadequate and their new model significantly improves nearest neighbor search performance.
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.