一种基于景区评论的静态热词提取模型  

A static hot word extraction model based on scenic spot comments

作　　者：王大睿 张超群[2,3] 郝小芳 完颜兵 李晓翔 WANG Da-rui;ZHANG Chao-qun;HAO Xiao-fang;WANYAN Bing;LI Xiao-xiang(College of Electronic Information,Guangxi Minzu University,Nanning 530006,China;College of Artificial Intelligence,Guangxi Minzu University,Nanning 530006,China;Guangxi Key Laboratory of Hybrid Computation and IC Design Analysis,Nanning 530006,China)

机构地区：[1]广西民族大学电子信息学院,南宁530006 [2]广西民族大学人工智能学院,南宁530006 [3]广西混杂计算与集成电路设计分析重点实验室,南宁530006

出　　处：《信息技术》2024年第6期15-21,共7页Information Technology

基　　金：国家自然科学基金(62062011);广西自然科学基金项目(2019GXNSFAA185017);广西民族大学研究生科研创新项目(gxun-chxps202088,gxun-chxs2021066)。

摘　　要：热词提取对于景区的发展具有重要意义,目前热词提取方法仍存在分词效果不佳、训练模型耗费大等问题,文中提出一种基于景区评论的静态热词提取模型CRF+TBTT。该模型利用新型算法流程过滤非关键词,分析高频词和特色词,提取候选词,最后得到准确的静态热词。通过对59107条景区评论数据进行实验,结果表明,CRF+TBTT模型的性能均明显优于比较模型,对景区前20个热词提取准确率达到90%,说明该模型对静态热词提取的效果较好,有助于旅游部门对景区进行有效管理和规划。Hot word extraction is of great significance to the development of scenic spots.At present,hot word extraction methods still have problems such as poor word segmentation effect and high cost of training models,a static hot word extraction model called CRF+TBTT is proposed based on scenic comments.The model uses a new algorithm process to filter non-keywords,analyzes high-frequency words and featured words,extracts candidate words,and finally obtains accurate static hot words.The experiments based on 59107 scenic spot comments show that the performance of the CRF+TBTT model is significantly better than that of the competitors,and the accuracy rate of extracting the top 20 hot words in the scenic spot reaches 90%.These results suggest that the new model has a good effect on extracting static hot words,which can help tourism departments to effectively manage and plan scenic spots.

关 键 词：景区评论 CRF+TBTT模型 TextRank算法 TF-IDF算法 静态热词 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] F592[自动化与计算机技术—计算机科学与技术]