Regresar

Topic Modelling for Automatically Identification of Relevant Concepts Discussed in Academic Documents

Abstract:

A combination of natural language processing and topic modeling identifies topic terms a collection of documents. Latent Dirichlet Allocation (LDA) is an algorithm widely used to infer topics that the document belongs to, on the basis of words contains in it. This research applied LDA to identify topics automatically from academic documents as a way of validating the relevant concepts discussed in the curriculum literature. The experiment involved academic documents about Knowledge Building. We apply some techniques to prepare the data, after data training and validation. Topic modeling was used to identify the topic terms that can be used in the knowledge-building dialogue. Those concepts are meaningful to both the teacher or students because they provide a visualization of the content coherence.