A Case Study of Failure Mode Analysis with Text Mining Methods

The maintenance dataset provided by SunWater contains information about failed assets also known as components and their corresponding failure modes. Currently, extraction of this information from the dataset been conducted in a manual manner, which is very tedious, time consuming and cumbersome work. It is necessary to discover an automatic method to decide/extract the failure mode. This paper presents three methods that were attempted in an effort to solve this problem. The performance of each method is analysed in detail and suggestions for how the outcomes can be improved are also proposed. .

• Type: Conference
• Year: 2007
• Where: AUSDM
• Authors: Lin Chen, Richi Nayak
• Keywords: AUSDM 2007, Corresponding Failure Modes, Data Mining, Failure Mode, Maintenance Dataset