Concept map construction applying natural language processing on text extracted from e-commerce web pages

摘要

To understand the theory behind a given subject can be an expensive task depending on the domain complexity. In the products case, the quantity of functionalities and variants can be a big impediment for the user take a decision, since he/she has to understand each detail before choose from an offers variety. Moreover, there is a huge amount of information available in the internet and tutorials to explain each concept that is involved in each subject. The concepts are related to each other to create the main idea of the subject. This paper presents a methodology to extract these concepts in a way that it is easier to an end user to rapidly understand the subject using concept maps. Its also aims to extract the relationships between the concepts showing how they are related. Many techniques were combined to provide this knowledge mechanism, like natural language processing and information extraction. In the web documents case the information is presented with particularities, like tables, than it should be maintained to better access the associations.