As promoters of tourism,university students play a crucial role in the industry.Addressing the significant variability in review data for tourist attractions around Wenzhou University Town,and the issues of imprecise ...As promoters of tourism,university students play a crucial role in the industry.Addressing the significant variability in review data for tourist attractions around Wenzhou University Town,and the issues of imprecise sentiment extraction and single influencing factors in traditional language sentiment analysis models,this study proposes a method that combines algorithms and considers multiple factors for sentiment analysis.First,data is collected using Python technology and analyzed based on seasonal time factors.The TF-IDF algorithm is then used to evaluate keywords in the reviews,followed by the TextRank algorithm to calculate the weight of each word,obtaining summary terms.Finally,the results of the TF-IDF and TextRank algorithms are combined,and sentiment analysis is conducted using the SnowNLP library in Python.Compared to traditional sentiment analysis,which uses only a single model or considers only time factors,this study combines the TF-IDF and TextRank algorithms and incorporates time factors,thereby expanding the range of sentiment influencing factors.This approach results in more accurate and rational sentiment analysis outcomes.展开更多
基金support from the Innovation and Entrepreneurship Training Program for College Students at Wenzhou Business College for this research.The project title is"A Study on the Evaluation of Tourist Attractions Around Wenzhou University Town Based on Big Data Analysis,"and the project number is 202313637017.
文摘As promoters of tourism,university students play a crucial role in the industry.Addressing the significant variability in review data for tourist attractions around Wenzhou University Town,and the issues of imprecise sentiment extraction and single influencing factors in traditional language sentiment analysis models,this study proposes a method that combines algorithms and considers multiple factors for sentiment analysis.First,data is collected using Python technology and analyzed based on seasonal time factors.The TF-IDF algorithm is then used to evaluate keywords in the reviews,followed by the TextRank algorithm to calculate the weight of each word,obtaining summary terms.Finally,the results of the TF-IDF and TextRank algorithms are combined,and sentiment analysis is conducted using the SnowNLP library in Python.Compared to traditional sentiment analysis,which uses only a single model or considers only time factors,this study combines the TF-IDF and TextRank algorithms and incorporates time factors,thereby expanding the range of sentiment influencing factors.This approach results in more accurate and rational sentiment analysis outcomes.
