ABSTRACT

Although search engines are currently the most effective and popular tools for Information Retrieval on the Web, there is now a consensus that it is still possible to exploit more effectively the potential of these systems. This is particularly true in the current scenario of expansion of social networks, consolidation of the Web 2.0, and emergence of the so called Web of Data. This finding led to the emergence of multiple proposals to increase the expressive power of queries over Web content, both from the syntactical point of view, for example, by the adoption of the XML technology, and from the semantic point of view, for example, through the adoption of the resources collectively known as the Semantic Web.

Although very promising, some of these proposals have run into a difficulty in the adoption of standards, which is an inherent characteristic of the nature of the Web. In this talk we focus on another possible perspective to address this issue: the development of methods and techniques for automatically gathering, extracting and exploiting (semi) structured data that are implicitly available in the vast unstructured textual content on the Web. Works that seek to effectively exploit these data have appeared in the literature for over a decade. However, a series of recent advances in Information Retrieval, Machine Learning and Data Mining, gave this issue a new impulse on the scientific community. This can be evidenced by the considerable space that venues of important areas such as Databases, Information Retrieval and Artificial Intelligence have devoted to research work related to it. Such an interest is justified not only by the challenging problems that arise, but mainly by the growing demand from industry to solve these problems. This makes the results of research on this subject not only immediately applicable, but also motivate a continuous feedback for the scientific investigation around it. The theme involves several classes of problems, and some of these classes of problems will be addressed here, namely: Data Extraction from Textual Sources, Focused Crawling of Web Pages, Integration of Data available in Textual Web Sources and Web Search Considering Structural Features.

EVENT DETAILS