Obtaining feedback for indexing from highlighted text

The indexing techniques most commonly applied in information retrieval systems associate weights to the terms of documents; this association is carried out in order to quantify the term’s capacity for representing the document. Subsequently, these weights are used in query processing, playing an important role in the performance of the system. This work presents a method for the adjustment of index weights by processing text that has been highlighted by users. Text in a document is highlighted using an application which provides annotation facilities in order to support active reading. Highlighted text is supposed to be of special relevance and the method tries to improve the index weights of terms located in these highlighted fragments.