Graphical Analysis on Text Mining Unstructured Data Using D-Matrix

Abstract

Fault dependency (D-matrix) is used as a diagnostic model that identifies the fault system data and its causal relationship at the hierarchical system-level. It consists of dependencies and relationship between identified failure modes and symptoms related to a system. Constructing such D-matrix fault detection model is time overwhelming task .A system is proposed that describes associate ontology based text mining on unstructured data using D-matrix for automatically constructing D-matrix by mining many repair verbatim text data (typically written in unstructured text) collected throughout the identification process. And also graphical model generation for each generated D-matrix. Initially we construct fault diagnosis ontology and then text mining techniques are applied to spot dependencies among failure modes and identified symptom. D-matrix is represented in graph so analysis gets easier and faulty parts becomes simply detectable. The proposed methodology are implemented as a prototype tool and validated by using real-life information collected from the automobile domain.