移动互联网时代，随着计算机与网络技术的蓬勃发展，海量的碎片化信息涌现，传统大众传播时代依靠专业化媒体统一对信息进行筛选、加工，再以点对面的方式传递给受众的模式已不再适用，媒介内容分发不得不经历着一轮又一轮的剧变。从偏重人工编辑的媒体型分发到基于智能算法实现信息与受众相匹配的算法型分发，凭借算法强大的数据处理能力，隐藏在社会角落里的“无效信息”得到有效激活，传统社交关系下话语权的垄断被打破，用户日趋个人化、精准化的信息需求正在逐渐被满足。然而，智能算法发展尚不充分，由机器筛选把关的信息质量得不到保证，通过算法流显示的碎片化信息无法给受众提 

供适合深度阅读思考的信息环境，无法带来深层次的愉悦的媒介体验。在信息盈余的传播时代，媒介除了通过传播事实、建构意义来劝服或改变受众态度的功能性作用之外，其对于个人的情感、娱乐性需求的满足同样显得尤为重要。 

威廉·斯蒂芬森于上世纪提出的“传播游戏理论”强调了媒介的娱乐性功能，关注具有自由意志的个体在媒介使用过程中的体验与感受。媒介生产的内容与附加在内容之上的场景建构和情感投射，共同构成了完整的媒介产品，以场景为切入点，以多样化的人机交互为沟通机制，为用户提供全方位的、沉浸式的阅读体验。在众多人机交互模式中，对话式交互最大程度模拟了真实的人际交流，广泛应用于多种网络场景中。这种对话式的交互方式如何实现与用户的互动？又以怎样的机制深入影响用户的认知体验甚至是态度以及行为的改变？本文试图从多重维度解决上述问题。 

本文从认知专注理论切入，聚焦于用户体验的认知层面，综合问卷调查与认知神经科学方法，探讨移动新闻阅读场景下，相比传统网页式交互，对话式交互在用户认知层面会产生哪些不同影响。针对不同内容属性的新闻，对话式交互的效果是否会有所不同。同时，结合EEG脑电数据，比较不同实验条件下用户的问卷数据与脑电数据是否存在关联。 

研究发现，对话式交互条件下，用户的认知专注程度更高，具体体现在用户感知到的沉浸感、愉悦感、控制感、好奇感更强。在EEG指标上，对话式交互条件下，用户的额区EEG偏侧化程度更高，表明用户表达和体验的正向情绪更多。同时，用户脑电信号的theta频段、delta频段更活跃，表明用户对与任务相关的信息投入更多的认知资源和注意分配，为深入加工信息提供了基础。然而，据个体认知与记忆相关的EEG指标显示，对话式交互组的认知性能表现并没有显著高于网页式交互组，用户的认知效果可能还受到信息交互方式之外的其他因素的影响，有待后续进一步深入研究。

In the era of the mobile Internet, with the vigorous development of computer and network technologies, massive amounts of fragmented information have emerged. In the era of traditional mass communication, specialized media are used to select and process information, and then pass it to the audience in a point-to-face manner. This model is no longer applicable, and media content distribution has to go through rounds of upheaval. From the medium distribution which emphasizes manual editing to the algorithm distribution which realizes the matching of information and audience based on intelligent algorithm, with the powerful data processing capabilities of the algorithm, ‘invalid information’ hidden in social corners is effectively activated, the monopoly of rights to speak has been broken, and users’ increasingly personal and precise needs of information are gradually being met. However, the development of intelligent algorithms is still insufficient, and the quality of information filtered by machines cannot be guaranteed. The fragmented information displayed through algorithmic streams cannot provide audiences with an information environment suitable for deep reading thinking, which can not bring a deep and pleasant media experience. In the era of the dissemination of information surplus, in addition to the functional role in persuading or changing the audience’s attitudes by disseminating facts and constructing meaning, it is also particularly important for media to meet users’ personal emotional and entertainment needs. 

The ‘communication game theory’ put forward by Stephenson in the last century emphasized the entertainment function of the media and focused on the experiences and feelings of individuals with free will during the process of media use. The media-produced content and the constructed scene and emotional projection attached to the content together constitute a complete media product, with scenes as the starting point and diverse human computer interaction as the communication mechanism, providing users with an all-encompassing and immersive reading experience. Among many human-computer interaction modes, conversational interaction simulates real interpersonal communication to the greatest extent, and is widely used in various network scenarios. How does this conversational interaction approach interact withusers? What kind of mechanism will deeply influence users’ cognitive experience and even attitude and behavior changes? This article attempts to solve the above problems from multiple dimensions. 

Starting from the theory of cognitive absorption, this paper focuses on the cognitive aspect of user experience. By integrating questionnaire survey and cognitive neuroscience methods, this paper discusses the different effects of conversational interaction on user cognition in the context of mobile news reading compared with traditional web page interaction. Whether the effect of conversational interaction will be different for news with different content attributes. At the same time, the correlation between the questionnaire data and EEG data was compared under different experimental conditions.

The study found that under the condition of conversational interaction, users’ cognitive absorption is higher, which is specifically reflected in users’ sense of focused immersion, heightened enjoyment, control, and curiosity. From EEG indicators, under the condition of dialogue interaction, users’ EEG lateralization in frontal area is more intense, indicating that users express and experience more positive emotions. At the same time, the theta band and delta band of EEG are more active, which indicates that users have invested more cognitive resources and attention on task-related information, providing a basis for further processing information.  However, according to EEG indicators related to individual cognition and memory, the cognitive performance of the conversational interaction group is not significantly higher than that of the web interaction group. Users’ cognitive effect may also be affected by factors other than the way of information interaction, which needs further study in the future.