基于日志挖掘的移动应用可用性评价研究

发布时间：2018-03-28 14:04

本文选题：可用性　切入点：日志挖掘　出处：《南京大学》2016年硕士论文

【摘要】：《第37次中国互联网络发展状况统计报告》显示,截至2015年底我国手机网民数量已经达到6亿,移动应用的总体使用量上涨了58%。作为用户网络体验的重要部分——可用性已经得到了包括腾讯、百度、阿里巴巴等互联网企业的高度关注。目前,国外对网站和移动应用的可用性研究,无论在理论层面还是在产品的具体评价方面均取得了丰硕的成果。相较而言,我国在该领域的研究主要涉及导航设计的优化、用户体验的提升、基于使用效率的可用性研究等方面。围绕移动应用的可用性主题尚缺乏重量级研究成果。据此,本文开展了以下研究工作：(1)阐述了可用性的概念和评价模型,分析了移动互联网产品的特点,对现阶段移动应用可用性的研究属性、内容和方法进行分析对比,对移动产品的特性进行归纳总结。在借鉴前人研究的基础上,采用扎根理论,对文献中提到的可用性评价指标和测量参数进行开放式编码,提取了95个概念,30个范畴,通过范畴之间的关联关系进行了主轴编码,确定了6个主研究范畴,建立了移动互联网产品的可用性评价模型。(2)采用层次分析法,通过专家打分确定了模型里每个指标对可用性的影响系数。其中对可用性产生影响最大的分别是用户完成任务的比例、完成任务的时间、用户粘度、出错率,所占权重都在10%以上。(3)通过日志数据发现用户的访问行为和访问态度。本文采用的日志数据是“掌上南大”移动应用的阶段性数据,具体工作包括：数据清洗、会话识别等；在得到结构化的数据之后,对用户访问模式进行关联规则挖掘,发现频繁访问集,构建路径访问图；通过路径访问图,分析用户的访问模式,对产品优化提出建议。(4)提取日志里作为可用性评价的参数,对产品整体可用性进行评价研究,即分析新旧版本在有效性、效率性、满意度等方面的表现情况,把可用性评价的指标运用到产品设计里,提出了产品设计应该遵循的基本原则。研究发现,本文构建的评价体系与日志挖掘方法的结合可以大大提升可用性评价的效率和准确性。
[Abstract]:The 37th Statistical report on the Development of China's Internet Network shows that by the end of 2015, the number of mobile phone users in China had reached 600 million. The overall usage of mobile applications has risen by 58%. As an important part of the user's network experience-usability has received high attention from Internet companies, including Tencent, Baidu, Alibaba, etc. At present, Research on usability of websites and mobile applications abroad has achieved fruitful results both in theory and in product evaluation. In contrast, the research in this field in China mainly involves the optimization of navigation design. The improvement of user experience, usability research based on usage efficiency and so on. There is a lack of heavyweight research results around the usability topic of mobile applications. Accordingly, the following research work is carried out in this paper: 1) the concept and evaluation model of usability are expounded. This paper analyzes the characteristics of mobile Internet products, analyzes and compares the research properties, contents and methods of mobile application usability at present, sums up the characteristics of mobile products. Based on the open coding of the usability evaluation indexes and measurement parameters mentioned in the literature, 95 concepts and 30 categories were extracted, and six main research categories were determined by the correlation relationship between categories. The usability evaluation model of mobile Internet products is established. (2) Analytic hierarchy process (AHP) is used to evaluate the usability of mobile Internet products. The influence coefficient of each index on usability in the model is determined by experts, among which, the proportion of tasks completed by the user, the time to complete the task, the viscosity of the user, the error rate, the proportion of the users who have the greatest impact on the usability are respectively. The weight is above 10%.) the user's access behavior and attitude are discovered through log data. The log data used in this paper are the periodic data of "palm Nanda" mobile application, the specific work includes: data cleaning, session identification and so on; After obtaining structured data, mining association rules for user access patterns, discovering frequent access sets, constructing path access diagrams, analyzing user access patterns through path access diagrams, Put forward suggestions for product optimization. (4) extract the log as the parameters of usability evaluation, evaluate the overall usability of the product, that is, analyze the performance of the new and old versions in the aspects of effectiveness, efficiency, satisfaction, etc. The usability evaluation index is applied to product design, and the basic principles of product design are put forward. The combination of the evaluation system and log mining method can greatly improve the efficiency and accuracy of usability evaluation.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：G254

【参考文献】