TopSig: Topology Preserving Document Signatures
This provides a viable alternative to inverted files for search engines, though it appears incremental as it builds on existing semantic hashing and dimensionality reduction techniques.
The paper tackles the problem of file signatures being inferior to inverted files for text retrieval by introducing TopSig, a new approach that achieves performance comparable to state-of-the-art inverted file systems like Language models and BM25.
Performance comparisons between File Signatures and Inverted Files for text retrieval have previously shown several significant shortcomings of file signatures relative to inverted files. The inverted file approach underpins most state-of-the-art search engine algorithms, such as Language and Probabilistic models. It has been widely accepted that traditional file signatures are inferior alternatives to inverted files. This paper describes TopSig, a new approach to the construction of file signatures. Many advances in semantic hashing and dimensionality reduction have been made in recent times, but these were not so far linked to general purpose, signature file based, search engines. This paper introduces a different signature file approach that builds upon and extends these recent advances. We are able to demonstrate significant improvements in the performance of signature file based indexing and retrieval, performance that is comparable to that of state of the art inverted file based systems, including Language models and BM25. These findings suggest that file signatures offer a viable alternative to inverted files in suitable settings and from the theoretical perspective it positions the file signatures model in the class of Vector Space retrieval models.