A Novel Approach to Find Author's Research Areas of Interests Using Graph Database

2. Related Works

The subject area of the published papers of an author have a strong role in determining his coauthor networks. An author usually collaborates with authors belong to his research area of interest (Zhang et al., 2018). Finding experts in a specific field is a key research area. Document features like author names can be used to identify subject experts, if the subject area of the paper is known which is detailed in (Zhu et al., 2009). Clique percolation method is used to find research areas of scientists in (Imran et al., 2018). First overlapping communities were identified by merging maximal cliques and then subject area of each community is determined using frequent words and bi-gram approach from the titles and journal names of authors. Identification of Subject experts using PubMed articles dataset is described in (Singh et al., 2013). The authors used terms of MeSH, medical subject headings contained in each published articles of the author to determine the subject expertise of an author. Finding subject experts at a specific time is explained in (Daud 2012), where authors used a time topic modeling approach. Article textual contents, author’s details and time information were used simultaneously to find dynamic research interests utilizing Latent Dirichlet Allocation (LDA). Authors follow an intelligent approach using knowledge organization systems in (Pei-van et al., 2019) to find academic subject experts by using a combination of Latent Dirichlet Allocation (LDA) and TF-IDF approach. Dynamic creation of author profile is explained in (Vivacqua et al., 2009). The authors used information given by the target authors and their observed behaviour to infer profiles.