Discovering and Ranking Relevant Comment for Chinese Automatic Question-Answering System

Intelligent customer service system is timely, efficient, and accurate, which more popular in grid electric power companies, the amount of consultation increasing day by day. It infeasible for human to answer these questions on time, so an automatic question-answering great help company. The queries from company very different open-domain questions: problems questioned tend be a specific device or within enterprise operation problem. Most companies provide customers with communication platform where can get guidance using equipment business process. comments platforms are valuable resources answering questions. In our work, we use three neural network models excavate potential answers comments. One key challenges, however, difficulty matching To solve this problem, propose method based deep learning find related generate accurate reliable answers. Experiments prove that performed well