摘要
随着早教机需求的日益增加,如何从纷繁复杂的在线评论中找出有用的在线评论,了解顾客需求和关注点成为一个非常有价值的问题。采用logistic回归模型,在控制其他因素的影响下,提炼出有用的特征词,并在此基础上运用词向量模型深入研究消费者在使用这些特征词时关注的角度与内容。研究结果表明:包含对话、联网和语音等特征词的评论正向影响评论有用性。消费者在关注人机对话功能时,对英汉翻译更为在意;关注早教机联网时,侧重于关注联网的成功率与便捷性;而在语音操作方面,则主要关注语音通话、语音互动等语音具体展现形式。这些发现对顾客优化购买决策以及厂商和电商平台了解并满足顾客需求有着重要的指导意义。
With the increasing demand for early education machines, how to find useful online comments from numerous and complicated online comments and finding customer needs and concerns has become a highly valuable issue. Logistic regression model is used to extract useful characteristic words under the influence of controlling other factors. On this basis, word vector model is used to study the angle and content that consumers pay attention to when using these characteristic words. The results show that comments with characteristic words such as conversation, networking, and voice, positively affect comment usefulness. Consumers pay more attention to English-Chinese translation when heeding man-machine conversation function; they take note of success rate and convenience of networking when focusing on networking of morning teachers and computers; in terms of voice operation, they highlightvoice communication, voice interaction and other specific forms of voice presentation. These findings have implications for customers to optimize purchase decisions and for manufacturers and e-commerce platforms to understand and meet customer needs.
作者
朱振涛
张志威
ZHU Zhen-tao;ZHANG Zhi-wei(School of Economics and Management,Nanjing Institute of Technology,Nanjing 211167,China)
出处
《南京工程学院学报(社会科学版)》
2018年第4期39-46,共8页
Journal of Nanjing Institute of Technology:Social Science Edition
基金
国家自然科学基金项目(71471084)
国家自然科学基金青年基金项目(71402071)
江苏高校哲学社会科学基金项目(2018SJA0395)
国家自然科学基金(71571099)
江苏省高校哲学社会科学优秀创新团队建设项目(2017ZSTD025)
南京工程学院校级科研基金项目(KJ201306)