Information behavior and practices research informing information systems design

Abstract

Information behavior and practices (IBP) research has been repeatedly criticized for having little impact on information systems development (ISD). Claiming that there is a complete disconnect would be an exaggeration but it is apparent that it is not always easy to translate findings of IBP research to workable design recommendations. Based on a reading of earlier literature and a closer investigation of three illustrative example contexts, this article underlines that the value of IBP research for ISD lies in its capability to inform ISD of the variety of ways people deal with information beyond individual systems, their own wants and designers' assumptions. Moreover, it highlights that the implications of information systems go beyond their primary users. Instead of overemphasizing the contextuality of findings, a part of IBP research would benefit from an increased focus on explicating its epistemological extents and limits and identifying, which findings are transferable, what distinguishes specific contexts, what are their defining constraints and priorities, and what aspects of their uniqueness are assumptions and simple clichés.

Explainable Artificial Intelligence (XAI)

“…10 Critical algorithm studies position machine learning as socio-techno-informational systems. 11 As such, a definition of XAI must encompass not just the techniques, as important and necessary as they are, but also the context within which XAI operates.…”

To pre-filter, or not to pre-filter, that is the query: A multi-campus big data study

“…This study highlights the importance of expanding students' understanding of the larger breadth of search possibilities as students seem unaware of individual search features within specific databases. Huvilla et al (Huvilla et al, 2022) investigated the disconnect between information behavior and practices literature and information system development, and a "need for a better and more holistic understanding of user needs and perspectives" (p. 1043). These authors mention the need for findings to be contextualized as "explanations and recommendations have the most value when they are connected to specific services and contexts of use -even if there are many general traits in how people interact with information" (p. 1052).…”

Can Information Behaviour Inform Machine Learning?

“…Information behaviour research has been adopted by other disciplines including computer science (Huvila et al, 2021;Makri, 2020;Wilson, 2018Wilson, , 2020b, but it has not influenced the subfield of machine learning. A search for papers discussing topics such as "information behaviour", "information-seeking behaviour", "information-seeking", or "information needs" in the Digital Library of the Association for Computing Machinery (dl.acm.org and the arXiv pre-print repository (arXiv.org) returned few correlations with machine learning research and none with references to information behaviour research.…”

“…It is explicit in machine learning and implicated in information behaviour (Case and Given, 2016;Wilson, 2016). As a result, these fields share a "socio-techno-informational" nexus (Eriksson-Backa et al, 2021;Huvila et al, 2021) that offers the basis for exchange.…”