Text mining in bioinformatics: Past, present and future

Abstract

The peer-reviewed articles and textual data are main source of data in biology. Text mining is solution to extract information from textual data sources that are usually in bulky quantities, messy and disorganized. Dealing with this situation needs to deploy innovative methods and techniques. In this paper, we identify the current heavily used methods for biomedical text mining, their capabilities and developments, some proposed solution and how to evaluate performance. Biomedical specific challenges in text mining context have been studied with respect to proposed answers and then main future trends based on current needs and requirements have been discussed.

Mineração de textos biomédicos: uma revisão bibliométrica

“…A mineração de textos pode ser entendida como o estudo e a prática de extrair padrões, regras e tendências a partir do texto completo de artigos científicos digitais, usando princípios da linguística computacional e métodos analíticos (SULLIVAN, 2001;FAIIAZEE et al, 2012). No domínio biomédico, os trabalhos de mineração de textos buscam, em sua grande maioria, extrair relacionamentos proteína-proteína (ZHANG et al, 2011), gene-gene (TIWARI;CHEN, 2009), gene-proteína (NAEEM et al, 2010, droga-proteína (SHU; HUANG; ZHU, 2012), gene-doença (GONG et al, 2012), entre outros, visando encontrar novos diagnósticos, prevenções e tratamentos HERSH, 2005).…”