2014 Ebola’s outbreak has been widely discussed in the scholarly publishing industry and it is illustrative of the current debates about the Open Access to academic knowledge.
Reports dating back from 1980’s warned about the health risk of Ebola in the region. If West African healthcare professionnals had access to that knowledge, many deaths could have been prevented .
To achieve such a result, Open access combined with information retrieval tools were necessary.
We have explored several ways to illustrate how the information contained in Open Access Full texts could have been used to track and follow weak signals, those that are difficult to identify but that could lead an attentive observer to react earlier.
On our corpus of Open Article that deals with Ebola, we show here how the number of Full-texts and the number of abstracts evolve within time.
Before the outbreak, we can see a rise in the number of Full-text that mention these three countries. Not in the abstracts.
This tag cloud is constitued by articles keywords that are only included in the Full Texts and not in the abstracts.
It shows some co-occuring decease and environmental factors that are not central in the articles but are nevertheless often mentionned. Weak signal to investigate…
A simple classification pipeline using either abstracts or fulltexts as the entry data was set-up.
Full-text are shown to give better results at simple classification tasks such as keyword retrieval.
This illustrates how Open Access Fulltexts can be leveraged.
 Dahn, B., Mussah, V., & Nutt, C. (2015, April 7). Yes, We Were Warned About Ebola. The New York Times.
 Crotty, D. (2019). Access Alone Isn’t Enough: Revisiting Calls for Discovery, Infrastructure, Technology, and Training. Retrieved February 10, 2019