Thematical and Impact Differences Between National and International Collaboration on Artificial Intelligence Research

In this paper, an analysis to discover the impact differences in terms of citations between three levels of research collaboration is carried out. Thus, the objective, further than discover what kind of collaboration is more rewarding in terms of impact, is to determine if researchers cover different themes when they collaborate with people from other countries or universities within the same country, and if those themes get more citations. To this end, 12610 documents belonging to the "Computer Science & Artificial Intelligence" Web of Science subject category, and published in 2015 were analyzed. The whole corpus was divided into three different datasets, according to the defined collaboration levels (local, national and international). Results indicate that papers by authors from different countries receive more citations in average, but also, there exist some differences with respect to the research themes between the datasets.