Sustainable Fault Diagnosis of Imbalanced Text Mining for CTCS-3 Data Preprocessing

At present, the method for fault diagnosis and maintenance of CTCS-3 (Chinese Train Control System Level 3) electronic equipment relies too heavily on expert knowledge. Moreover, use historical data is not valued. This paper proposes a sustainable model based imbalanced text mining. First, to process from field recorded in natural language, language processing technology used extract feature words. Then, term frequency-inverse document frequency transform words extracted database into vectors. It worth noting that imbalance samples affects accuracy this model. To solve problem, we borderline-synthetic minority over-sampling technique step predicting train components, also backpropagation neural network proposed naive Bayesian which commonly as classification model, compare prediction these two algorithms. The experimental results perform well, proves using can further assist engineers complete timely repair work. research has played very important role technical support intelligent dispatching command, will play positive automatic operation urban rail transit under prevention control new coronavirus.