藉由機器學習模型自動將使用案例規格轉換為任務模型

Abstract

在軟體專案的開發過程中，我們通常會在收到需求後撰寫軟體規格書，然後進行軟體設計和實作。這整個流程非常耗時且需要大量人力。為了解決這一問題，我們致力於開發一個任務自動化系統，以實現整個流程的自動化。在這過程中，產生任務模型（Task Model）是一個關鍵步驟。任務模型是一種可以用來表示使用者操作和系統反應的工具，用於描述系統功能和交互行為。 我的主要任務是自動化從使用案例規格到任務模型生成的過程。由於使用案例規格中的任務具有多樣性和複雜性，很難通過規則的方法生成完全對應的任務模型資訊，因此機器學習技術是必不可少的。我們特別重視自然語言處理（NLP）的應用，因為我們的任務涉及到文本的預測和分類問題。生成的任務模型可以用於後續的前端介面設計和網路應用程式開發，從而提高整個開發流程的效率和質量。

In the software project development process, we usually write software specification documents after getting the requirements, followed by software design and implementation. This entire process is very time-consuming and requires significant human resources. To solve this problem, we are committed to developing a task automation system to simplify the entire process. Generating task models is a key step in this process. Task models are tools used to represent user actions and system responses, which are essential for describing system functionalities and interaction behaviors. My primary task is to automate the process from use case specifications to task model generation. Due to the diversity and complexity of tasks within use case specifications, it is challenging to generate fully corresponding task model information through rule-based methods. Consequently, machine learning techniques are essential. We particularly focus on the application of natural language processing (NLP) because our task involves text prediction and classification issues. The generated task models can be employed for subsequent front-end interface design and web application development, thereby enhancing the efficiency and quality of the entire development process.