User-driven efforts in creating knowledge graph information system

Abstract

Purpose This paper presents the results of a qualitative study that involved students of an interdisciplinary PhD program. The study objective was to gather requirements to create a knowledge graph information system. The purpose of this study was to determine information-seeking practices and information needs of this community, to inform the functionalities of a proposed system, intended to help students with relevant resource discovery and decision-making. Design/methodology/approach The study design included semi-structured interviews with eight members of the community, followed by a website usability study with the same student participants. Findings Two main information-seeking styles are recognized and reported through user personas of international and domestic (USA) students. The findings show that the useful information resides within the community and not so much on the program website. Students rely on peer communication, although they report lack of opportunities to connect. Students’ information needs and information seeking are dependent on their progress through the program, as well as their motivation and the projected timeline. Practical implications Considering the current information needs and practices, a knowledge graph hosting both information on social networks and the knowledge produced by the activities of the community members would be useful. By recording data on their activities (for example, collaboration with professors and coursework), students would reveal further useful system functionalities and facilitate transfer of tacit knowledge. Originality/value Aside from the practical value of this research that is directly influencing the design of a system, it contributes to the body of knowledge on interdisciplinary PhD programs.

Community Design of a Knowledge Graph to Support Interdisciplinary PhD Students

“…This knowledge is often referred to as experiential or tacit knowledge, here operationalized as the values and quality of the resource of interest, as perceived by students [14]. Resources that help students discover appropriate faculty to collaborate with are especially critical for this population [3,4,10], as well as information about research topics of interest [3,7,11].…”

“…Faculty profile pages can be a lightweight data source of potential connections between researchers and students from diverse fields, and community stakeholders [19], and can provide a way to identify one's place in the interdisciplinary community. However, in the case of the program studied here, students reported that faculty profiles are often not up to date, and too numerous to inspect thoroughly [10]. Also, expert finding applications in the scholarly domain are usually based on recommending acclaimed researchers based on their publication history [i.e.…”

“…Consequently, we chose a multilayer/multiplex network as the medium to address the raised issues and create a knowledge graph combining information on social and knowledge traces in a community. The aim of building this graph is to facilitate explicit and tacit information discovery and informed decision-making related to the factors identified as meaningful for this population: i) establishing collaboration with faculty [3,4,10], ii) information and resources available about research topic of interest [3,7,11], and iii) interaction with peers, including community building [3,4]; while creating a testbed for exploring multiplex network visualization and navigation options [12].…”

“…As noted by Hogan et al [13], the definition of a knowledge graph remains contentious, but in short, it can be described as "a graph of data intended to accumulate and convey knowledge of the real world, whose nodes represent entities of interest and whose edges represent relations between these entities". Before building the graph, we created the database model to support tacit knowledge exchange-such as first-hand impressions of other students-that have been shown to be of crucial importance for this population [7,9,10].…”

“…The While the flexible, decentralized structure of the program provides more possible research avenues for PhD students, it also requires students to find and navigate their own path through program requirements [4,10]. This is especially acute since the selection of a dissertation chair and other faculty mentors is the crucial factor contributing to PhD student retention and satisfaction in general [3,31].…”