I am interested in doing research in data mining, specifically in the areas of health in developing countries. As you are aware, data is mainly stored in a manual format as in either file/paper-based. In addition, medical records are very protected. How best can I achieve my goal?
My two cents on the topic:
1. My first advice would be to focus on a specific research area, as "health" is a very broad field. Find an application area of your interest, e.g. breast-cancer, depression, cardio-vascular disease, etc.
2. Next, if you are interested in collecting your own dataset you should form a cooperation with a relevant clinic. This is highly co-dependent with the first point, as the focus could also depend on the availability of relevant partners.
3. Very important point is the formal procedures which need to be followed in order to LEGALLY obtain access to such data. Top tier journals require permits to be explicitly stated in articles that make use of medical datasets in order to be considered for publication. Procedures differ significantly based on the country. For example, for the European Union legislation has become more strict with the GDPR law which became effective since May 2018. Thus, you need to search for the relevant legislation of the country the research is being carried.
Data collection is a long-term goal, designing the collection methods, running pilots, obtaining permits, having people engaging in participating, etc. Your expertise area on your profile is computer science, thus I suppose you are mostly interested in the processing of the data and the development of the relevant algorithms, that the actual data. If I am right in my assumption you could also investigate the option of employing a free dataset in an area which interests you, rather than collecting the data on your own.
As a first step I'd recommend to explore some of the accessible datasources.
For example, Global Health Data Exchange (GHDx) ontains surveys, censuses, vital statistics, and other health-related data from many countries (http://ghdx.healthdata.org/countries). The Global Health Observatory data repository supported by the World Health Organization contains health-related statistics for more than 1000 indicators for its 194 Member States aggregated by topic, country, etc. (http://apps.who.int/gho/data/node.home).
Even if you don't find all the information you need in them, they might be a good starting point.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining