تحلیل احساس و کاوش عقیده از متن به روش بدون نظارت با استفاده از منطق فازی: کاربرد در تحقیقات بازار (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
تجزیه و تحلیل احساسات کاربران در حوزه بازاریابی و بهبود تجربه مشتریان نقشی اساسی دارد و به تدوین استراتژی های بازاریابی کمک می کند. این تحلیل یکی از عوامل اساسی در ارزیابی کارآیی و عملکرد رسانه های اجتماعی به عنوان ابزارهای ارتباطی است. پژوهش حاضر از لحاظ هدف، کاربردی، از لحاظ زمان، مقطعی، از لحاظ متغیر پژوهشی، کمّی و کیفی و از لحاظ طرح پژوهش، توصیفی است. محققان این پژوهش با بهره گیری از سیستم فازی و رویکرد بدون نظارت، بدون نیاز به دانش پیشین و داده های برچسب گذاری شده توییت های کاربران را دسته بندی کردند. این پیشرفت در تحلیل احساسات به محققان و صنعت گران امکان می دهد تا بدون نیاز به داده برچسب گذاری شده اطلاعات مفیدی را از نظر ها و احساسات کاربران در زمینه و بستر دلخواه خود استخراج و از آن در فرآیند تصمیم گیری های تجاری برای دستیابی به سود بیشتر استفاده کنند. محققان در پژوهش حاضر تجربه های کاربران را درباره تلفن های همراه شرکت های سامسونگ و اپل از سال 2022 تاکنون بررسی و تحلیل و آنها را به شکل مثبت، منفی و خنثی دسته بندی کردند. بدین منظور از سه ابزار تحلیل احساسات واژگان SentiWordNet، AFINN و VADER برای تعیین قطبیت توییت ها استفاده شد. نتایج دسته بندی نشان داد که میزان رضایت کاربران از تلفن های همراه شرکت سامسونگ نسبت به شرکت اپل بیشتر بوده است.Fuzzy Logic-Based Unsupervised Sentiment Analysis and Opinion Mining: Applications in Market Research
Analyzing user sentiments in marketing and enhancing customer experiences are essential for developing effective marketing strategies. This analysis is crucial for assessing the performance of social media platforms as communication tools. This research was practical in nature and cross-sectional in time, while utilizing both quantitative and qualitative variables within a descriptive research design. The study categorized user tweets without relying on prior knowledge or labeled data, employing fuzzy systems and an unsupervised approach. This advancement in sentiment analysis enabled researchers and practitioners to extract valuable insights from user opinions and emotions within their respective domains and platforms, thereby facilitating informed business decisions aimed at maximizing profitability. As a case study, this empirical research examined user experiences with Samsung and Apple mobile phones from 2022 to the present, classifying sentiments into positive, negative, and neutral categories. Three sentiment analysis tools—SentiWordNet, AFINN, and VADER—were employed to determine the polarity of the tweets. The classification results revealed a higher level of user satisfaction with Samsung mobile phones compared to Apple.IntroductionSentiment analysis plays a vital role in marketing by enabling businesses to extract emotional insights from text data, allowing for a deeper understanding of customer reactions to products and services. As social media and online platforms continue to expand rapidly, the challenge of analyzing vast amounts of textual data has intensified, highlighting the need for efficient and accurate analytical methods. This study aimed to explore and introduce an innovative approach to sentiment analysis within the context of market research and marketing strategies. By leveraging fuzzy logic—a technique adept at managing imprecise and ambiguous opinions—this research proposed a novel method for categorizing user tweets without relying on labeled data. This approach offered greater flexibility and adaptability compared to traditional methods. The primary objective was to provide managers and industry professionals with actionable insights that could inform commercial decision-making and enhance marketing strategies. Furthermore, this study addressed significant challenges associated with conventional sentiment analysis techniques, such as the time-consuming and costly nature of manual data processing and the limited availability of labeled datasets. By proposing a fuzzy logic-based system, the research aimed to overcome these limitations and offer a more efficient alternative. The central research question investigated whether a fuzzy logic approach could surpass the existing sentiment analysis methods in market studies, potentially leading to more accurate and insightful outcomes. This innovative approach has the potential to revolutionize sentiment analysis in marketing, making it more accessible and effective in understanding customer sentiments.Materials & MethodsThis study employed a comprehensive fuzzy rule-based sentiment analysis system to evaluate user opinions on Twitter. The system encompasses several detailed processes, including data collection, text preprocessing, sentiment lexicon analysis, and fuzzy classification. Data Collection: Data were collected from Twitter, focusing on tweets related to Apple and Samsung smartphones from 2022 onwards. Approximately 100 tweets for each company were selected and stored in separate CSV files. Text Preprocessing: During preprocessing, URLs and @ symbols were removed and common contractions, such as "can’t", were expanded to "cannot". Hashtags were stripped of the "#" symbol to prepare the text for analysis. These steps reduced ambiguity and enhanced sentiment interpretation. Sentiment Lexicon Analysis: Three sentiment lexicon tools—SentiWordNet, AFINN, and VADER—were utilized to calculate positive and negative scores for the tweets. These tools facilitated the labeling and analysis of sentiment. Fuzzy System: A fuzzy rule-based system with 9 proposed rules was developed to determine sentiment polarity. This unsupervised system classified sentiment based on a set of carefully defined rules. Comparison of Lexicon Tools: The performance of the sentiment analysis tools was evaluated using datasets from Samsung and Apple. Results indicated that AFINN demonstrated the highest accuracy, recall, and F1 scores among the tools, proving to be the most effective for analyzing social media posts. AFINN outperformed SentiWordNet and VADER in both precision and recall. The choice of the best lexicon tool depended on the evaluation metrics and the characteristics of the dataset.Research FindingsThis section presented a comprehensive analysis of the performance of the fuzzy rule-based sentiment analysis system, utilizing the AFINN lexicon to evaluate Twitter data related to Samsung and Apple. The sentiment analysis revealed a 60% satisfaction rate for Samsung products, while Apple products had a satisfaction rate of 50%. Negative comments accounted for 20% of tweets related to Samsung compared to 14% for Apple. Neutral comments represented 20 and 36% of Samsung- and Apple-related tweets, respectivdely. This distribution indicated that Samsung users generally expressed more positive sentiments toward the brand compared to Apple users. However, Apple received a higher percentage of neutral comments, suggesting a more nuanced and varied perception of the brand. The lower percentage of negative comments for Apple might imply that while fewer users were dissatisfied, there was a greater level of indifference or neutrality compared to Samsung. The analysis highlighted the effectiveness of the AFINN tool in sentiment classification. Compared to other sentiment analysis tools, AFINN demonstrated superior accuracy and efficiency in processing Twitter data. The results indicated that the classification of AFINN and scoring of sentiments were both reliable and consistent, reinforcing its value as a tool for social media sentiment analysis. This effectiveness was crucial for gaining accurate insights into user opinions and brand perceptions, providing a valuable resource for marketers seeking to understand and respond to consumer sentiments more effectively.Discussion of Results & ConclusionThis study highlighted the effectiveness of an unsupervised fuzzy system in accurately identifying sentiments in tweets related to Samsung and Apple, achieving accuracies of 74.44 and 77.16%, respectively. The fuzzy system operated independently of prior training data and demonstrated high precision in sentiment classification, making it a highly efficient tool for analyzing large-scale data. This approach was particularly valuable for handling the vast and diverse nature of social media data, where traditional supervised methods might fall short. Additionally, the AFINN lexicon outperformed SentiWordNet and VADER in terms of precision, recall, and F1 score. This validation underscored the effectiveness of AFINN in capturing nuanced sentiment expressions, which was crucial for accurate sentiment analysis. The findings indicated that Samsung products generally achieved a higher level of customer satisfaction compared to Apple products. This insight could be instrumental for Apple management, providing a clear indication of areas that required improvement. The sentiment analysis enabled both companies to identify strengths and weaknesses in their products and allowed for a strategic focus on positive attributes in marketing campaigns. By leveraging detailed customer feedback, businesses can gain a better understanding of market trends and more accurately predict customer behavior. The fuzzy system’s cost-effectiveness and resource efficiency further enhance its value, supporting improved managerial decision-making and strategic planning. Overall, this approach provides a robust and scalable solution for sentiment analysis, offering significant advantages over traditional methods.
تبلیغات
