Artificial Intelligence Lab, Department of MIS, University of Arizona / Recommender systems are being widely applied in many application settings to suggest products, services, and information items to potential consumers. Collaborative filtering, the most successful recommendation approach, makes recommendations based on past transactions and feedback from consumers sharing similar interests. A major problem limiting the usefulness of collaborative filtering is the sparsity problem, which refers to a situation in which transactional or feedback data is sparse and insufficient to identify similarities in consumer interests. In this article, we propose to deal with this sparsity problem by applying an associative retrieval framework and related spreading activation algorithms to explore transitive associations among consumers through their past transactions and feedback. Such transitive associations are a valuable source of information to help infer consumer interests and can be explored to deal with the sparsity problem. To evaluate the effectiveness of our approach, we have conducted an experimental study using a data set from an online bookstore. We experimented with three spreading activation algorithms including a constrained Leaky Capacitor algorithm, a branch-and-bound serial symbolic search algorithm, and a Hopfield net parallel relaxation search algorithm. These algorithms were compared with several collaborative filtering approaches that do not consider the transitive associations: a simple graph search approach, two variations of the user-based approach, and an item-based approach. Our experimental results indicate that spreading activation-based approaches significantly outperformed the other collaborative filtering methods as measured by recommendation precision, recall, the F-measure, and the rank score.We also observed the over-activation effect of the spreading activation approach, that is, incorporating transitive associations with past transactional data that is not sparse may “dilute” the data used to infer user preferences and lead to degradation in recommendation performance.

Artificial Intelligence Lab, Department of MIS, University of Arizona / Spiders are the software agents that search engines use to collect content for their databases. We investigated algorithms to improve the performance of vertical search engine spiders. The investigation addressed three approaches: a breadth-first graph-traversal algorithm with no heuristics to refine the search process, a best-first traversal algorithm that used a hyperlink-analysis heuristic, and a spreading-activation algorithm based on modeling the Web as a neural network.

Artificial Intelligence Lab, Department of MIS, University of Arizona / Areas related to nanotechnology, or nanoscale science and engineering (NSSE), have experienced tremendous growth over the past few years. While there are a large variety of useful resources available on the Web, such information are usually distributed and difficult to locate, resulting in the problem of information overload. To address the problem, we developed the NanoPort system, an integrated Web portal aiming to provide a one-stop shopping service to satisfy the information needs of researchers and practitioners in the field of NSSE [1]. We believe that the approaches taken also can be applied to other domains.

The authors use an exploratory project involving Web resources related to Alzheimer’s Disease to explore ways in RDF metadata can more effectively translate the virtues of the traditional vertical file to a Web environment form using Semantic Web descriptive standards. In so doing, they argue against the separation of “bibliographic control” from the socially-embedded institutional practices of reference work, collection development, and the management of information ephemera. Libraries of the future will use specific Web technologies that lend themselves to sophisticated and rigorous knowledge structures, and link them with librarians’ skills in information harvesting and evaluation.

This paper describes an exploratory, multisite case study of the production of textual content for state agency Web sites. The qualitative field study explored internal agency Web staff characterizations of textual Web content and staff perceptions of factors affecting the production of content. Study results suggest that staff characterize content in terms of its format, its style age, its rate of change, its degree of change, its owner, and the degree to which it is sensitive. Staff described nine factors affecting content production including information intensity, public education mission, public inquiry burden, top-down directives, existing maintenance burden, review and approval process, resources, management interest and goals, and support from other program staff. A better understanding of how internal agency staff perceive and treat content is important because staff play a large role in determining what content is produced and what characteristics the content contains. The inclusion or exclusion of certain characteristics in content has important implications for information usability, costs, citizen participation in agency policymaking, government transparency, and public trust in government.

This study assesses one Web site, Veggie-mon.org. This Web site aims to effectively communciate health information to children, resulting in user learning and an intent to change health behavior. Fourth- through eighth-grade pupils were interviewed before and after perusing the Web site for up to 20 minutes, and then they participated in focus groups. A majority of participants learned what the Web site was about, had previous health knowledge reinforced, discovered new health information, and said it made a difference in their health choices. Readability tests performed on seven text passages on the site indicated seventh- and eighth-grade reading levels, which is too high for the majority of the targeted audience.

Over a decade ago, there was no standardised method for information sharing during emergency situations. Governments, first responders, and emergency practitioners often had to rely on what little technology that was available to them. This situation slowed down communications, putting entire recovery operations, and lives at stake. The Emergency Data Exchange Language (EDXL) is the umbrella standard for several emergency communication standards that are being developed to address this issue. The Semantic Web is slowly, but steadily becoming a natural extension of the present-day Web. Thanks to efforts from researchers, and corporations such as Google, Facebook, etc., we are seeing more, and more semantics aware applications on the Web. These applications have been successful in bringing Semantic Web technologies to the common user to a large extent. Semantic Web technologies have found applications in a wide range of domains, from medical research to media management. However, a study to see if EDXL messaging standards can benefit from Semantic Web technologies has not yet been made. In this thesis, we investigate the possibility of enabling Semantic Web technologies for EDXL standards, specifically the EDXL Resource Messaging (EDXL-RM) standard, and explore the benefits that can come out of it. The possibility of converting XML based EDXL-RM messages to semantic formats is explored at first. This step is achieved through the evaluation of existing tools and technologies. Based on the outcome of this study, an EDXL to OWL converter that works in two stages is developed. The motivation for enabling semantic support for EDXL standards is illustrated through several use cases.

<p> The purpose of this quantitative research study was to evaluate and determine if significant associations and linear correlations exist between reader socio-demographics, levels of trust and affinity in online citizen writer news story article content, brand loyalty, and acuity in newspaper organizations transitioning from print to online in a heterogeneous population. Descriptive and inferential statistics for dependent and independent variables were analyzed through <i>t</i> tests, Pearson <i>r</i> correlation coefficients, cross tabulation, Pearson Chi-Square, analysis of variance (ANOVA), multivariate analysis of variance (MANOVA), and factor analysis. Data results showed significant relationships for each of the five hypotheses between variables of readers' socio-demographics and level of trust, socio-demographics and acuity, readers' level of trust and level of desire, readers' news brand loyalty and engagement, and readers' engagement and types of online news article content. A post evaluation of observed multivariate associations was conducted through a multivariate analysis of variance (MANOVA) and factor analysis.</p>

<p> This paper examines coverage of the 2012 presidential election by television news organizations on Twitter and their legacy medium. Motivated by the possibility that those who are not interested in politics might pick up relevant political information from the individuals or news organization they follow, this study found that tweets about the election by CBS, CNN, ABC, MSNBC, NBC, and FNN were nearly devoid of substantive political information. This research will be important as news consumption habits continue to evolve and the Twitter audience continues to grow, it is important that we understand the content that is being put forward through these mediums. This will help to evaluate the democratic value of news content in terms of its role in informing the American public.</p>

No description available.