Identifying, Indexing, and Ranking Chemical Formulae and Chemical Names in Digital Documents

Abstract

End users use chemical search engines to search for chemical formulae and chemical names. Chemical search engines identify and index chemical formulae and chemical names appearing in text documents to support efficient search and retrieval in the future. Identifying chemical formulae and chemical names in text automatically has been a hard problem that has met with varying degrees of success in the past. We propose algorithms for chemical formula and chemical name tagging using Conditional Random Fields (CRFs) and Support Vector Machines (SVMs) that achieve higher accuracy than existing (published) methods. After chemical entities have been identified in text documents, they must be indexed. In order to support user-provided search queries that require a partial match between the chemical name segment used as a keyword or a partial chemical formula, all possible (or a significant number of) sub-formulae of formulae that appear in any document and all possible sub-terms (e.g., 'methyl') of chemical names (e.g., 'methylethyl ketone') must be indexed. Indexing all possible sub-formulae and sub-terms results in an exponential increase in the storage and memory requirements, and the time taken to process the indices. We propose techniques to prune the indices significantly without reducing the quality of the returned results significantly. Finally, we propose multiple query semantics to allow users to pose different types of partial search queries for chemical entities. We demonstrate empirically that our search engines improve the relevance of the returned results for search queries involving chemical entities.

Chemical entity extraction using CRF and an ensemble of extractors

“…Our interest was initially in the Chemical Entity Mention (CEM) task as it is the prerequisite to the following Chemical Document Indexing (CDI) task. We started by running a distribution of ChemXSeer's formula and name extractor [ 7 - 9 ] that is released for the general public on the training and development datasets. The tagger is based on Conditional Random Fields (CRF) [ 4 ] models, with additional rules for pre and post processing documents.…”

Identification of Inorganic Chemical Formulas based on Support Vector Machine and SURF Key Point Descriptor

“…The model uses the concept of trees to store individual IS which was more complex in terms of memory and time. Sun et al [6] have proposed an algorithm for ranking chemical formulae and tagging chemical names in digital documents. This algorithm uses Conditional Random Fields (CRFs) and Support Vector Machines (SVMs).…”

A Corpus of Very Short Scientific Summaries

“…We, likewise, employ customised tokenisation and named entity recognition while pre-processing the corpus (Corbett and Boyle, 2018) to enable future researchers to forgo the NER step. Further related applications of NLP include information retrieval (Sun et al, 2011;Hawizy et al, 2011) and literature mining (Zaslavsky et al, 2017;Öztürk et al, 2020), while lessons learned from deep learning in NLP have been applied to strings representing chemical structures to successfully discover new potential antibiotics (Stokes et al, 2020).…”