Tourism online comment text mining based on BiLSTM and LDA models
More and more tourists are accustomed to using online comment information from major tourism websites as an im-portant reference for travel decisions.Therefore,how to guide scientific decision-making in scenic areas by mining online com-ments has become an urgent problem that scenic area managers need to solve.Firstly,we crawl the online comment data of tourists from various tourism websites.Then,based on the BiLSTM model,we analyze the sentiment tendency of comment texts.We use the LDA model to cluster and extract feature keywords for positive and negative comment texts,in order to explore the advantages and disadvantages of the dimensions that tourists are most concerned about.We use data-driven approaches to provide data support for scientific decision-making and precise improvement in scenic area management,thereby improving the service quality and tourist satisfaction of scenic areas.The experimental results show that the text vector of BiLSTM model performs well in sentiment classifi-cation tasks,and the LDA topic model has strong expressive ability,which can effectively explore the existing advantages and po-tential shortcomings of scenic spots.
万方数据
