Attribute interactions are the irreducible dependencies between attributes. Interactions underlie feature relevance and selection, the structure of joint probability and classification models: if and only if the attributes interact, they should be connected. While the issue of 2-way interactions, especially of those between an attribute and the label, has already been addressed, we introduce an operational definition of a generalized n-way interaction by highlighting two models: the reductionistic part-to-whole approximation, where the model of the whole is reconstructed from models of the parts, and the holistic reference model, where the whole is modelled directly.

International co-authorship relations and university-industry-government (“Triple Helix”) relations have hitherto been studied separately. Using Japanese (ISI) publication data for the period 1981-2004, we were able to study both kinds of relations in a single design. In the Japanese file, 1,277,823 articles with at least one Japanese address were attributed to the three sectors, and we know additionally whether these papers were co-authored internationally. Using the mutual information in three and four dimensions, respectively, we show that the Japanese Triple-Helix system has continuously been eroded at the national level. However, since the middle of the 1990s, international co-authorship relations have contributed to a reduction of the uncertainty. In other words, the national publication system of Japan has developed a capacity to retain surplus value generated internationally. In a final section, we compare these results with an analysis based on similar data for Canada. A relative uncoupling of local university-industry relations because of international collaborations is indicated in both national systems.

Can the knowledge base of an economy be measured? In this study, we combine the perspective of regional economics on the interrelationships among technology, organization, and territory with the triple-helix model, and offer the mutual information in three dimensions as an indicator of the configuration. When this probabilistic entropy is negative, the configuration reduces the uncertainty that prevails at the systems level. Data about more than a million Dutch companies are used for testing the indicator. The data contain postal codes (geography), sector codes (proxy for technology), and firm sizes in terms of number of employees (proxy for organization). The configurations are mapped at three levels: national (NUTS-1), provincial (NUTS-2), and regional (NUTS-3). The levels are cross-tabled with the knowledge-intensive sectors and services. The results suggest that medium-tech sectors contribute to the knowledge base of an economy more than high-tech ones. Knowledgeintensive services have an uncoupling effect, but less so at the high-tech end of these services.

Abstract not found

Psychological and social systems provide us with a natural domain for the study of anticipations because these systems are based on and operate in terms of intentionality. Psychological systems can be expected to contain a model of themselves and their environments; social systems can be strongly anticipatory and therefore co-construct their environments, for example, in techno-economic (co-)evolutions. Using Dubois’s hyper-incursive and incursive formulations of the logistic equation, these two types of systems and their couplings can be simulated. In addition to their structural coupling, psychological and social systems are also coupled by providing meaning reflexively to each other’s meaning-processing. Luhmann’s distinctions among (1) interactions between intentions at the micro-level, (2) organization at the meso-level, and (3) self-organization of the fluxes of meaningful communication at the global level can be modeled and simulated using three hyper-incursive equations. The global level of self-organizing interactions among fluxes of communication is retained at the meso-level of organization. In a knowledge-based economy, these two levels of anticipatory structuration can be expected to propel each other at the supra-individual level.

Abstract not found

This study examines the longitudinal trend of systemness in networked research relations in South Korea using a triple-helix (TH) indicator of university-industry-government (UIG) relations. The data were harvested from the Science Citation Index (SCI) and its counterparts in the social sciences (SSCI) and the arts and humanities (A&HCI). The total number of Korean SCI publications has grown rapidly since 1965. However, the TH indicator shows that the network dynamics have varied considerably according to the research policies of the national government. The collaboration patterns, as measured by co-authorship relations in the SCI noticeably increased, with some variation, from the mid-1970s to the mid-1990s. However, interinstitutional collaboration in the first decade of the 21 st century was negatively influenced by the new national science and technology (S&T) research policies that evaluated domestic scientists and research groups based on their international publication numbers rather than on the level of cooperation among academic, private, and public domains. The results reveal that Korea has failed to boost its national research capacity by neglecting the network effects in science, technology, and industry.

2. The Triple Helix as a model of the knowledge-based economy......................................................... 3 3. Knowledge as a social coordination mechanism.................................................................................8 4. Neo-evolutionary dynamics in a Triple Helix of coordination mechanism...................................... 12

Abstract. The article presents the information-theoretic based feature information interaction, a measure that can describe complex feature dependencies in multivariate settings. According to the theoretical development, feature interactions are more accurate than current, bivariate dependence measures due to their stable and unambiguous definition. In experiments with artificial and real data we compare the empirical estimates of correlation, mutual information and 3-way feature interaction. We can conclude that feature interactions give a more detailed and accurate description of data structures that should be exploited for information fusion in multimedia problems. 1

Research article Learning transcriptional regulatory networks from high throughput gene expression data using continuous three-way mutual information