随着电子化不断发展，越来越多政府工作都由手工升级为电脑自动处理，电子政务应运而生。电子政务已成为政府工作自动化和智能化的重要手段，电子政务应用软件反映了政府工作效率及处理行政事务的能力，事关广大人民的切身利益，对社会发展有重要意义。随着中国电子政务应用软件市场前所未有的发展，其质量，尤其是可靠性成为了建设电子政务需要考虑的重要方面。

软件测试是保证软件质量的有效方法，而软件可靠性测试是提高软件可靠性水平的重要手段。国家在2005年颁布了《电子政务的标准化指南》，要求电子政务项目应通过国家认证的第三方机构测评，同时将电子政务应用软件的可靠性单独提出。第三方机构可以提供面向全行业的测评工作，包括软件确认测试、验收测试以及专项测试（如软件可靠性测试），但缺乏根据具体行业特点提出针对性的软件测评服务，在电子政务领域更为明显。

针对该问题，本文重点研究了电子政务应用软件的第三方测评策略。由于测评中心对电子政务应用软件可靠性的特殊要求，本文将专门一章详细阐述软件的可靠性测试。对于电子政务应用软件来讲，除可靠性测试外的其他测试，本文称为常规测试，主要包括在确认及验收阶段对软件进行的功能、性能、安全及兼容性等方面的测试。

本文深入细致的研究了电子政务应用软件的特点，在常规测试方面，分析了常规测试的流程、方法及自动化测试工具，并根据不同类型电子政务应用软件分别制定了相应的常规测试策略；在可靠性测试方面，深入分析可靠性测试的流程，包括自顶向下^[2]构建运行剖面及随机选取可靠性测试用例^[11]，有针对的选择可靠性模型，围绕模型提出了适合电子政务的可靠性评估及验证测试方法。基于电子政务应用软件的常规测试策略和可靠性测试策略的研究，我们开发了电子政务应用软件测试策略辅助工具，并详细说明了工具的设计结构，结合界面截图详细的介绍了工具使用。用户根据需求使用工具选择常规测试策略，也可以通过输入运行剖面数据自动模拟可靠性测试过程。

Electronic government (e-Gov.) has appeared along with electronization as increasingly amount of government work has changed from manual operations to computerized processes in recent years. E-Gov. has become indispensable tools for automation and intelligent. The application software of e-Gov. is different from other software in that it reflects the working efficiency and transaction ability in the government, which generates great impact on the society. With unprecedented development of the e-government in China, software quality (software reliability in particular) has played an important role in developing e-government projects.

Software testing is an effective method to guarantee software quality while software reliability test is the key to improve software reliability. Chinese government has issued the “Electronic Government Standardization Guideline” in 2005, which requires e-Gov. projects should be tested by third-party organizations of National Certification. The guideline also raised the focus on the reliability of e-Gov. application. Third-party organizations can provide tests for all across the software industry, including software validation test, acceptance test, and subject test (such as reliability test). However, the limitation of tests offered by the third party is that they are not able to provide specific service for each different industry, which is quite severe in the field of e-government.

Based on the above situation, the thesis has conducted researches on the third-party testing strategies of e-Gov. application and also developed the corresponding tools. Because of the special requirements of software reliability, the thesis defines normal testing and describes the difference between reliability testing and normal testing. Normal testing is the collection of test methods during software validation test and acceptance test, including function test, performance test, security test, and compatibility test. For the e-Gov. application, normal testing means that the tests which were run by the third-party organizations for checking functions, performances, security and compatibility during the period of validation and acceptance.

Furthermore, the thesis presents the characteristics of e-Gov. applications. For normal testing, we analyze the test processes, methods, and automation test tools, and then we recommend corresponding test strategies. For reliability testing, we analyze the test processes including how to build operational profile using top-down method, randomly select test cases, and choose models accordingly. The analysis also covers how to evaluate the reliability and validation processes in terms of the selected model for e-Gov. projects. Based on the above researches, we have developed the deployment tools for e-Gov. application, specifically introduced the design structure of the tools, and explained the usage of the tools with attached screenshots. Thus, users can choose their normal test strategies according to their needs, and they can also use the tools to complete automatic simulation for the process of reliability testing by typing in operational profile data.