海南大学学报(社会科学版)

随着自然语言处理等人工智能技术的发展，作为其最主要应用的语义识别技术也从以往局限于单一模态的信息识别拓展到跨越模态异构鸿沟的多模态语义融合识别。这一技术突破也促进了诸多领域有效发挥自然语言处理的优势，通过开展多模态信息识别来提升本领域各项工作的效率。然而作为我国国民经济战略性支柱产业的旅游业，特别是在旅游安全领域，丰富的互联网社会语言资源有待深入挖掘，多模态语义识别等自然语言处理先进技术尚未得到切实转化。随着社会发展和技术进步，这些多模态社会语言资源将成为探究旅游安全事件发展规律的重要数据资源，基于多模态语义识别的旅游安全预警系统构建将成为旅游业发展的迫切需求。

With the development of artificial intelligence technologies such as the natural language processing system, semantic recognition technology has expanded from the information recognition of the single modality to the multi-modal semantic fusion recognition that spans modal heterogeneity. This technological breakthrough has promoted many fields to effectively exploit the advantages of natural language processing,enabling them to improve the work efficiency through the multi-modal information recognition. However,rich social language resources on the Internet still need to be further explored in tourism, a strategic pillar industry of China's national economy especially in the field of tourism security, which await the transformation and application of such advanced natural language processing technologies as multi-modal semantic recognition.With the social development and technological progress, these multi-modal social language resources will become the important data resource for exploring the development rules of tourism security events. The construction of tourism security early warning system based on the multi-modal semantic recognition will become an urgent demand for tourism development.