Evaluation of User Interaction with In-Vehicle Systems - A Graph-Based Exploratory Literature Review Method Focused on Data Connections

The dataset contains a graph-based representation of research papers from the research area of human-computer interaction for in-vehicle systems. For the proposed approach several thousand papers from different conferences, journals and monographs in the domain of in-vehicle interaction were systematically filtered, classified, and stored in a graph database. An exploratory survey shows a trend for usability assessment methods with direct involvement of users, especially the observation of users and performance-related measurements, as well as questionnaires and interviews. The used graph database platform neo4j allows to query the developed dataset while focusing on connections between the captured data. The data was manually classified from three conferences, eleven journals and three monographs between 2013 to 2017. The relevant articles were selected through different selection criteria. An explicit thematic relation to one or more us- ability methods, a practical application of the method in an empirical part of the study, and a relation to the context of use for in-vehicle interaction.本数据集包含面向车载系统的人机交互（Human-Computer Interaction）研究领域内科研论文的图结构表示形式。针对本研究提出的方法，研究团队从车载交互领域的各类学术会议、期刊及专著中筛选出数千篇论文，经系统化过滤、分类后存储至图数据库（graph database）中。一项探索性调查显示，直接引入用户参与的可用性评估方法呈现出发展趋势，其中尤以用户观察、绩效相关测量，以及问卷与访谈法为代表。本次研究所使用的图数据库平台Neo4j（neo4j）支持针对所构建数据集的查询，且可聚焦于捕获数据间的关联关系。 2013年至2017年间，研究团队从3个学术会议、11种期刊及3部专著中选取相关文献，并通过多项遴选标准完成筛选。具体筛选标准包括：文献需与一种或多种可用性评估方法存在明确的主题关联，所提及的方法需在研究的实证部分得到实际应用，且需与车载交互的使用场景存在关联。