Reviewed by
Mario Paolucci
Institute of Cognitive Science and Technology, CNR
This book with a long and intricate title collects contributions on the subject of several cognitive artefacts in the context of online information. The application focus is on knowledge markets, defined as a "secondary and very active informal marketplace of ideas and knowledge" (preface and p.4). Here, secondary refers to authority; those markets are not regulated by institutional roles, but, at least to some extent, they let those roles emerge internally. The main example is that of Wikipedia, but also the other classic social networks are considered.The book appeared thanks to the support of KredibleNet, an NSF-financed project, and reports in the preface and in the last chapter on the opening KredibleNet workshop held at Purdue University in April 2013. It is thus a report from the start of the project, and not a final result. This approach is meanwhile a common procedure in project development, perhaps because of the need to capitalize on the increasing effort required in project preparation.
As such, the book does not report on new findings, but contains for the larger part research problems and suggested recipes to tackle them, in addition to a report on the state-of-the-art of research on information collection and exchange. Some subjects would benefit from more precision in definitions and in conceptual boundaries. For example, the market analogy in the title, while widely applied, is misleading. Places where (copiable) information production and consumption are exchanged freely are not well described by a market analogy - indeed, the authors use at times a different formulation as "knowledge building spaces".
The reader will find out that the initial workshop was organised around eight questions (p. vi). Some of the questions concerned with a description of current (and a speculation on future) ideas about knowledge and social media, network shape, and central individuals. I select the two most interesting questions:
5. How do functional roles, reputation, and authority emerge on social media knowledge generation projects and how can they be operationalised, measured and explained?
6. How do trust and knowledge credibility connect to specific functional roles and authority structures?
The book, thus, touches also on topics of interest for a cognitive approach to knowledge, and aims to touch subjects like reputation, roles, and even authority - although focusing only on the distributed form of the above, without addressing their centralised aspect as one would with an institutional approach.
In fact, questions 5 and 6 are especially suitable to be addressed by simulation modelling. JASSS readers will find several papers touching on the subject matter - a brief review points to the kene model of innovation (Gilbert, Pyka and Ahrweiler 2001), that can be adapted to use kenes as a model to represent knowledge; a Wikipedia simulator developed by Klaus Troitzsch in the EMIL project (Xenitidou et al., 2014), and a study of web credibility carried out in a collaboration led by the Polish-Japanese Institute of Information Technology (Wierzbicki et al. 2014).
The book structure as presented in its introduction is organised in five parts: the first and the last chapter illustrate the KredibleNet approach. After the project has been introduced, the focus has first been set on expertise and reputation. Homage is paid to labels one expects to find there – for example, Big Data and interdisciplinary approach. In the remainder of this review, I will browse through these parts, highlighting some contributions that appeal to my personal interests and that I find useful for the JASSS audience. The preface of the book also contains a useful summary.
The introduction section consists of two chapters on which the KredibleNet project is sketched out with a focus on functional role analysis performed through the study of network structure. With the example of Wikipedia in mind, the purpose is to collect measures of characteristics of article modifications – called edits – such as size, frequency, place, and interaction. These measures are employed to derive information about the role of the edit‘s author. In order to recognise "functional roles of editorial, administrative, or political consequence" (p.17) the network approach might be not enough, and an important part could be played by an explicit modeling of the cognitive structure of agents.
A novel tool, the measurement of social entropy, is also presented in this first part (pp. 23). It is calculated as the sum of agents’ contributions, weighed by the logarithm of the same. As with information entropy, it is greater when all contributions are of the same size. In other words, it measures the diversity in some quantity measure of posts. It is a bit disappointing that any reference to social entropy disappears after the first chapter, but I attribute this to the nature of the book as a project, not a result.
Still in the first part, Ben Shneiderman introduces concepts such as trust and credibility. The lack of references to cognitive theories of reputation is disappointing; in addition, the publisher didn't check the citation style, mismatching number-and author-citation. Matei and others conclude the book with a report on the first KredibleNet workshop, touching on the same points as the introduction already did, which is why I mention it here.
The second part of the book focuses on methods which explore trust and credibility, and consists of four chapters. I recommend Chapter 4 by Brian Keegan, containing case studies of editor activity in reaction to breaking news. The evidence presented as individual editor’s trajectories is intriguing but anecdotal; no attempt is made to evaluate its representativeness. It is argued (but not demonstrated) that the pattern of editors’ roles emerging from interactions during incidents "reflect more of the interactionist dimension of disaster response teams rather than the regeneration of collaborative infrastructures found in ER teams." (p. 74).
The representation chosen for the edits, that is, the idea of tracking the path of an editor through the pages she has modified, changes name (pp. 61 and following) between being called "user trajectory" and "sociotechnical trajectory". The second seems only justified by the fancy of the name just for confusing matters (there's nothing social in it. Technical? Perhaps).
The third and fourth part of the book present tools for increasing trust and transparency, and novel research directions. Of some note, Bernie Hogan argues that the sorting approach used from all the information retrieval tools is comparable to an ideology, and that it prevents the onset the of different approaches based on reputation and credibility. Corruption in organisations, a hot topic, is addressed by Howard Welser in a useful review, starting with the unavoidability of it (mentioning classic corruption cases, "we should remember that they behaved exactly as we should predict, given the opportunities presented by their organizational position," p. 122). The review that mentions several online tools and tries to compare distributed social control with the more traditional kind of control found in organizations, concludes with a call to arms for organisational design: "We need a global distributed effort, a space race for high-fidelity organizational design." (p. 142).
The paper by Mei Kobayashi is a compact overview of cultural differences in online space, employing cognitive artefacts as trust and reputation, although the author seems to consider them equivalent. The topic of authority is also approached.
To sum up, the book offers a huge variety of viewpoints on current approaches to online information exchange as facilitated by cognitive artefacts. It does not contain new research, but it constitutes a suitable starting point for the researcher with an interest in the field.
Finally, it is worth mentioning that the book appears in the "computational social sciences" series of Springer, which is of certain interest for the JASSS reader. The list of titles published there can be found here: http://www.springer.com/series/11784?detailsPage=titles.
ReferencesWIERZBICKI A., Adamska P., Abramczuk K., Papaioannou T., Aberer K. and Rejmund E. (2014). 'Studying Web Content Credibility by Social Simulation'. In: Journal of Artificial Societies and Social Simulation 17 (3) 6 https://www.jasss.org/17/3/6.html.
XENITIDOU M. A., Emde R. O., Villard J.E, Lotzmann U., Troitzsch K. G. (2014). Demonstrating the theory: The case of Wikipedia. In: Minding norms. Mechanisms and dynamics of social order in agent societies. Oxford University Press, pp.127-152.
Return to Contents
of this issue