联系我们

论文查询结果	返回搜索

论文编号：	15428
作者编号：	2120223822
上传时间：	2025/6/11 11:52:50
中文题目：	基于Kano-QFD方法的网络游戏智能客服系统优化研究
英文题目：	Research on Optimization of Question Answering System for Online Games Based on Kano-QFD Method
指导老师：	李颖
中文关键字：	用户需求；网络游戏；人工智能+；智能问答系统；Kano-QFD
英文关键字：	user demands; online games; artificial intelligence plus; question answering system; Kano-QFD method
中文摘要：	我国游戏行业正逐步由增量市场转向存量市场，亟待提升用户体验以增强用户黏性。游戏客服对于用户体验和黏性存在重要影响。伴随着“AI+”时代的悄然来临，我国网络游戏陆续接入基于AI的智能客服系统。当前，由于缺乏对用户需求的有效分析，以及对于AI大模型及相关生成式AI工具的发掘和应用不够充分等原因，我国网络游戏智能客服系统存在功能项点建设不完善，用户交互体验不理想，解决问题效率、能力不足等普遍问题。 本文首先基于Kano模型，对用户需求进行了调研、梳理与分析，得到系统安全性等6项需重点避免用户不满意的必备需求；多媒体交互等4项是否满足对于用户满意与否均存在显著影响的期望需求；情境理解能力等2项可重点着力以提升用户满意度的魅力需求；界面美观性等6项对于用户满意度影响较不显著的无差异需求；系统稳定性等16项需结合所具备的不同需求类型特征加以综合分析、研判的混合类型需求。 进一步基于各需求项点的Kano类型和Better-Worse系数，运用QFD方法分析各需求项点同技术特征间的相关性。QFD分析结果表明，模型性能、数据处理及数据集成具备相对较高的技术重要性，系统基础设施建设，人机交互、系统架构建设及运维管理等亦是需要得到重视的技术要素。其中，模型的语义分析、逻辑推理及信息组织能力，系统知识库构建及相关数据的集成与处理等具体细分技术特征指标尤为重要。另考虑到技术特征内在关联，硬件资源、网络服务，模型的学习迭代能力等亦得到足够重视。 基于上述Kano-QFD方法对用户需求的调研、分析与转换的结果，笔者在模型选择、部署及系统支撑，系统知识库建设、数据集成及数据处理，情感、多模态交互性能优化及模型的迭代提升，系统业务侧建设等方面对网络游戏智能客服系统的建设和优化完善提出相关意见或建议，以期持续保障网络游戏客服系统对用户需求的满足能力，持续提升用户体验。
英文摘要：	The gaming industry in China is gradually shifting from an incremental market to a stock market, and there is an urgent need to improve user experience to enhance user stickiness. Game customer service has a significant impact on user experience and stickiness. With the quiet arrival of the "AI+" era, China's online games have gradually integrated AI based question answering systems. Due to the lack of effective analysis of user demands and insufficient exploration and application of AI models and related generative AI tools, there are common problems in the current question answering systems for online games in China, such as incomplete construction of functional items, unsatisfactory user interaction experience, and insufficient efficiency and ability to solve problems. Firstly, based on the Kano model, this article conducted research, sorting, and analysis on user demands, and identified six essential requirements for avoiding user dissatisfaction, including system security; Whether the four aspects of multimedia interaction meet the expected needs that have a significant impact on user satisfaction or not; Two key areas that can be focused on to enhance user satisfaction are situational understanding ability and charm needs; There is no significant difference in the impact of six factors, including interface aesthetics, on user satisfaction; 16 mixed type requirements, including system stability, need to be comprehensively analyzed and judged based on the characteristics of different types of requirements. Further based on the Kano type and Better Worse coefficient of each demand item point, the QFD method is used to analyze the correlation between each demand item point and technical features. The QFD analysis results indicate that model performance, data processing, and data integration have relatively high technical importance. System infrastructure construction, human-computer interaction, system architecture construction, and operation and maintenance management are also technical elements that need to be given attention. Among them, the semantic analysis, logical reasoning, and information organization capabilities of the model, as well as the specific technical indicators such as the construction of the system knowledge base and the integration and processing of related data, are particularly important. Considering the inherent correlation between technical features, hardware resources, network services, and the learning and iteration capabilities of models have also received sufficient attention. Based on the research, analysis, and transformation results of user demands mentioned above through the Kano-QFD method, the author puts forward relevant opinions or suggestions on the construction and optimization of the online game question answering systems in terms of model selection, deployment, system support, system knowledge base construction, data integration and processing, emotional and multimodal interaction performance optimization and model iteration improvement, system business side construction, etc., in order to continuously ensure the ability of the online game customer service system to meet user demands and continuously improve user experience.
查看全文：	预览  下载(下载需要进行登录)