Design And Implementation Of Comment Mining Supported MOOC System

With the rapid development of Internet technology, a combination of Internet plus education has attracted widespread attention. Online education can make people break through the geographical and time constraints to obtain the best educational resources for learning. Of all the popular online education format, the MOOC mode is the most popular one because of its openness and mass.MOOC system provides the Internet users with best teaching resources. At the same time, it has accumulated a great number of user comments. These comments have recorded the opinions of online courses and the evaluation of the teaching and have very important significance.However, the current mainstream MOOC system do not make good use of these information. Therefore, a comment mining supported MOOC system, which can optimize the MOOC system service and bring a better learning experience by taking advantage of comment mining technology,has an important significance for the development of online education.A large number of scholars have studied on the emotional polarity of comment. However, the emotional polarity are not strong enough to affect the user's behavior decision-making because of the limited information.Therefore, this paper takes the user view contained by the comments as mining object. Aiming at the problem of the divergence of comment content and the diversity of expression patterns, this paper propose a new technical route and a view extract solution based on view pattern.According to the technical route, firstly we construct a keyword vocabulary which represent all course characteristics. After that,considering the dependency relationship, we can learn the inner relationship between the key word and the view word automatically by using the training set. After that, we can abstract the inner relationship into view mode and mine new views by view mode. The accuracy of this new method reached 60.54% and its recall rate reached 71.77%.This paper proposes a new method of view extraction, at the same time, it also apply the new method to MOOC system, design and implement a comment mining supported MOOC system. This paper analyze system functions and describe system level and module division in detail. Most importantly, it elaborates the working mechanism and work flow of comment mining module in the system and explains how the view extraction method applied to the actual system.