A cognitive knowledge base for learning disabilities using concept analysis

Abstract

In recent days, the amount of unstructured data has been widely increased, and it is a challenging task to derive meaningful insights and knowledge from the data. A Cognitive Knowledge Base (CKB) is constructed to derive the hidden semantic knowledge from the unstructured text data. A Knowledge Base (KB) is a collective structure of knowledge developed by ontology or rule-based techniques. A CKB is an Object-Attribute-Relational (OAR) model constituted by a weighted hierarchical concept network. CKB relies on the denotational mathematical structure of formal concepts like concepts, objects, attributes, and knowledge. This paper deals with the construction of a CKB using Formal Concept Analysis (FCA). Construction of CKB using FCA consists of four steps, namely (i) topic modeling guided concept extraction, (ii) determination of objects and attributes, (iii) determination of relevant objects, relevant attributes, hyper-attributes, and (iv) construction of relational hierarchy. The proposed framework derives meaningful insights by extracting main concepts and representing the weighted hierarchical concept network as a relational hierarchy. This paper demonstrates a CKB developed for dyslexic children.