Text and Data Mining in Medicine

Image credit: Open Knowledge P3260857, https://flic.kr/p/e87SvU

For the first time today we had a visit from a user who wanted advice on how to undertake text and data mining using already created data. There has been a rise in the profile of Clinical Informatics as a discipline. It has been said that the patient is the most underused resource in medicine, and clinical informatics experts are now able to interact with not only the reports of research projects but the data from those projects, as well as data in other databases involving data from the users of health services and other databases.

The ethical and legal issues around this work are complex and important, for instance the necessity to ensure that the data is anonymised. At Cambridge University we now have an Office of Scholarly Communication to which researchers can be directed. This office  advises on  Open Access and the Data repositories, and Data Management . They train researchers and librarians to be aware of some of these issues and signpost them to the more specific services where appropriate. I was pleased to find a Text and Data Mining library guide, including some helpful videos, to which I could direct my reader. I also directed him to our Hospital's Research and Development team (R&D. I found him some useful books on our library catalogue Handbook, and some articles that dealt with data mining his own specialist healthcare area.

  In Cambridge researchers use Symplectic Elements to manage their research data. There are also journals such as 'Medical Informatics' or the 'Journal of Biomedical Informatics' that deal with this area. Before a library reader can undertake embarking on a career in Clinical Informatics they need to understand the basic concepts that all library users need to know, so I told him about the database training that we offer and explained that he would need this to understand, for instance, how to use MESH subject headings. We looked and discovered that there was a MESH subject heading for Data mining. https://www.ncbi.nlm.nih.gov/mesh/?term=data+mining and one for Medical informatics a MESH term for Medical Informatics https://www.ncbi.nlm.nih.gov/mesh/?term=medical+informatics. In Cambridge the The Cambridge Big Data Strategic Research Initiative brings together researchers from across the University to address challenges presented by our access to unprecedented volumes of data, and a Data Mining symposium was held recently to address these issues.

Books on the subject seem to be The Cambridge Text Mining Handbook  also Data mining : practical machine learning tools and techniques / Ian H. Witten, Eibe Frank [and] Mark A. Hall.