Abstract: Every complex system interacts with its environment to perform number of tasks within acceptable tolerances. Any change in performance or unacceptable result is treated as fault or error. The Fault Detection and Diagnosis (FDD) technique is used to determine such faults or errors, to find the root cause and to take necessary actions to prevent such error to occur in the system. This will protect the system from damage and will give maximum throughput. Fault Dependency D-matrix is one such approach to find faults at different levels which consists of dependencies observed and actual failure symptoms associated with the system. But in practical itís difficult to construct d-matrix . In this paper we define a text mining method based on ontology for automatically constructing D-matrix by mining thousands of unstructured data. In this method we first construct the fault diagnosis ontology consisting of concepts and relationships usually observed and examined in the fault diagnosis domain. Then we use the text mining algorithms that make use of the ontology concept to develop the D-matrix. Then we use the text mining algorithms that make use of the ontology concept to develop the D-matrix. Next Graph is created for every D Matrix & from all generated graphs similarity graph is created.

Keywords: Text mining, D-matrix, fault detection and diagnosis, text processing.