DevOps开发过程中用户群智需求的理解、挖掘应用技术研究

DevOps作为日渐兴起的软件开发过程模型，通过紧密结合开发和运维过程，使得软件得以有效的进行持续演化。基于DevOps的互联网类应用中来自广大用户群体的需求是推动持续演化的源泉。尽管个体用户需求天差地别，但群体需求却存在共性，而共性需求的形成则是群体智能的表现。目前，群体需求的体现通常包含于用户群体的行为反馈中，通过用户产生数据挖掘群智需求，具有良好的研究和应用价值,其中有两大重要挑战：1）如何对异质数据进行统一建模和融合理解；2）基于计算机对用户产生数据的理解，如何对其中群智需求进行挖掘。目前，为DevOps提升其自动化程度的技术研究以及用户使用软件的行为分析等方面有着一定的进展，但鲜有研究重点关注基于用户产生数据的用户群体群智需求挖掘。为此，本研究希望借助数据挖掘相关技术在群智需求挖掘这一领域进行系统性的探索，在解决DevOps中群智需求挖掘的相关研究问题的同时，促进技术本身的发展。

With the improvement of software development, DevOps aims to achieve continuous software revolution by unifying software development (Dev) and software operation (Ops). Currently, many Internet companies use DevOps in their web-based applications. The requirements from users, which as the product of collective intelligence, are the key force for continuous delivery. Therefore, it is worthy mining collective requirements from the user generated content (UGC), and leads two main challenges: 1). How to understand and model the heterogeneous data from UGC, and 2) how to mining collective requirements? Recently, existing work focuses on advocating automation and monitoring in DevOps. Few work targets to mining collective requirements from UGC. Therefore, we aim to discover mining collective requirements systematically by leveraging the techniques of data mining, which in turn solve the target problem in this application.