Tracking Behavioral Construct Use through Citations: A Relation Extraction Approach

During the past few decades, social and behavioral sciences experienced a proliferation of vibrant research communities, leading to a rapid accumulation of theories, articles, and constructs (Lee et al. 2004; Straker 2008). Many efforts have been dedicated to exploration and integration of this large network of behavioral research data. For example, the inter-nomological network, a construct search engine, improves retrieval of relevant constructs in the behavioral sciences. It allows researchers to search and create their own hierarchies of over 70,000 constructs manually collected by a large team of research assistants (Larsen and Hovorka 2012; Larsen et al. 2010). Furthermore, by using state-of-art information extraction (IE) techniques, Li and Larsen (2011; 2012) developed the automated nomological network extraction (ANNE) IE tool which automatically extracts hypotheses, constructs, and nomological networks from behavioral articles. This tool has the potential to significantly ease the time and labor limitations of INN. Additionally, by visualizing a large-scale nomological net which connected a number of quantitative theories published in two leading journals in the information systems field, Hovorka and Larsen (2012) conducted a meta-theoretic analysis to locate and identify theory domains, which demonstrates a potential pathway to integration and extension of theories within and across disciplines.