Workshop ID W07
Workshop title Text mining for systematic reviews
   
Methods Group N/A
Facilitators James Thomas, EPPI-Centre, Social Science Research Unit, Institute of Education, London, UK
Contact James Thomas
   
Time Monday 12 Oct, 14:00-15:30h
Room 311
   
Topic category Searching and information retrieval
Target audience Review authors, methodologists, Trials Search Coordinators, information scientists, researchers
Level of knowledge Intermediate
Type of workshop Discussion
   
Workshop abstract Objectives:
A significant proportion of systematic reviewing time is spent searching for research and deciding whether or not it is relevant. This workshop will describe, and then be a forum for discussing, emerging technologies to alleviate this problem. Objectives:
  • To present text mining services to participants.
  • To discuss their use in systematic reviews.

Description:
We will begin by describing and demonstrating new tools – using 'text mining' – to help researchers to collect, maintain, interpret, curate, and discover hidden knowledge in text. We will present:

  • Term extraction, named entity recognition, query expansion and document classification: locating relevant documents in poorly indexed electronic sources.
  • Association mining to find snippets of information.
  • Summarisation: selecting significant information from the selected documents to 'map' research activity quickly and efficiently.
The workshop will then move to a structured discussion covering both potential uses of text mining tools and also issues of bias and transparency that their use might raise. Demonstrations of existing services will be presented to the users for hands on experience. Participants will learn about new technologies that might assist them in their work, and their views will feed through into the future development of the tools.
   
Other contributors Sophia Ananiadou, National Centre of Text Mining, UK