Open Data User Needs: A Preliminary Synthesis Auriol Degbelo degbelo@uni-muenster.de Institute for Geoinformatics, University of Münster Münster, Germany ABSTRACT Open city data is critical for smart cities and is becoming increas- ingly available, thanks to open government initiatives. Yet, we still know little about the needs of users concerning open data reuse. Taking fragmented and scattered issues reported by users in sev- eral works as a starting point, this article formulates user needs statements for research at the intersection of the topics of web, city, and data reuse. The 27 user needs statements proposed can inform the design and evaluation of tools that facilitate open data reuse. CCS CONCEPTS · Human-centered computing; · Information systems → Web searching and information discovery; Web applications; KEYWORDS Smart city, open government data, dataset search, user needs state- ments, human-computer interaction theory, web intelligence ACM Reference Format: Auriol Degbelo. 2020. Open Data User Needs: A Preliminary Synthesis. In Companion Proceedings of the Web Conference 2020 (WWW ’20 Companion), April 20ś24, 2020, Taipei, Taiwan. ACM, New York, NY, USA, 6 pages. https: //doi.org/10.1145/3366424.3386586 1 INTRODUCTION A smart city relies on advanced data processing, with the goals of making city governance more efcient, citizens happier, businesses more prosperous and the environment more sustainable (see [43]). Open data has been recognized as a critical element of smart cities in e.g. [12, 15, 27]. Besides, there are several approaches to modeling smart cities, but people is a dimension that is part of most mod- els (see [3]). Advancing research on smart cities thus necessitates theoretical foundations on user needs regarding open data usage. Yet, as mentioned in [45, 46], there is a dearth of research on the needs of users of open government data. Data publishers do not do that research (see [45]), nor has the scientifc community actively endeavored to close that gap. This discussion paper presents ideas of how these needs can be identifed and provides a preliminary taxonomy of these needs for open data usage. The size and complexity of interlinked web pages were the key factors behind Web Intelligence (WI) as a research feld [42], about 20 years ago. Nowadays, open datasets form an important part of the Web and add to its size and complexity. Ergo, the scope of WI should This paper is published under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license. Authors reserve their rights to disseminate the work on their personal and corporate Web sites with the appropriate attribution. WWW ’20 Companion, April 20ś24, 2020, Taipei, Taiwan © 2020 IW3C2 (International World Wide Web Conference Committee), published under Creative Commons CC-BY 4.0 License. ACM ISBN 978-1-4503-7024-0/20/04. https://doi.org/10.1145/3366424.3386586 be broadened to include ensuing issues of the increasing availability of datasets. Despite an increasing number of tools available (and user studies conducted to evaluate them), insight about what users actually need is less evident to glean. At best, these needs may be scattered across the literature. As previous work [34] reminded, developing a łhandbook of human needsž is one of the current grand challenges for research in human-computer interaction. There are at least two reasons why addressing user needs more directly would be benefcial to smart city and WI research. First, tools have a very short half-life, user needs last much longer (some needs may even not have a half-life at all). Second, addressing user needs is undoubtedly an important aspect of tool performance. Pre- vious work [37] reported that user-centered design (UCD) methods are generally considered to improve product usefulness and usabil- ity. There is a spectrum of ways in which users are involved in UCD, but the bottom line of all methods is that users are involved in some way (see [1]). Thus, extracting user needs explicitly may help better predict and explain why some tools perform better than others. In sum, explicit identifcation of user needs is a much needed theoretical contribution to smart city and WI research. As ‘need’ is a key term throughout the work, a defnition is in order at this point. From the point of view of human-computer interaction, a value-based defnition of ‘need’ is not helpful, as Miller [26] commented. It leads indeed to a regress that dead-ends only with the agreement that all that humans really need are food, water, and a place to sleep. Instead, a more pragmatic approach is needed. This pragmatic view takes the stance that ‘need is what is demanded’ [26]; it is the intent behind a user query (see e.g. [7]). This view allows the formulation of ‘user need statements’ [17], that is, statements following the pattern: ‘[A user] needs [need] in order to accomplish [goal]’. The objective of this work is to advance the formulation of such statements in the context of open government data (OGD) usage. With the assumption that problems point at needs (i.e. unsatisfed demands), the article synthesizes what we can learn about the needs of OGD users, based on a critical review of empirically-derived OGD usage issues in the literature. The main contribution of this paper is a synthesis of current fndings from the literature into 27 user needs statements. Since there is a strong conceptual overlap between OGD and city data (see [14]), and most OGD is available on the Web, these user needs statements are relevant to research at the intersection of the topics of web, city, and data reuse. A discussion of their usefulness for web intelligence research is provided towards the end of the article. 2 BACKGROUND This section briefy reviews what we know, so far, about usage and users of open (government) data. The works mentioned here have, for the most part, derived their insights through empirical studies in some form or another. Topics covered are: the impact of OGD,