Unsupervised Information Extraction and Retrieval

Unsupervised Classification

Unsupervised classification is a division of data without a learning procedure; it is an automatic process which extracts common features or relationships. This type of classification is also called self-adaptive because the division of data is modified to fit the incoming data set.

The modification are occurred while come in the data even when are a model of the in data.

The modes of search are:

• Likeness level among concepts, this mode looks for close concepts. In retrieval information means compare concepts and structure information.

• Features and relationships extraction, features of concepts and the relationships among them are sought. Similar concepts should make similar results.

• Main feature analysis and extraction, this method chooses singularity and fundamental features in concepts to have enough information for the retrieval, this is important because it reduces the amount of search and the volume of initial data.

• Modelling to group and search the prototype that symbolizes the whole group. Unsupervised classification is a division without a teacher, that is, an automatic process that extract usual characteristics or relationships. That kind of classification, also called self-adaptive because the modification are occurred while come in the data even when are a model of the in data.

This site is developed for the assignement Information and Retrieval Organization of the subject Information Retrieval and Access over Computer Engineering of Carlos III University of Madrid.

Available versions:

Download the DOC article

Download the PDF article