联系我们

论文查询结果	返回搜索

论文编号：	15419
作者编号：	2120223830
上传时间：	2025/6/11 0:03:51
中文题目：	基于用户评论挖掘的新能源汽车智能驾驶市场需求研究
英文题目：	Research on Market Demand for Intelligent Driving of New Energy Vehicles Based on User Review Mining
指导老师：	王芳教授
中文关键字：	情感分析；主题分析；新能源汽车；智能驾驶；用户需求；竞争情报
英文关键字：	Sentiment Analysis; Topic Analysis; New Energy Vehicles; Intelligent Driving; User Demand; Competitive Intelligence
中文摘要：	通过挖掘用户评论来了解市场需求，是竞争情报领域的重要研究课题。近年来，随着全球能源结构的转型和碳中和目标的推进，新能源汽车市场迅速扩张，智能驾驶技术成为汽车产业变革的关键因素。深入分析市场中的用户需求，将为车企制定产品策略和明确市场定位提供重要依据。然而，现有研究在系统性分析用户需求方面仍显不足，尤其缺乏对用户情感和具体需求的细致剖析。为此，本研究聚焦新能源汽车智能驾驶领域，通过大数据分析手段深入挖掘用户需求，旨在为车企提供科学的产品优化建议，同时为市场竞争情报分析提供框架。 研究数据来源于汽车垂直媒体平台“懂车帝”，共采集了341款新能源汽车的1.3万余条用户评论数据。通过八爪鱼爬虫技术获取数据后，进行了数据清洗、分词、去重等预处理。运用TF-IDF关键词抽取、LDA主题分类和k-means文本聚类等自然语言处理技术，归纳出新能源汽车领域的10个核心主题，其中包括智能驾驶。进一步采用方面级分析方法，对智能驾驶相关评论进行更精细的主题词表和方面词表构建，细分为8个具体方面。通过短句切分确保每个短句精准对应具体方面，并利用微调后的BERT模型对各主题和方面下的短句情感进行分类分析。本研究受KANO模型的启发，以期望确认理论、感知价值理论为基础，构建了需求模型，综合用户关注度和满意度，挖掘影响市场中用户体验的关键因素。 研究显示，智能驾驶领域的用户需求在不同方面存在显著差异，主要集中在安全性、驾驶辅助功能和交互体验等方面。针对不同城市级别和价格区间的用户需求差异，提出差异化市场策略：一线城市强调智能化体验，三四线城市侧重性价比；低价位车型优化续航和质量，高价位车型强化智能驾驶能力和高端品牌服务。同时，建议加强充电网络布局，提升用户体验。 本研究不仅为新能源汽车领域的产品创新提供了数据支持，也为智能驾驶技术的发展方向提供了重要参考。通过精准识别用户需求，车企能够优化产品设计，提升市场竞争力。此外，本研究为市场竞争情报分析提供了情报驱动的分析框架，有助于车企更好地理解用户需求，制定精准的市场策略。未来研究可结合实地用户调研和A/B测试，进一步验证结论的可行性，并提升数据驱动的市场分析能力。
英文摘要：	Mining user reviews to understand market demand is a critical research topic in the field of competitive intelligence. In recent years, with the global transition in energy structures and the advancement of carbon neutrality goals, the new energy vehicle (NEV) market has expanded rapidly. Intelligent driving technology has emerged as a key driver of transformation in the automotive industry. Analyzing user demand in this market plays a vital role in supporting automotive enterprises in formulating product strategies and determining market positioning. However, existing research remains limited in its systematic analysis of user needs, particularly in terms of detailed insights into user sentiments and specific requirements. To address this gap, this study focuses on the intelligent driving domain within the NEV sector, employing big data analytics to deeply mine user demands. The objective is to provide scientifically grounded recommendations for product optimization and to offer an analytical framework for competitive intelligence in the market. The research data were sourced from the vertical automotive media platform "Dong Che Di," with a total of over 13,000 user reviews collected from 341 NEV models. After data acquisition via Octopus web scraping technology, a series of preprocessing steps, including data cleaning, word segmentation, and deduplication, were conducted. Natural language processing (NLP) techniques, such as TF-IDF keyword extraction, Latent Dirichlet Allocation (LDA) topic modeling, and k-means text clustering, were applied to identify ten core themes in the NEV domain, including intelligent driving. Further aspect-level analysis was employed to construct a refined topic and aspect lexicon for intelligent driving-related reviews, which were categorized into eight specific aspects. To ensure precise aspect alignment, sentence segmentation techniques were applied so that each sentence accurately corresponds to a particular aspect. A fine-tuned BERT model was then utilized to perform sentiment classification on short sentences under different topics and aspects. This study integrates the KANO model with the Expectation Confirmation Theory (ECT) and Perceived Value Theory (PVT) to develop a user demand model. By incorporating both user attention and satisfaction, the study identifies key factors affecting user experience. The findings reveal significant differences in user demands across different aspects of intelligent driving, primarily focusing on safety, driving assistance functions, and interactive experiences. Specifically, fundamental user needs include driving assistance and intelligent interaction, while expected needs involve parking assistance. Moreover, connectivity and driving experience lie at the boundary between fundamental and expected needs, whereas driving enjoyment represents an excitement need, with autonomous driving and safety positioned at the intersection of expected and excitement needs. To address user demand variations across city tiers and price segments, the study proposes differentiated market strategies: first-tier cities should emphasize intelligent experiences, while third- and fourth-tier cities should prioritize cost-effectiveness. For low-priced models, improvements in battery range and quality are recommended, whereas high-end models should enhance intelligent driving capabilities and premium brand services. Additionally, strengthening the charging network infrastructure is suggested to further improve user experience. This study not only provides data-driven support for product innovation in the NEV sector but also offers valuable insights into the development direction of intelligent driving technology. By accurately identifying user needs, automakers can optimize product design and enhance market competitiveness. Furthermore, the study introduces an intelligence-driven analytical framework for market competitive intelligence, assisting automakers in better understanding user demands and formulating precise market strategies. Future research may incorporate field user studies and A/B testing to further validate the conclusions and enhance data-driven market analysis capabilities.
查看全文：	预览  下载(下载需要进行登录)