Transportation is closely related to people's lives. The Ministry of Transportation and Communications (MOTC), which is responsible for national transportation affairs, is not only in charge of traffic construction, but also committed to providing better transportation services for the people. Therefore, all units of the MOTC pay great attention to public opinion mailboxes, take the feedback and suggestions from the people as the important reference for governance and improvement.
With the increasing enthusiasm of the people involved in transportation affairs, the opinions received by the public opinion mailboxes of various departments of the MOTC have increased significantly in recent years. In the past year (2017), the MOTC received more than 20,000 pieces of feedback in the public opinion mailboxes. Due to the rapid growth of the number, the traditional way which relies on manual sending, receiving and replying, although it can ensure the quality of the reply, but with the growth of the number, there may be a potential crisis of overburdening. Take 20,000 pieces last year for example, an average of 55 pieces of public opinion were processed a day. If only the working days are calculated, the average number of transactions per day will be more than 80 pieces. In the foreseeable future, the number of letters will continue to grow, which will be a challenge to maintain the quality and efficiency of replying the opinions of the people.
In the case of the growing number of letters, the MOTC hoped to find solutions that can balance the quality and efficiency of public opinion mailbox operations. This coincides with the National Development Council's continued promotion of data application workshops this year, assisting the MOTC in the use of data science and applications to identify possible solutions.
Photo: The discussion meeting of the MOTC.
The MOTC data application workshop was officially launched on 12th June. After many physical meetings and online discussions, the MOTC and the executive team established that if the technology or data science is going to be used to solve the efficiency problem of public opinion mailbox operations, the text mining of the letters of the public opinion mailbox should be first conducted by the data application workshop. Due to the huge amount of information of the public opinion mailbox of the MOTC, it is impossible to analyze it by simple manual or statistical methods. The text mining technology helps to find out the topics and keywords of the problems that most people respond, and establishes related thesaurus as the basis for follow-up technology or system assistance.
After three months of joint work, the initial outcome of the application, including the establishment of the public opinion mailbox thesaurus, and the experimental template of search interface based on the thesaurus, which helps the colleagues of the MOTC search for relevant replies when dealing with the public opinion mailboxes, and find out the key trends of traffic concerns that the public is concerned about. Based on the achievements, a common Q&A that meets the needs of the public, or use the thesaurus to deploy smart Q&A robots could be established to help people quickly find the information they need.
Photo: The experimental template of search interface of the public opinion mailbox (developing).