A Natural Language Processing and deep learning based model for automated vehicle diagnostics using free-text customer service reports

Initial fault detection and diagnostics are imperative measures to improve the efficiency, safety, stability of vehicle operation. In recent years, numerous studies have investigated data-driven approaches process using available data. Moreover, methods employed enhance customer-service agent interactions. this study, we demonstrate a machine learning pipeline automated diagnostics. First, Natural Language Processing (NLP) is used automate extraction crucial information from free-text failure reports (generated during customers’ calls service department). Then, deep algorithms validate requests filter vague or misleading claims. Ultimately, different classification implemented classify so that valid can be directed relevant department. The proposed model – Bidirectional Long Short Term Memory (BiLSTM) along with Convolution Neural Network (CNN) shows more than 18% accuracy improvement in validating compared technicians’ capabilities. addition, domain-based NLP techniques at preprocessing feature stages CNN-BiLSTM based request validation enhanced (>25%), sensitivity (>39%), specificity (>11%), precision (>11%) Gradient Tree Boosting (GTB) model. Receiver Operating Characteristic Area Under Curve (ROC-AUC) reached 0.82.