Abstract
We proposed the method which collects the analogous failure reports using text-processing and ontology. The method has two steps. First, text-processing for the failure reports is performed. By analyzing the result of text-processing, the malfunction information is normalized to a pair of component and malfunction. Second, the words which have similar meaning to the user query are listed by ontology. The analogous failure reports are retrieved by these words from the malfunction information. Proposed method is evaluated using actual report in a shipyard. The result shows the proposed method can obtain failure reports which cannot be found by full-text search.
