联系我们

论文查询结果	返回搜索

论文编号：	12366
作者编号：	2120183028
上传时间：	2021/6/9 16:54:27
中文题目：	信息茧房现象的影响因素及预测研究
英文题目：	Study on Influencing Factors and Prediction of Information Cocoons
指导老师：	李月琳
中文关键字：	信息茧房;认知因素;选择性心理
英文关键字：	Information Cocoons; Cognitive Factors; Selective Psychology
中文摘要：	面对着信息庞杂、时时更新的互联网环境，网民可能陷入大量同质化的信息中，形成个人的信息茧房，逐渐丧失独立思考与理性判断的能力。 为深入分析信息茧房现象的产生，本研究从当前网络媒体环境出发，探究研究现象的形成原因，因而提出两个研究问题，分别是信息茧房现象形成的影响因素主要有哪些?这些影响因素对信息茧房现象形成产生的影响程度是否不同?为了回答上述问题，本研究参考已有文献，提炼出识别网民产生信息茧房现象的评价指标和影响因素的相关量表，探究信息茧房现象形成的主要因素。 同时，本研究有效回收 963 份问卷。基于所获得的数据，使用因子分析和基于 最小二乘法的线性回归等量化分析方法，探寻不同因素对茧房现象形成的影响 程度，并通过线性回归的结果和机器学习技术对茧房现象进行预测。 研究发现，信息茧房现象形成的主要影响因素可以分为 4 个维度，8 个因素，分别是网民的行为因素(信息获取渠道、信息软件的使用习惯)、认知因素 (兴趣爱好和价值选择)、网络信息因素(信息的有用性、信息的同质性)和网络环境因素(社群影响和算法推荐)。在上述 8 个影响因素中，网络用户的认知因素(兴趣爱好与价值选择)是对信息茧房现象产生影响程度最大的因素；其次，信息同质化与算法推荐等外部因素是对该现象产生影响较大的因素；而其他四个影响因素(信息获取软件的使用习惯、社群影响、信息有用性、信息获取渠道)对该现象的产生也有不小的影响。 本研究通过基于最小二乘法的线性回归的方法验证了因子分析得出的权重体系，并且进一步明确了不同影响因素的权重大小。其后分别采用线性回归的 结果和机器学习的技术进行预测，现象预测的成功率超过90%。本研究一方面丰富了信息茧房现象形成因素的理论体系;另一方面，其研究结果有利于为应对信息茧房现象的产生提供对策，具有一定的实践意义。
英文摘要：	Faced with the complex and constantly updated Internet environment, netizens are easily trapped in a large amount of homogeneous information, forming a personal information cocoon, and gradually losing the ability to think independently and make rational judgments. In order to analyze the phenomenon of information cocoon in-depth, this research starts from the current network media environment and proposes two research questions in view of existing research phenomena. What are the main factors affecting the formation of an information cocoon? Do these influencing factors have different degrees of influence on the formation of the information cocoon room? In order to answer the above questions, this study extracts the evaluation indicators and influencing factors related to the phenomenon of information cocoons generated by netizens, and explores the main factors of the phenomenon of an information cocoon. This study effectively retrieved 963 questionnaires. Based on the obtained data, quantitative analysis methods such as factor analysis and linear regression based on the least square method are used to explore the degree of influence of different factors on the formation of the information cocoons, which is also predicted through machine learning technology. The study finds that the main influencing factors for the formation of the information cocoon can be divided into 4 dimensions, 8 factors, namely the behavioral factors of netizens (information acquisition channels, information software usage habits), cognitive factors (interests and hobbies and value choices), network information factors (usefulness of information, homogeneity of information) and network environment factors (community influence and algorithm recommendation). Among the above 8 influencing factors, the cognitive factors (hobbies and value choices) of Internet users are the factors that have the greatest impact on the phenomenon of information cocoons; secondly, external factors such as information homogeneity and algorithm recommendation have a relatively great impact on the phenomenon. The other four influencing factors (the usage habits of information acquisition software, community influence, information usefulness, and information acquisition channels) also have a considerable impact on the occurrence of this phenomenon. This study verifies the weighting system obtained by factor analysis through the method of linear regression based on the least-squares method and further clarifies the weights of different influencing factors, and then uses the results of linear regression and machine learning technology to predict the information cocoons. The success rate of phenomenon prediction exceeds 90%. On the one hand, this research has enriched the theoretical system of the forming factors of the information cocoons; on the other hand, the research results are conducive to providing countermeasures to deal with the information cocoons, which has enough practical significance.
查看全文：	预览  下载(下载需要进行登录)