Twitter & Academic Healthcare Content, More Noise than Signal?



Over an eight week period, tweets containing hashtags specific to Men’s and Women’s health and lifestyle issues were compiled using freely available open source tools. The content of mined tweets was further analysed to identify user profiles, additional hashtags and urls. As twitter frequently uses shortened urls, these were resolved to reveal the original syntax. A search optimisation engine (SEO) spider was then used to gather key onsite elements of urls such as page title, Meta data description and keywords.

The raw data was then analysed and contextualised output taxonomies for each of the subject areas were developed. This helped the researchers identify the following:

1. Identification of Top Tweeters in specific subject areas to determine who the influencers in the community are. A review of Twitter Profiles helped in the development of archetypes for each area.
- Advocate
- Survivor/ Previvor
- Academic
- Professional Affiliation
- Health Professional
- Medical Researcher
- News/Media

2. The statistical analysis of raw data (all data in a tweet and all tweets from the Top Tweeters identified the most frequently used hashtags. The top 50 hashtags constitute 80% of all hashtags for each area. Associated hashtags provide an insight into contextualised words used by users, even though there is a diverse sets of hashtags, there frequency of the most popular is produced a textual pattern.

3. The next of extraction and analysis was from the top hashtags identified linkable URL’s. The URL content extraction was linked to the subject specific taxonomy. This analysis has provided insight
- Into the ways URL’s are categorised
- Relationships between the most frequently tweeted link and the primary resource for tweeters.
- The relationship between the most frequently tweeted links from specific domains demonstrate patterns related to frequency of use of academic content (journals, conference papers, peer reviewed charity information, and related blogs.

A review of Twitter content has applications in multiple settings including medial education and healthcare. For the Medical Education Environment it provides a reference resource which identifies Top Tweeters, top URL’s and related content as well as identification of primary content generators, Individual (affiliated or not), Official Body (Medical Research/ Academic) or Media/Press.

While other tweet aggregation services exist, our method allows for the identification and categorisation of urls shared via twitter and the exploration of who and how this information is shared.


All talks:


If you require information about any aspect of the conference please contact: Fi Coyle: fiona.coyle@heanet.ie / +353 1 6609040