[Objective/Significance] Through fine-grained emotion analysis of domestic and foreign tourists’ comments on museum services, this paper explores tourists’ needs and preferences, and compares the differences in domestic and foreign tourists’ image perception and satisfaction towards domestic history museums so as to provide references for museum managers to formulate more targeted service strategies. [Method/Process] Firstly, it collected online reviews of museums from both domestic and foreign tourists, and extracted attribute words of museum services and attribute-level statements. Then, by fine-tuning multiple large language models and comparing the effectiveness in extracting museum users’ fine-grained reviews, it identified GPT-3 and Llama2 as the optimal classification effect for Chinese and foreign reviews, respectively. Then it employed the optimal large language model to conduct fine-grained sentiment analysis of attribute-level statements. Finally, based on the results, it analyzed the satisfaction and differences between the two groups. [Results/Conclusion] The categories of museum services reviewed by domestic tourists include cultural creation, service facilities, guided explanation, museum staff, ticket security, online service. Foreign tourists, on the other hand, focus on guided explanation, in-museum service, ticket security, online service, and shopping. The analysis reveals that both domestic and foreign tourists show the highest levels of attention and satisfaction with guided explanation. Domestic tourists have the lowest satisfaction on the staff service, while foreign tourists express the least satisfaction with the ticket security. It summarizes the key factors influencing tourist satisfaction in each service category.
[1] 王棋壹, 雷青, 康祖怡.基于“故宫文创”的产品价值感知与消费需求影响研究[J]. 包装工程, 2023, 44(14): 224-234. (WANG Q Y, LEI Q, KANG Z Y. Product value perception and consumption demand influence based on the “cultural and creative products of the palace museum” [J]. Journal of packaging engineering, 2023, 44(14): 224-234.)
[2] 瞭望周刊. 中国博物馆区域发展指数报告（2022年）[EB/OL]. [2024-01-25]. https://www.xhinst.net/ggzc/research/public-policy/20220518/29264_%E4%B8%AD%E5%9B%BD%E5%8D%9A%E7%89%A9%E9%A6%86%E5%8C%BA%E5%9F%9F%E5%8F%91%E5%B1%95%E6%8C%87%E6%95%B0%E6%8A%A5%E5%91%8A2022%E5%B9%B4.html. (Outlook weekly. Museum of Chinese regional development index report (2022) [EB/OL]. [2024-01-25]. https://www.xhinst.net/ggzc/research/public-policy/20220518/29264_%E4%B8%AD%E5% 9B%BD%E5%8D%9A%E7%89%A9%E9%A6%86%E5%8C%BA%E5%9F%9F%E5%8F%91%E5%B1%95%E6%8C%87%E6%95%B0%E6%8A%A5%E5%91%8A2022%E5%B9%B4.ht ml.)
[3] 国家文物局.《关于推进博物馆改革发展的指导意见》发布[EB/OL]. [2024-01-25]. http://www.ce.cn/culture/gd/202105/28/t20210528_36597891.shtml. (The state administration of cultural heritage. The guidance about promote reform and development of the museum issued [EB/OL]. [2024-01-25]. http://www.ce.cn/culture/gd/202105/28/t20210528_36597891.shtml.)
[4] 谢法举, 刘臣, 唐莉. 在线评论情感分析研究综述[J]. 软件导刊, 2018, 17(2): 1-4, 7. (FAJU X, CHEN L, LI T. A review of sentiment analysis in online reviews [J]. Journal of software guide, 2018, 17(2): 1-4, 7.)
[5] 孙晓东, 倪荣鑫. 中国邮轮游客的产品认知、情感表达与品牌形象感知——基于在线点评的内容分析[J]. 地理研究, 2018, 37(6): 1159-1180. (XIAODONG S, RONGXIN N. Chinese cruisers’ product cognition, emotional expression and brand image perception: a web content analysis[J]. Geographical research, 2018, 37(6): 1159-1180.)
[6] 周清清, 章成志. 在线用户评论细粒度属性抽取[J]. 情报学报, 2017, 36(5): 484-493. (QINGQING Z, CHENGZHI Z. Fine-grained aspect extraction from online customer reviews [J]. Journal of the China Society for Scientific and Technical Information, 2017, 36(5): 484-493.)
[7] 郑淞尹, 王萍, 丁恒, 等. 基于方面级情感分析的博物馆数字化服务用户体验研究[J]. 情报科学, 2022, 40(4): 171-178. (SONGYIN Z, PING W, HENG D, et al. Exploring factors influencing user experience of museum digital services via aspect-level sentiment analysis[J]. Information science, 2022, 40(4): 171-178.)
[8] 张建华. 基于LDA和词性句法规则的用户评论情感分析研究[D]. 南宁: 广西大学, 2014. (JIAN HUA Z. Research on sentiment analysis of user comments based on LDA and part-of-speech syntax rules [D]. Nanning: Guangxi University, 2014.)
[9] TITOV I, MCDONALD R. Modeling online reviews with multi-grain topic models[C]// Proceedings of the 17th international conference on World Wide Web. New York: ACM, 2008: 111-120.
[10] JAKOB N, GUREVYCH I. Extracting opinion targets in a single- and cross-domain setting with conditional random fields[C]//Proceedings of conference on empirical methods in natural language processing. Massachusetts: Association for Computational Linguistics, 2010: 1035-1045.
[11] 徐冰, 赵铁军, 王山雨, 等. 基于浅层句法特征的评价对象抽取研究[J]. 自动化学报, 2011, 37(10): 1241-1247. (BING X, TIEJUN Z, SHANYU W, et al. Extraction of opinion targets based on shallow parsing features[J]. Acta automatica sinica, 2011, 37(10): 1241-1247.)
[12] AAGHAR M Z, AHMAD S, QASIM M, et al. SentiHealth: creating health-related sentiment lexicon using hybrid approach[J]. SpringerPlus, 2016, 5:1139.
[13] 宗宇, 方朝阳, 吴波. 面向在线评论的领域情感词典的自动构建[J]. 现代计算机, 2021(18): 79-84. (YU Z, ZHAOYANG F, BO W. Automatic construction of a domain emotion dictionary for online reviews [J]. Modern computer, 2021(18): 79-84.)
[14] 巴桑卓玛, 李苗苗, 高定国. 基于词向量的藏文情感词典的构建方法研究[J]. 电子技术与软件工程, 2017(20): 132-134. (ZHUOMA B, MIAOMIAO L, DINGGUO G. Research on the construction method of tibetan emotion dictionary based on word vector [J]. Electronic technology and software engineering, 2017(20): 132-134.)
[15] 柳位平, 朱艳辉, 栗春亮, 等. 中文基础情感词词典构建方法研究[J]. 计算机应用, 2009, 29(10): 2875-2877. (WEIPING L, YANHUI Z, LICHUN L, et al. Research on building Chinese basic semantic lexicon: research on building Chinese basic semantic lexicon[J]. Journal of computer applications, 2009, 29(10): 2875-2877.)
[16] Turney P D, Littman M L. Unsupervised learning of semantic orientation from a hundred-billion-word corpus[J]. arXiv preprint cs/0212012, 2002.
[17] 单睿康, 江涛, 张婷婷, 等. 基于改进SO-PMI的藏文情感词典自动构建方法[J]. 高原科学研究, 2022, 6(1): 104-112. (RUIKANG S, TAO J, TINGTING Z, et al. Study on the method of automatic construction of tibetan sentiment lexicon based on the improved SO-PMI [J]. Plateau science research, 2022, 6(1): 104-112.)
[18] 马子洲. 基于改良SO-PMI算法的在线评论情感倾向性分析研究[D]. 大连: 东北财经大学, 2018. (ZIZHOU M. Research on sentiment tendency analysis of online comments based on improved SO-PMI algorithm [D]. Dalian: Dongbei University of Finance and Economics, 2018.)
[19] MULYO B M, WIDYANTORO D H. Aspect-based sentiment analysis approach with CNN[C]//5th international conference on electrical engineering, computer science and informatics. Malang: Universitas Muhammadiyah Malang, 2018: 142-147.
[20] JIANG M, ZHANG W, ZHANG M, et al. An LSTM-CNN attention approach for aspect-level sentiment classification[J]. Journal of computational methods in sciences and engineering, 2019, 19(4): 1-10.
[21] DEVLIN J, CHANG M W, LEE K, et al. Bert: pre-training of deep bidirectional transformers for language understanding[J]. ArXiv preprint arXiv:1810.04805, 2018.
[22] JINGFENG Y, HONGYE J, RUIXIANG T, et al. Harnessing the power of llms in practice: a survey on chatgpt and beyond[J]. ArXiv preprint arXiv:2304.13712, 2023.
[23] OpenAI. Models-GPT-3[EB/OL]. [2024-01-25]. https://platform.openai.com/docs/models/gpt-3.
[24] TOUVRON H, MARTIN L, STONE K, et al. Llama 2: Open foundation and fine-tuned chat models. ArXiv preprint arXiv:2307.09288, 2023.
[25] MUNIKAR M, SHAKYA S, SHRESTHA A. Fine-grained sentiment classification using BERT[C]//2019 Artificial Intelligence for Transforming Business and Society (AITB). Kathmandu, Nepal: IEEE, 2019: 1-5.
[26] RAY A. An ensemble-based hotel recommender system using sentiment analysis and aspect categorization of hotel reviews[J]. Applied soft computing, 2021, 98(1): 106935.
[27] 王娟, 史吉祥. 河南博物院观众满意度调查报告[J]. 中原文物, 2005(2): 91-96. (JUAN W, JIXIANG S. Investigation of satisfaction degree of the audience of Henan museum [J]. Cultural relics of central China, 2005(2): 91-96.)
[28] KOWALSKA N, OSTRĘGA A. Using SERVQUAL method to assess tourist service quality by the example of the silesian museum established on the post-mining area[J]. Land, 2020, 9(9): 333.
[29] 彭秋平, 吴思洋. 公共图书馆用户在线口碑情感分析——以副省级市图书馆大众点评网评论为中心[J]. 图书馆论坛, 2021, 41(6): 141-149. (QIUPING P, SIYANG W. Users’ evaluation of public libraries based on sentiment analysis of comments from Dianping.com [J]. Library tribune, 2021, 41(6): 141-149.)
[30] 钱成. 基于网络评论的旅游目的地游客满意度分析[D]. 武汉: 中南财经政法大学, 2019. (CHENG Q. Analysis of tourist satisfaction in tourist destinations based on internet reviews [D]. Wuhan: Zhongnan University of Economics and Law, 2019.)
[31] COSTA S, MORO S, RITA P, et al. Customer experience through online reviews from TripAdvisor: the case of Orlando theme parks[J]. International journal of technology marketing, 2023, 17(1): 48-77.
[32] MAREK N. Evaluating a museum as a tourist product using the servqual method[J]. Museum management and curatorship, 2005, 20(3): 235-250.
[33] YAOHUA S. Contemplating museums’ service failure: extracting the service quality dimensions of museums from negative on-line reviews[J]. Tourism management, 2018(69): 214-222.
[34] 毛太田, 吴宇豪, 黄文佳. 自治区民族博物馆的游客在线点评内容挖掘与情感分析[J]. 经济地理, 2023, 43(8): 229-236. (TAITIAN M, YUHAO W, WENJIA H. Content mining and sentiment analysis of online comments for ethnic museum in autonomous regions[J]. Journal of economic geography, 2023, 43(8): 229-236.)
[35] yeyupiaoling. PPASR[EB/OL]. (2023-01-23) [2023-04-10]. https://github.com/yeyupiaoling/PPASR.
[36] PONTIKI M, PAVLOPOULOS J. SemEval-2014 Task 4: Aspect Based Sentiment Analysis [C]// Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014). Dublin: Association for Computational Linguistics, 2014: 27-35.
[37] 张雪涛, 孙蒙, 王金双. 基于操作码的安卓恶意代码多粒度快速检测方法[J]. 网络与信息安全学报, 2019, 5(6): 85-94. (XUETAO Z, MENG S, JINGSHUANG W. Multi-granularity android malware fast detection based on opcode [J]. Journal of network and information security, 2019, 5(6): 85-94.)
