Identifying Diachronic Topic-Based Research Communities by Clustering Shared Research Trajectories

Abstract

Abstract. Communities of academic authors are usually identified by means of standard community detection algorithms, which exploit 'static' relations, such as co-authorship or citation networks. In contrast with these approaches, here we focus on diachronic topic-based communities -i.e., communities of people who appear to work on semantically related topics at the same time. These communities are interesting because their analysis allows us to make sense of the dynamics of the research world -e.g., migration of researchers from one topic to another, new communities being spawn by older ones, communities splitting, merging, ceasing to exist, etc. To this purpose, we are interested in developing clustering methods that are able to handle correctly the dynamic aspects of topic-based community formation, prioritizing the relationship between researchers who appear to follow the same research trajectories. We thus present a novel approach called Temporal Semantic Topic-Based Clustering (TST), which exploits a novel metric for clustering researchers according to their research trajectories, defined as distributions of semantic topics over time. The approach has been evaluated through an empirical study involving 25 experts from the Semantic Web and Human-Computer Interaction areas. The evaluation shows that TST exhibits a performance comparable to the one achieved by human experts.

Klink-2: Integrating Multiple Web Sources to Generate Semantic Topic Networks

“…For example, they can enhance semantically many information extraction techniques, such as trend detection [6] and community detection [7,10]. They also make it possible to improve search results and their presentation, e.g., by supporting semantic faceted search [20].…”

“…This indicates that two topics can be treated as equivalent for the purpose of exploring research data -e.g., "ontology mapping" and "ontology matching". Skos:broaderGeneric and relatedEquivalent are necessary to build a taxonomy of topics and to handle different labels for the same research areas, while contributesTo provides an additional relationship that can be used to assist the user in browsing research topics [5] and analyzing research data -e.g., for identifying topic-based research communities [10].…”

“…However, there is growing consensus that semantic technologies can help to overcome this problem by improving our ability to discover, query, explore, annotate and visualize research information on the web [4,5,6,7,8,9,10]. Nonetheless, we still face some important technical challenges before this vision can be realized.…”

“…For example, while it is easy to retrieve all the co-authors of Enrico Motta, it is much more difficult to identify all the papers of Enrico Motta which are relevant to research on the Semantic Web or one of its sub-areas. For this reason many popular systems for the exploration of research data, such as Google Scholar 8 , Microsoft Academic Search 9 and Scopus 10 , sidestep the challenge of identifying research topics and linking them to other relevant research entities, and simply use keywords as proxy. Unfortunately, this purely syntactic solution is unsatisfactory, as it fails i) to distinguish research topics from other keywords which can be used to annotate papers; ii) to deal with situations where multiple labels exist for the same research area; iii) to deal with the fact that a keyword may denote different topics depending on the context, and iv) to model and take advantage of the semantic relationships that hold between research areas, treating them instead as lists of unstructured keywords.…”

Rexplore: Unveiling the dynamics of scholarly data

“…A user can thus gain an immediate knowledge of the history, the main groups of authors, the collaborations and the organizations active in each research area. Technically, this is achieved by TST [4], an algorithm which identifies communities of researchers who appear to follow a similar research trajectory. For example, Figure 1 shows a graph view of the main Semantic Web research communities in which the user clicked on MIT to obtain additional details.…”