Catálogo de publicaciones - libros

In this paper, an adaptive tool for providing suggestions during the information search process is presented. The tool uses case-based reasoning techniques to find the most useful suggestions for a given situation by comparing them to a case base of previous situations and adapting the solution. The tool can learn from user participation.

A small, preliminary evaluation showed a high acceptance of the tool, even if improvements are still needed.

XML information retrieval (XML-IR) systems search for relevant document fragments in XML documents for given queries. In top- search, users control the size of output by an integer . In XML-IR, however, each output element varies widely in size. Consequently, total output size of top- elements is uncontrollable by simply giving an integer . In addition, search results may have nesting elements. If a system orders result elements simply by their relevance, we may browse the same content more than once due to the nestings. To handle these problems, we propose a new ranking method that enables us to browse search results of XML-IR systems efficiently by introducing the concepts of and . We also propose an evaluation metrics based on and , and compared the metrics with existing XML-IR metrics by experiments.

In this study, we performed a comprehensive evaluation of pseudo-relevance feedback technique for automatic query expansion using OHSUMED test collection. The well-known term sorting methods for the selection of expansion terms were tested in our experiments. We also proposed a new term reweighting method for further performance improvements. Through the multiple sets of test, we suggested that local context analysis was probably the most effective method of selecting good expansion terms from a set of MEDLINE documents given enough feedback documents. Both term sorting and term reweighting method might need to be carefully considered to achieve maximum performance improvements.

Most of the previous works that disambiguate personal names in Web search results often employ agglomerative clustering approaches. In contrast, we have adopted a semi-supervised clustering approach in order to guide the clustering more appropriately. Our proposed semi-supervised clustering approach is novel in that it controls the fluctuation of the centroid of a cluster, and achieved a purity of 0.72 and inverse purity of 0.81, and their harmonic mean was 0.76.

Over the last years, digital preservation has become a particularly active research area. While several initiatives are dealing with the preservation of standard document formats, the challenges of preserving multimedia objects and pieces of electronic art are still to be tackled. This paper presents the findings of a pilot project for preserving born-digital interactive multimedia art. We describe the specific challenges the collection poses to digital preservation and the results of a case study identifying requirements on the preservation of interactive artworks.

Many experiments and studies have been conducted on the application of FRBR as an implementation model for bibliographic databases, in order to improve the services of resource discovery and transmit better perception of the information spaces represented in catalogues. One of these applications is the attempt to identify the FRBR work instances shared by several bibliographic records. In our work we evaluate the applicability to this problem of techniques based on string similarity, used in duplicate detection procedures mainly by the database research community. We describe the particularities of the application of these techniques to bibliographic data, and empirically compare the results obtained with these techniques to those obtained by current techniques, which are based on exact matching. Experiments performed on the Portuguese national union catalogue show a significant improvement over currently used approaches.

Social bookmark services like enable easy annotation for users to organize their resources. Collaborative tagging provides useful index for information retrieval. However, lack of sufficient tags for the developing documents, in particular for new arrivals, hides important documents from being retrieved at the earlier stages. This paper proposes a spreading activation approach to predict social annotation based on document contents and users’ tagging records. Total 28,792 mature documents selected from are taken as answer keys. The experimental results show that this approach predicts 71.28% of a 100 users’ tag set with only 5 users’ tagging records, and 84.76% of a 13-month tag set with only 1-month tagging record under the precision rates of 82.43% and 89.67%, respectively.

Mobile tagging is an extension of social tagging that allows users to associate location-sensitive information with physical objects in the real world. This paper presents MoTag, a mobile tagging application that is used to help people with disabilities share up-to-date accessibility information about buildings and other physical structures to help them navigate their environment. MoTag integrates with G-Portal, a geospatial digital library for storing, managing and retrieving tags.

In the age of Web 2.0, users are increasingly familar with social tagging or bookmarking where comments and ratings are added by users to objects on the web for public consumption. Such comments and ratings are represented in bookmarks which can be used for information or opinion sharing, user interest discovery, and content recommendation. In this paper, we investigate social bookmarking in digital libraries and derive the design requirements for digital library incorporating social bookmarking. Instead of implementing social bookmarking functions in digital library systems from ground zero, we have chosen to explore the possibilities of integrating pre-existing digital library systems with pre-existing social bookmarking systems, and to derive a feasible system architectural design. We also present a case study where , a geography digital library system, is integrated with , an open source social bookmarking system.

With an exponential growth of Weblogs (or blogs), many blog directories have appeared to help users to locate topical blogs. As tags are commonly used to describe blogs, we study the effectiveness of tags in blog classification. Compared with titles and descriptions, our experiments, using 24,247 blogs, showed that tags could lead to better classification accuracy. It is interesting to observe that more tags did not necessarily lead to better classification accuracy. To better describe blogs, we have also proposed a tag expansion algorithm that assigns a blog more tags that are often co-occur with those already associated with the blog. Our experiments showed that tag expansion helped to improve the recall of blog classification with the price of precision degradation.