基于多模态数据语义融合的旅游在线评论有用性识别研究

Research on Usefulness Recognition of Tourism Online Reviews Based on Multimodal Data Semantic Fusion

为有效识别旅游产品在线评论中多模态数据对在线评论感知有用性的影响因素,探究基于用户生成内容的在线旅游产品优化方法,从数据融合分析角度出发,对旅游产品在线评论中的多模态数据进行特征融合。以旅游产品的真实在线评论数据作为研究对象,进行描述性统计分析,同时使用机器学习和深度学习方法,进行文本向量嵌入与图片内容识别,融合图文特征向量,构建多模态在线评论有用性分类模型,进行模型测试。实验结果表明,与仅包含文本或仅包含图片的单模态评论相比,图文结合的多模态评论能够更好地进行在线评论有用性预测,结合评论激励机制,提高在线评论质量,能够充分发挥用户生成内容潜在价值,为产品提供者提供优化思路,为产品消费者提供决策支持。

In order to identify the influencing factors of multimodal data in online reviews of tourism products effectively,the online tourism product optimization method based on user-generated content is explored. From the perspective of data fusion analysis, we apply feature fusion methods to multimodal data in the online review of tourism products. Based on real online review data of tourism products, first, we performed a descriptive statistical analysis. Then, we utilized machine learning and deep learning methods, performed text vector embedding, image content recognition, and fusion of graphic feature vectors. Finally, we constructed multimodal online comment usefulness to classify models for model testing. The experimental results show that compared with single-modal comments containing text only or images only, the multimodal comments combining images and texts can predict the usefulness of online reviews better, and combine the incentive mechanism of comments better to improve the quality of online comments. In addition, the multimodal comments combining images and texts can be used to leverage the potential value of user-generated content, provide optimization ideas for product providers, and provide decision support for product consumers.