ABSTRACT
A network text analysis is a way to extract knowledge from texts and generate a network of words. A central premise is that the network represents the mental model of the author. After transforming an unstructured text into a structured network, text analytic methods can be used to analyze the network conducted by specific networks. Moreover, this kind of information representation can be one technique to achieve the underlying semantic structure of a text and make the mental models of different authors comparable. In evolving knowledge resources such as wiki articles, the extracted networks can be utilized to compare the uncovering of misconceptions, knowledge conflicts between authors, or the identification of latent relations between concepts in a particular knowledge domain. A network text analysis and visualization are used for the concept network. There are three main steps in the process – concept identification, relationship identification, and network generation. Various techniques are available for each of these steps. Identified concepts for extracting concepts and relations are based on an open information extraction tool (ClausIE). Three steps are supported to extract labeled relations between concepts: extraction of candidate relations and a-posteriori filtering by the user. The solution, which can be easily incorporated into existing process chains for network extraction from texts, is compatible with arbitrary approaches for concept extraction. In this chapter, we review the existing research articles on concept networks and network text analysis to find some research gaps and discuss the methods of applying concept networks and network text analysis.
