Knowledge Graph for Academic Libraries: A User-Centered Design Approach

Gardasevic, Stanislava; Lamba, Manika

Permalink

https://hdl.handle.net/2142/123231 Copy

Description

Title

Knowledge Graph for Academic Libraries: A User-Centered Design Approach

Author(s)

• Gardasevic, Stanislava
• Lamba, Manika

Issue Date

2024-06-30

Keyword(s)

knowledge graph

Abstract

This poster presents the user-centered design research that engaged interdisciplinary Ph.D. students (the end-users) in developing the knowledge graph model and dataset using the Neo4J graph database and Bloom graph exploration tool, helping them progress toward completing the degree. The graph elicited from this research shows collaboration and knowledge networks in a community and involves automated and manual indexing of publications and dissertations. The methodology used in this research is a combination of semi-structured interviews, participatory workshops, and questionnaires with over 20 PhD students. The proposed technology is currently a prototype that was tested with the user population. It has proven helpful in decision-making, especially when identifying and choosing collaborators, supervisors/dissertation committee members, and appropriate classes based on the feedback and experiences of other students. Users found this approach helpful for quick access to data aggregated from various websites, metadata of publications and dissertations, and crowdsourced data from other students. End users can explore current co-authorships, dissertation-mentorship, and research/lab collaboration networks, among other networks interwoven in this complex graph. Furthermore, they could discover potential collaborators among those who research in the same domain of application or research area and read about insights and experiences of other students in their program. With this poster, we will present in a visually appealing way (i) the user-centered design methods that led to modeling and populating this graph, (ii) the knowledge graph model that can be applied to interdisciplinary and other domains of science, (iii) showcase selected helpful queries, and (iv) study results showing how the graph is impactful for interdisciplinary researchers and students. This poster proposes a promising avenue for academic librarians, who can contribute to building such graphs to serve mono and interdisciplinary Ph.D. programs, students, and other interested stakeholders (e.g. faculty, program coordinators, administrators).

Type of Resource

• still image
• Poster

Language

eng

Owning Collections