Web genres as tools for selective information retrieval

Abstract

Information about genres seems to be able to improve the information search and selection processes, especially from large collections of documents (Crowston y Kwasnik, 2003). It is assumed that genre reveals the intended use of a document within certain user communities. Research has either focused on all the web genres available in the Web, or on single web genres, such as blogs (Herring et al., 2005) or online newspapers (Ihlström and Henfridsson, 2005). However, attempts at studying specific sets of web genres and the ties between them are less common. In particular, in this research it is our intention to find out those that appear to be the most popular among users, or at least among certain groups of users. In order to do so, we define a specific context and user group, which is often left undetermined in much research into genre. The context of our research is a university f, i.e. the Faculty of Documentation of the Complutense University of Madrid. The participants web users are a selection of the people working and studying there. The participants are interviewed in sessions of 15 minutes approximately. They comment on some of the web pages or web resources they routinely use, explain why they find them useful; and how they use them. Complementarily, the two researchers classified these web pages into web genres. The web genres represent to a certain extent some information needs of the user group considered, at least those information needs that they satisfy using web-based resources. and give us an insight into the way users utilize web based information. This set of web genres can be further confirmed and extended to other user communities. It can also serve as a basis for training algorithms that can recognize exclusively predetermined sets of web genres. The implications of these results can improve selective information retrieval.