【作者】 刘伟；

【导师】 刘正捷；

【作者基本信息】 大连海事大学 ， 计算机科学与技术， 2021， 硕士

【摘要】 智能手机的普及使得移动APP被广泛使用。对移动APP进行可用性评估,可以帮助产品团队发现其中的可用性问题,提升产品的用户体验。由于移动APP使用情境多变,使得传统基于实验室的可用性评估难以很好地发现其在真实情境中的可用性问题,因此需要进行远程可用性评估。但目前的远程可用性评估方法大多限于可用性测试场景,不适用于自然场景,且存在过多依赖人工分析、较少考虑情境因素等问题。为解决上述问题,本文作者基于情境感知的用户数据采集工具CAUX,探索建立新的移动APP远程可用性评估方法。本文工作分为三部分:(1)利用CAUX,以情境感知的方式自动捕获用户使用APP时的录屏和情境数据。其中为了方便对数据做进一步的自动分析,采用图像处理技术,将录屏数据自动转换为用户行为序列。(2)优化现有基于可用性测试场景的远程可用性评估方法。根据所捕获的用户执行可用性测试任务时的行为和情境数据,开发和建立半自动化的用户任务分析方法,自动识别用户任务中存在逻辑和时间偏差的异常点,并以此为线索,在可视化技术支持下,结合情境信息,引导可用性评估人员分析和发现可用性问题,并进行了实际的应用案例验证。(3)建立基于自然场景的远程可用性评估方法。根据所捕获的用户日常使用移动APP时的行为和情境数据,开发和建立半自动化的用户行为分析方法,自动将用户行为序列拆分识别为对应于产品功能的用户任务,并结合情境信息从任务执行频度和执行模式两个方面,提取和分析用户行为模式,发现用户的行为规律以及移动APP中存在的可用性问题,并进行了实际的应用案例验证。研究结果表明,本文所提出的移动APP用户行为捕获方法,能以较高的数据准确度实现对用户行为和情境数据的自动获取;所建立的面向可用性测试场景的远程可用性评估方法,能够结合情境信息对用户行为进行半自动化分析,使发现的可用性问题更加详细具体,且有效地提高了分析和评估效率;所建立的面向自然场景的远程可用性评估方法,突破了现有方法仅限于可用性测试场景的局限性,拓展了移动APP远程可用性评估方法的应用范围。更多还原

【Abstract】 Due to the popularity of smart phones,mobile APP is widely used.The usability evaluation of mobile app can help the product team find its usability problems and improve the user experience of the product.Due to the variety of mobile app usage situations,it is difficult to find the usability problems in real situations by traditional laboratory-based usability evaluation.Therefore,it is necessary to use remote evaluation method.However,most of the current remote usability evaluation methods are limited to usability testing scenarios,which are not suitable for natural scenarios,and there are some problems such as relying too much on manual analysis and less considering the situational factors.To solve the above problems,the author of this thesis explores and establishes new remote usability evaluation method of mobile app based on the user data acquisition tool CAUX based on situational perception.This thesis is divided into three parts:(1)CAUX is used to capture user’s screen recording and situation data in the way of context awareness,and image processing technology is used to automatically convert the screen recording data into user behavior sequence.(2)Optimize existing remote usability evaluation methods based on usability test scenarios.According to the behavior and situation data of the captured user performing usability test task,a semi-automatic user task analysis method is developed and established to automatically identify the outliers with logical and time deviation in the user task.Based on this clue,with the support of visualization technology,combined with situational information,the usability evaluators are guided to analyze and find usability problems,and the actual application case is verified.(3)Establish a remote usability evaluation method for mobile apps based on natural scenes.According to the captured behavior and situation data of users’ daily use of APP,semi-automatic user behavior analysis method is developed and established to automatically convert user behavior sequences into user tasks representing functional use.Combined with context information,the user behavior patterns are extracted and analyzed from two aspects of task execution frequency and execution mode to help usability evaluators find user behavior rules and usability problems in APP,and the actual application case is verified.The results of research shows that the proposed method of user behavior capture in mobile APP can automatically capture user behavior and contextual data with high data accuracy.The established remote usability evaluation method for usability test scenarios can semi-automatically analyze user behavior data in combination with context information,find more detailed and specific usability problems,and effectively improve the efficiency of analysis and evaluation.The established remote usability evaluation method for natural scenes breaks through the limitations of existing methods that are limited to usability test scenarios,and expands the application range of mobile APP remote usability evaluation methods.更多还原