Extracting the Names of Genes and Gene Products with a Hidden Markov Model

By Nigel Collier, Chikashi Nobata, Junichi Tsujii. International Conference on Computational Linguistics, 2000

298 citations recorded by the source corpus.

We report the results of a study into the use of a linear interpolating hidden Markov model (HMM) for the task of extracting technical terminology from MEDLINE abstracts and texts in the molecular-biology domain. This is the first stage in a system that will extract event information for automatically updating biology databases. We trained the HMM entirely with bigrams based on lexical and character features in a relatively small corpus of 100 MEDLINE abstracts that were marked-up by domain experts with term classes such as proteins and DNA. Using cross-validation methods we achieved an F-score of 0.73 and we examine the contribution made by each part of the interpolation model to overcoming data sparseness.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/5a10986a068c8cf3c43e433fea451bda93e19661

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Molecular Biology

Citations in the corpus

An architecture for biological information extraction and representation (2004, unknown): “Common methods for extracting information from text vary from simple statistical methods such as term co-occurrences [7, 19] or Hidden Markov Models (HMMs) for term identification and classification [2] to computationally intensive structural methods such as Natural Language Processing (NLP) techniq”
Biomedical Named Entity Recognition using Conditional Random Fields and Rich Feature Sets (2004, methodology): “In a previous attempt to use a Hidden Markov Model to simultaneously recognize multiple biomedical entities (Collier et al., 2000), HMM performance for a particular entity seemed more or less proportional to its frequency in the data.”
A Brief Survey of Text Mining: Classification, Clustering and Extraction Techniques (2017, background): “Hidden Markov Model (HMM) [32, 129, 150], Maximum Entropy Markov Model [33] and Conditional Random Fields (CRF) [128] are the most common sequence-based approaches and CRFs have frequently demonstrated to be better statistical biomedical NER systems.”
The GENIA corpus: an annotated research abstract corpus in molecular biology domain (2002, methodology): “The preliminary version of the corpus[15] was used as a learning and test set of a system that extracts the names of substances involved in and their sources from MEDLINE abstracts[3].”
ProMiner: rule-based protein and gene entity recognition (2005, methodology): “Rule-based approaches [3,4] as well as machine-learning based techniques [5,6] have been proposed.”
Term identification in the biomedical literature (2004, methodology): “For example, Collier and colleagues [33] used Hidden Markov models (HMM) and specific orthographic features (e.”
Accomplishments and challenges in literature data mining for biology (2002, methodology): “(2000), and techniques like Hidden Markov Models (Collier et al., 2000) or Bayesian classifiers trained on k-grams (Wilbur et al., 1999), would further raise sensitivity and specificity in recognizing protein names. The field has now progressed beyond recognizing names and has entered the realm of r”
Comparative experiments on learning information extractors for proteins and their interactions (2005, methodology): “Token classification methods like k-NN, TBL, SVM, and MaxEnt are not directly applicable to extracting interactions; however, we plan to test HMMs on extracting interactions in the near future.”
Bidirectional RNN for Medical Event Detection in Electronic Health Records (2016, background): “, 2001), Hidden Markov Models (Collier et al., 2000), and Max-entropy Markov Models (McCallum et al.”
Tuning support vector machines for biomedical named entity recognition (2002, methodology): “Nobata et al. (1999) and Collier et al. (2000) trained their model with only 100 annotated paper abstracts from the MEDLINE database (National Library of Medicine, 1999), and Yamada et al. (2000) used only 77 annotated paper abstracts.”