Using machine learning for concept extraction on clinical documents from multiple data sources

Abstract

Objective Concept extraction is a process to identify phrases referring to concepts of interests in unstructured text. It is a critical component in automated text processing. We investigate the performance of machine learning taggers for clinical concept extraction, particularly the portability of taggers across documents from multiple data sources. Methods We used BioTagger-GM to train machine learning taggers, which we originally developed for the detection of gene/protein names in the biology domain. Trained taggers were evaluated using the annotated clinical documents made available in the 2010 i2b2/VA Challenge workshop, consisting of documents from four data sources. Results As expected, performance of a tagger trained on one data source degraded when evaluated on another source, but the degradation of the performance varied depending on data sources. A tagger trained on multiple data sources was robust, and it achieved an F score as high as 0.890 on one data source. The results also suggest that performance of machine learning taggers is likely to improve if more annotated documents are available for training. Conclusion Our study shows how the performance of machine learning taggers is degraded when they are ported across clinical documents from different sources. The portability of taggers can be enhanced by training on datasets from multiple sources. The study also shows that BioTagger-GM can be easily extended to detect clinical concept mentions with good performance.

Bio-NER: Biomedical Named Entity Recognition using Rule-Based and Statistical Learners

“…In Expanding exploration or survey of various data source including protein, gene as well Bio Tagger was prepared on it in the domain of medical text mining. The Experimentation or Testing result demonstrated in which Bio Tagger conceivably valuable to extract the protein, gene in the form of huge dataset accommodated for the Training [12]. Content characteristics dependably assume a vital part in Named Entity Recognition; the framework"s execution can be considerably enhanced via expansion of many characteristics.…”

“…In past few researches, the use of orthographic features is widely advocated in [12]- [14]. Our used experimental orthographic features are shown in Table I.…”

“…Methods or techniques are consolidated in light of normal probabilities. In normal of probabilities, the likelihood can be accomplished as, ̂ ∑ whereas represents error probabilities and computed via ( √ ( ) ) and Are free or independent probabilities [12]. Inside and out an examination of order match has been completed which extend from two pairs combinatory to five pairs combinatory or blend.…”

“…The authors [13], [17] has gathered most common prefix and suffix from training data. Whilst [12], [18] the author gathered 23 categories of prefix and suffix data using statistical methods as their own distribution. Our experiment shows the significant improvement in contextual features affixes.…”

A Dual-Layer Semantic Role Labeling System

“…Many researchers have tried to extract concepts from texts (Gelfand et al, 1998;Hovy et al, 2009;Villalon and Calvo, 2009;Dinh and Tamine, 2011;Torii et al, 2011). Hovy narrowed the domain of interest into concepts "below" a given seed term.…”