The purpose of the present study is to develop an effective and objective knowledge extraction methodology with text mining technology especially for the ageing management in Nuclear Power Plants (NPPs). In the present study, eighteen out of fifty topics for latent material degradations (Generic Issues) discussed in the Proactive Materials Degradation Committee (PMDC) have been surveyed by interviewing users related to the PMDC and investigating the online forum of the committee. Then the records including keywords related to Generic Issues (Seeds) have been extracted from a public database NUCIA. Then knowledge extraction system developed in our group has been applied to extract records that have been similar to Seeds with statistical way. All the extracted records have been reviewed and summarized and the summaries for each Generic Issue has been clustered as network structures by text mining software "SPSS Text Analysis for Surveys". As a result, failure mechanisms not considered in "Code for Ageing Management Program in Japan" have been found by analyzing structures. These results suppose that the methodology developed in this study could support users to raise awareness to some cues related to latent failure mechanisms and they could be effective to support the discussion about proactive materials degradation management (PMDM).