On Collective Intelligence

The idea of collective unintelligence is examined in this paper to highlight some of the conceptual and practical problems faced in modeling groups. Examples drawn from international crises and economics provide illustrative problems of collective failures to act in intelligent ways, despite the inputs and efforts of many skilled and intelligent parties. Choices made of “appropriate” perceptions, analysis and evaluations are examined along with how these might be combined. A simple vector representation illustrates some of the issues and creative possibilities in multi-party actions. Revealed as manifest (un-)intelligence are the resolutions of various problems and potentials that arise in dealing with the “each and all” of a group (wherein items are necessarily non-parallel and of unequal valency). Such issues challenge those seeking to model collective intelligence, but much may be learned.

In order to be successful in terms of market share, sales, and profit, companies from different industries are detecting the innovative power of the customer network. Handing over more and more elements of the innovation process to the customer is accompanied by a loss of control of the innovating company thus creating quality uncertainty concerning the innovation process. According to New Institutional Economics these uncertainties can be overcome by building actor reputation within the web-based innovation network. Based on a short overview of the different stages of innovation process organization we will show how relevant actor reputation is for innovation networks. We develop an explanatory model of reputation building based on sociological theories of role modelling, interaction and communication and offer first considerations how the model can be tested. We conclude with a summary and an outlook on further research.

The rise of the so called Web 2.0 changed many classical business models considerably. New or changed business models systematically integrate the customer into the process of value-adding. The customer is not only consumer of products and services. He is rather directly or indirectly part of the production process. In the definitions of Web 2.0 this phenomenon is called collective intelligence. Roughly, in this context collective intelligence can be explained as a general term for user participation and the resulting added value. Examples like the T-Shirt retailer “Threadless” or the open innovation marketplace “InnoCentive” show the potential of these business models. Conventional methods and approaches for the visual representation of business models do not consider this new circumstance. The existing methods and approaches are inadequate, because they do not represent the special features of this kind of collective intelligence. This paper describes what web-based collective intelligence is, to get a common understanding of it and to have a definition for further work. Furthermore, an approach for the visual representation of business models using collective intelligence that represents these special features is presented.

Both, Open Innovation and Web 2.0, are concepts used in commerce in order to support the collaboration of different people and the emergence of new ideas. The approaches can be adapted to science, thus offering new opportunities for research and education. If necessary requirements are satisfied, Open Science 2.0 facilitates e.g. the public development of scientific papers and the conduct of public seminars, both harnessing collective intelligence. This way, it is not only possible to improve the individual outcomes, but also to encourage the exchange between theory and practice.

Social networks play an increasingly important role in knowledge management, information retrieval, and collaboration. In order to leverage the full potential of social networks, social networks need to be supported through technical systems. Within this paper, we introduce such a technical system. It is called AcaSoNet. It is a system for identifying and managing social networks of researchers. In particular, AcaSoNet employs a combination of techniques to extract co-author relationships between researchers and to detect groups of persons with similar interest. Past systems have used either search engines to extract information about social networks from the Web (Web mining) or have required people’s effort to enter their relationships to others into the system (as being done by most social network services). AcaSoNet, instead, uses a combination of these two types, thereby achieving data reliability and scalability. It extracts and collects data of researchers from the Web but allows researchers to modify the data. In the current version, our system can identify the social network based on publication lists and evaluate the publication activities of users within an academic community.

For today’s customers there is a need for integrated products and service bundles, e.g. a trip with flights, accommodation, local activities, home care during absence etc. To define such bundles, the collective intelligence of customers can be used. Thus, more customer-oriented products (bundles) can be created and the quality of bundles can be increased through the participation of many customers. By letting the collective in the form of customers participate, a solution is being created, which is better than any solution defined by a single company or any individual of the collective. To achieve this, swarm intelligence, as it can be found with social insects, is being transferred into an innovative digital platform. This platform enables customers to work together as a swarm and provides the means for customers to collectively design bundles and learn from each other.

Web2.0 users can create new services by combining existing Web2.0 services that offer open programming interfaces. This system of service composition forms a network, which we call the Web2.0 service network. A node of the Web2.0 service network represents a service. A link between two nodes exists, if another Web2.0 service (i.e. mashup) uses the linked services. The Web2.0 service network can be understood as an innovation system that creates value through the composition of services, representing the collective intelligence of users. Within this paper, we analyze the openness of the Web2.0 service network. Openness, which is an indicator for the innovation potential of a network, is measured using the Enhanced-EIS-Indexes. These indexes are based on Krackhardt and Stern’s EI-Index. The analysis results of the indexes show that the Web2.0 service network is not as open as the evolutionary analysis of the Web2.0 service network suggested. The slight closeness of the Web2.0 service network has been identified by the Agent Behavior Index EIS _a , which highlighted that relatively more links are created within subgroups than between subgroups. It indicates that factors such as service ownership and type of service have an impact on innovation within the network.

The socio-cognitive approach to teamwork has gained a lot of attention recently. Especially the concept of transactive memory (i.e., knowledge about each other’s knowledge) has been fruitfully applied to the team level. First, we extend the concept of transactive memory by considering a wider range of interpersonal aspects (e.g., personal traits, external relations, background knowledge). Second, we delineate practical approaches to develop transactive memory quickly. We distinguish between two training sequences: knowledge disclosure and knowledge updating. Whereas cross-training is an appropriate training approach at the beginning of teamwork, we refer to the after action review as an effective tool to update knowledge about each other in ongoing teamwork activities. Finally, open questions are discussed.

Knowledge representation is a key task of both computing science and programming practice. Suffice it to say that any program is a knowledge representation of a certain problem solution. However till now there are no means for the representation of application problems’ decision methods, for the representation of the environments making these problems, and for the representation of the communications between different knowledge environments. Today’s evolution of IT requires such knowledge representation tools. This paper proposes a knowledge representation language that allows a system of knowledge to be represented in a comfortable way for wide range of users and for automatic and semi-automatic problem solving in a suitable form.

This paper introduces a novel method to analyze the content of communication in social networks. Content clustering methods are used to extract a taxonomy of concepts from each analyzed communication archive. Those taxonomies are hierarchical categorizations of the concepts discussed in the analyzed communication archives. Concepts are based on terms extracted from the communication’s content. The resulting taxonomy provides insights into the communication not possible through conventional social network analysis.

Forecasting the demand for new products is becoming increasingly difficult in many markets. A new method to decrease the flop rate of new products is the idea to integrate customers deeply into the innovation process. This method of integrating the commitment of users to screen, evaluate and score new designs as a powerful mechanism to reduce flops of new products. The process starts when an idea for a product is posted on a dedicated web site by either a (potential) customer or just the designer of a product. Second, reactions and evaluations of other consumers towards the posted idea are encouraged in form of internet forums and opinion polls. Based on the results of this process, the manufacturer investigates the possibility of commercialization of the most popular designs. Is this evaluation positive, the company decides about a minimum amount of purchasers necessary to produce the item for a given sales price, covering its initial development and manufacturing costs (and the desired margin). The new product idea is then presented to the customer community, and interested customers are invited to express their commitment to this idea by voting for the design or even placing an order. Accordingly, only if the number of interested purchasers exceeds the minimum necessary lot size, investments in final product development are made, merchandising is settled and sales are commenced.