学生论文详情

中文摘要：随着全面健康理念的不断普及和深化，人们越来越意识到自己是个人“健康第一责任人”，主动对网络健康信息的搜索、获取与选择利用逐渐成为人们健康生活的重要组成部分。然而，我国网络用户的健康信息素养整体偏低，健康信息辨别能力也偏弱，不容易从质量参差不齐的网络信息海洋中搜索到高质量健康信息，这一问题严重影响了我国民众的健康生活，也成为党和国家关注的重要民生问题之一。也正因如此，如何帮助民众搜索并获取高质量健康信息、增长健康知识、提升解决健康问题的能力、保障健康生活水平已成为情报学领域关注的热点问题。而已有相关研究对不同用户在不同任务情境下网络健康信息交互行为、交互感知以及交互质量间的差异或关联关系分析、用户获取不同质量网络健康信息的影响因素、用户网络健康信息交互质量预测模型及有效干预策略与机制等方面的研究成果还很缺乏。基于这样的社会背景与学术背景，本研究希望通过分析并回答以下4个研究问题：①不同个体特征用户网络健康信息交互行为有哪些特征？②不同任务情境下用户网络健康信息交互行为有哪些特征？③不同任务情境下用户网络健康信息搜索时的交互行为、交互感知与交互质量间的关系是怎样的？④用户网络健康信息交互质量如何被有效预测，低质量交互行为又如何被引导？旨在促进健康信息系统个性化、智能化的设计与优化，从而帮助民众高效获取高质量健康信息、学习和积累健康知识，提升国民健康信息素养与生活健康水平，助力“健康中国2030”战略实现。 基于以上研究背景与问题，研究首先对相关概念、相关基础理论及已有相关研究进行了梳理与回顾，进一步分析了研究开展的必要性与可行性。然后，在借鉴已有研究基础上，采用用户实验方法，通过问卷、半结构化访谈、用户搜索行为记录软件等方式或工具对用户个体特征及用户网络健康信息搜索过程中的交互感知、交互行为、交互质量等多种不同类型数据进行了收集，并针对数据结构特征与研究目标，采取了相应的数据分析方法，对相关问题展开了分析，得到了以下结果与结论： ①不同个体特征用户（性别、学历、学科、信息检索学习经历、计算机使用频率、信息搜索经验、健康信息素养与认知风格）在选择不同健康信息类型（不同信息形式）与健康信息源时均存在一定偏好（研究问题1）；任务属性/类型对用户健康信息类型、健康信息源类型选择也存在显著影响（研究问题2）。②任务复杂度、任务产品类型与健康信息话题类型等任务属性不仅对用户网络健康信息搜索时的内容交互行为、系统交互行为有显著的主效应，而且不同属性间的交互效应也一定程度上影响了用户交互行为（研究问题2）。③感知任务困难程度、方法和过程熟悉程度、感知任务复杂度、搜索经历丰富程度、信心程度、信息有用性判断难度、获取信息的认知强度、确定有用信息努力程度、任务感知成功、任务感知挫败感、任务感知满意等交互感知子维度均与交互行为存在显著相关关系，而且鼠标点击最大间隔时间、任务持续时间、查询修改次数、点击链接次数、使用推荐查询数、鼠标移动像素量、保存信息条目/浏览信息条目、访问检索系统个数、鼠标滑轮滚动次数与鼠标左键点击次数等交互行为指标可作为自变量来解释和预测用户交互感知状态（研究问题3）。④不同任务属性对用户网络健康信息搜索过程中的任务交互质量有明显的主效应作用，不同任务属性间的交互效应对交互质量也有显著影响（研究问题3）。⑤用户搜索网络健康信息过程中的交互感知对交互质量有显著的影响，且多重线性回归分析结果表明信息有用性判断难度、获取信息的认知强度、方法和过程熟悉程度与感知任务复杂度能有效解释交互质量的高低变化（研究问题3）。⑥交互行为各项指标与交互质量存在不同程度的相关关系，且多重线性回归模型显示鼠标移动像素量、鼠标左键点击次数、浏览条目总数、鼠标滑轮滚动次数、使用推荐查询数、键盘输入最大间隔时间等交互行为指标可有效解释交互质量的变化（研究问题3）。⑦交互感知、交互行为对交互质量影响作用路径分析结果表明，交互感知?交互质量、交互行为?交互质量的直接效应显著，而交互感知?交互行为?交互质量的间接效应不具有显著的统计学意义（研究问题3）。⑧以相关性变量（包括线性回归关系自变量）与任务属性为输入变量的1层隐含层BP神经网络模型与以相关性变量（包括线性回归关系自变量）、任务属性与个体特征为输入变量的2层隐含层BP神经网络模型可以有效预测不同任务完整会话的交互质量，而基于2分钟时间切片的2层隐藏层BP神经网络模型在交互质量实时预测上有较好的表现（研究问题4）。 基于以上发现，研究进一步对用户网络健康信息交互行为引导机制进行了分析和探索，构建了基于完整任务会话预测模型与基于任务会话时间切片实时预测模型的交互行为引导机制，并对低质量交互与交互行为引导策略触发机制进行了分析（研究问题4）。考虑实验研究发现用户在进行不同健康信息搜索任务时交互质量整体偏低，且用户健康信息素养整体上不高。研究基于分析结果与结论，并结合社会认知理论，针对用户健康信息素养与交互质量提升策略也进行了探讨（研究问题4）。 本研究的相关结果与结论及其带来的启示，一方面，可以促进用户网络健康信息行为研究的深化发展与相关知识体系的完善，同时也为未来相关研究的推进提供了参考；另一方面，相关策略与机制的提出为健康信息系统个性化搜索功能设计与优化、用户健康信息素养与交互质量的提高提供了方法路径与实施策略，从而帮助用户高效获取高质量健康信息，学习健康知识，科学合理地解决健康问题，进而提高公民整体健康素养和健康生活水平。最后，文章根据研究的局限与不足对未来研究进行了展望。图105幅，表156个，参考文献404篇。  英文摘要：With the popularization and deepening of the concept of comprehensive health, people are increasingly aware that they are the first person responsible for their health, and the active searching, acquiring and selecting of health information on the Web are becoming an important part of people's healthy life. However, the health information literacy of Web users in China is low, and the discrimination ability of health information quality is also weak, which makes it difficult to search high-quality health information from the Web. All of this seriously affects the healthy living condition of Chinese people, resulting in a serious livelihood issues concerned by the national government. Under this background, how to help the public or web users to search and obtain high-quality health information, increase their health knowledge, improve the ability to solve health problems, and keep healthy has become a hot research issue in the field of Library and Information Science (LIS). Unfortunately, few existing researches have been conducted on the correlations or differences between web health information interactive behavior，interaction perception and interaction quality of different users in different task situation, and the factors which influencing users’ access to different quality of web health information. The prediction model of user interaction quality when they searching for web health information, and effective intervention strategies and mechanisms are also rarely involved. Considering these social and academic concerns, this study aims to find answers or solutions for the following four research questions: (1) What are the differences between Web health information interactive behaviors of users with different individual characteristics? (2) What are the differences between Web health information interactive behaviors while users are searching for different task? (3) What are the relationships between users' interactive behavior, interaction perception and interaction quality when searching for Web health information under different task situations? (4) How can the interaction quality of users searching for Web health information be effectively predicted, and how can low-quality interactive behaviors be led to a right way? By answering these questions, we hope to help people learn and increase health knowledge, improve the national health information literacy and the health level of people’s daily life, contributing to the realization of "Healthy China 2030" strategy. Based on the above research background and questions, this paper firstly reviews and summarizes relevant concepts, relevant basic theories and existing researches, and further analyzes the necessity and feasibility of current study. Then, it introduces how the user experiments was designed and conducted for answering the research questions. In the experiment, questionnaires, semi-structured interviews, user search behavior recording software are used to collect the data of user's individual characteristics, interaction perception, interactive behavior and interaction quality in the process of user searching Web health information. Based on the characteristics of data structure and research objectives, corresponding data analysis methods are adopted to analyze related problems and the following findings are obtained: (1) Users with different individual characteristics (gender, educational background, discipline, learning experience of information retrieval, frequency of computer using, information searching experience, health information literacy and cognitive style) have certain preferences when selecting different types of health information and health information sources(for RQ1), and task attribute/type also has significant influence on users’ selection of health information form and health information source(for RQ2). (2) Task attributes, such as task complexity, task product type and topic type of health information, not only have significant main effects on users' interactive behavior, including content interactive behavior and system interactive behavior, but also have the interaction effects between different attributes on users' interactive behaviors to some extent (for RQ2). (3) Except for familiarity with the subject, all the sub-dimensions of interaction perception are significantly correlated with interactive behavior. In addition, interactive behavior indicators, such as Maximum time between mouse clicks、Time on task、Number of query modification、Number of hyperlink clicks、Number of recommended query accepted、Mouse movement、Saving useful document or page/ Total number of items viewed、Number of IR system consulted、Wheel scrolling and Mouse clicks-left button, can be used as independent variables to explain and predict user interaction perception state (for RQ3). (4) Different task attributes have a significant main effect on task interaction quality, and the interaction effect between different task attributes also has a significant effect on task interaction quality (for RQ3). (5) The interaction perception that users have in the process of searching for Web health information has a significant impact on the interaction quality, and the results of multiple linear regression analysis show that the Difficulty of information usefulness judgment, the Cognitive load of acquiring information, the Familiarity of methods and processes, and Perceived task complexity can effectively explain the change of the interaction quality (for RQ3). (6) Various indicators of interactive behavior are correlated with interaction quality to different degrees, and the multiple linear regression model shows that Mouse movement、Mouse clicks-left button、Total number of items viewed、Wheel scrolling、Number of recommended query accepted、Maximum time between keystrokes can effectively explain the change of interaction quality (for RQ3). (7) The path analysis of interaction perception and interactive behavior on interaction quality shows that the direct effects of "interaction perception?interaction quality" and "interactive behavior?interaction quality" are significant, while the indirect effects of "interaction perception? interactive behavior? interaction quality" are not statistically significant (for RQ3). (8) One hidden layer of BP neural network model, with correlation variables and the task properties as input variables, and 2 hidden layer of BP neural network model, with the correlation variable, the task properties and individual characteristics as the input variables, can effectively predict the interaction quality of a complete task session. And the two-layer hidden layer BP neural network model based on 2-minute time slice has better performance in real-time prediction of interaction quality (for RQ4). Based on the above findings, users’ Web health information interaction behavior leading mechanism is analyzed and explored， and two interactive behavior leading mechanisms, basing on complete task session prediction model and real-time task session time slice prediction model, are constructed (for RQ4). Considering that current experimental study finding that the interaction quality of users in different health information search tasks are generally low, and the health information literacy of users are not good, therefore, this study explores strategies for improving user health information literacy and interaction quality, by revisting the results and conclusions, from the perspective of social cognition theory (for RQ4). The results and conclusions of this study, on the one hand, promote the deepening development of user Web health information behavior research and the improvement of relevant knowledge system, and also provide reference for the promotion of future research; on the other hand, the proposed strategies and mechanisms provide methods and practical guidance for the design and optimization of personalized information search function of health information system, as well as improving of users' health information literacy and interaction quality. Finally, the future research is prospected according to the limitations and shortcomings of current research. There are 105 figures, 156 tables and 404 references.  

相关知识

食疗养生论文（精选8篇）,中医食疗学论文
学科教育论文
大学生心理健康教育论文参考文献
运动康复论文（最新范文10篇）,体育保健学论文
贵州省医科大学预防医学专业详情
运动营养学论文（必读10篇）,营养学论文
大学健身塑形课论文
营养减肥论文（推荐6篇）,营养学论文
大学生心理健康论文2000字【精选9篇】
健康营养学论文题目推荐

网址: 学生论文详情 https://www.trfsz.com/newsview142972.html