Improvement of requirements quality and software process efficiency

建议的研究计划是在软件工程领域，更具体地说，是在研究用自然语言编写的软件需求、它们的质量和可度量性。众所周知，自然语言需求是容易出现缺陷的，为了提高效率，许多技术旨在通过在软件生命周期的早期识别缺陷来验证和确认需求质量。两个研究轴与功能需求的质量和可度量性有关，更具体地说：1)规范方法，2)用自然语言编写的需求的验证方法。我们的研究计划与这两个轴是一致的。实际上，我们的研究目标是首先通过使用功能大小度量（FSM）作为识别这些需求中的缺陷的方法来提高功能需求的质量。然后，我们的长期目标是通过早期包括FSM来改进需求说明方法，以避免可能对软件工程过程效率产生负面影响的某些类型的缺陷。

The proposed research program is in the field of software engineering, more specifically in the study of software requirements written in natural language, their quality and their measurability. It is well recognized that natural language requirements are defect-prone and a number of techniques aim to verify and validate the requirements quality through the identification of defects early in the software life cycle for efficiency purposes. Two research axes are related to quality and measurability of functional requirements, more specifically: 1) Specification methods, and 2) Verification methods of requirements written in natural languages. Our research program is aligned with these two axes. Indeed, our research goal is to first improve the quality of functional requirements by using functional size measurement (FSM) as a means to identify defects in these requirements. Then, our long-term goal is to improve requirements specification methods by including FSM early in order to avoid certain types of defects that could have a negative impact on the software engineering process efficiency. The long-term goal of our proposed research program is to improve methods for specifying requirements in order to prevent defects in the requirements. Our hypothesis is that improving requirements quality using FSM will have a positive impact on the software engineering process efficiency. Software process efficiency is measured by comparing the project effort with the software Functional Size (FS) resulting from that process and can be expressed as a number of hours spent per functional size unit. FSM is normally applied on requirements once they are completed. By moving the measurement earlier within the requirements phase, it would prevent some defects and contribute to requirements of a higher quality, hence better measureable software engineering process efficiency. To achieve this goal, we propose the following research objectives: Objective 1. To design techniques for defect identification in completed software requirements using FSM, in order to improve requirements quality. Objective 2. To investigate early defect detection within the requirements engineering phase and its relationship to the software engineering process efficiency. To achieve these objectives, research projects are defined in collaboration with industry partners and research groups and laboratories. The proposed research program is a natural evolution from progress reported in my recent Ph.D. thesis.