User preferences, query and document categorization, three important variables in relevance calculation

The quality of an information retrieval system depends largely on the satisfaction degree of users with the results obtained when executing a query, so it is essential to design processes that store the preferences patterns of each of them and vary the way in which the results are shown taking into account the specific characteristics of each user. The objective of this article was to present an algorithm for calculating the relevance of the documents provided to users, which used the variables: the user's search profile, the category of the documents and the category of the query as parameters, to customize the results provided by the search engine to the users. In addition, it used as impulse factors the degree of predominance of a search category in the user's profile and the categories to which the document belongs. To validate the algorithm, precision and recall metrics were applied to check that the results obtained are relevant to users.