Mapping computer science research in Africa: using academic networking sites for assessing research activity

Abstract

Much of the research conducted in Africa that can address social and economic challenges on the continent remains invisible when using standard bibliometric databases because these fail to capture research outputs that are in local languages or published in local journals. We turn to academic networking sites as data sources to uncover research outputs of sub-Saharan African countries to provide a fuller picture of the research activity in the region. We map computer science research productivity and collaboration in sub-Saharan Africa using traditional bibliometric databases (Scopus and Web of Science) and an alternative data source: Academia.edu. This alternative data source reveals additional research to that found in the bibliometric databases and provides insights into publication and collaboration patterns. We found 761 unique records not indexed in the bibliometric databases, constituting 6% of all the computer science research output for the region. These records were mostly articles in local and regional journals. We also found a higher percentage of regional collaborations. The top five countries for computer science research outputs in sub-Saharan Africa (excluding South Africa) using both traditional bibliometric and alternative data sources were Nigeria, Kenya, Sudan, Ghana, and Mauritius. We conclude that utilizing only traditional bibliometric databases for mapping computer science research activity in sub-Saharan Africa provides an incomplete picture of the research happening in this region. Using academic networking sites as data sources, despite being labor intensive and having other limitations, can help to fill in some of the gaps.