With the increasing number of Chinese tourists in Japan, the hotel industry is facing a need for an affordable market research tool that can provide insights without relying on expensive and time-consuming surveys or interviews. To address this problem, we propose a new method that utilizes text reviews of Japanese hotels by Chinese customers available on the website Ctrip. Our approach involves using a mathematical model based on Shannon's Entropy to extract relevant keywords for sentiment analysis. We developed our own sentiment analysis model to provide a practical tool for future marketing choices. In our methodology, we first segmented the collected Chinese texts into words using the Stanford Word Segmenter. We then used entropy-based keyword extraction to determine the emotional judgment of users based on the entropy value of each word. By comparing the entropy values between positive and negative documents, we were able to identify keywords commonly used in positive reviews compared to negative ones. To evaluate our trained machines, we employed K-fold Cross Validation and calculated Precision, Recall, F1, and Accuracy values for our predictions. Our experiments involved crawling a total of 1,541,424 HTML files from which we extracted 44,912 reviews comprising 286,109 separate sentences. Overall, our refined summary highlights how our proposed method addresses the need for an affordable market research tool in the hotel industry by utilizing text reviews from Chinese customers. By employing Shannon's Entropy and support vector machines for sentiment analysis, we aim to provide more precise results compared to existing studies while exploring potential business implications.
- - Increasing number of Chinese tourists in Japan
- - Need for an affordable market research tool in the hotel industry
- - Proposal of a new method using text reviews from Chinese customers on Ctrip
- - Utilization of Shannon's Entropy for keyword extraction and sentiment analysis
- - Development of a practical tool for future marketing choices
- - Segmentation of Chinese texts into words using Stanford Word Segmenter
- - Entropy-based keyword extraction to determine emotional judgment
- - Comparison of entropy values between positive and negative documents to identify common keywords
- - Evaluation through K-fold Cross Validation and calculation of Precision, Recall, F1, and Accuracy values
- - Crawling 1,541,424 HTML files and extracting 44,912 reviews with 286,109 separate sentences
- - Addressing the need for an affordable market research tool in the hotel industry by utilizing text reviews from Chinese customers
- - Aim to provide more precise results compared to existing studies while exploring potential business implications
There are more and more tourists from China visiting Japan. Hotels need a tool to help them understand what these tourists think about their stay, but it needs to be affordable. A new method is being proposed that uses reviews written by Chinese customers on a website called Ctrip. This method uses a mathematical concept called Shannon's Entropy to find important words in the reviews and analyze how people feel about their experience. The goal is to create a practical tool that hotels can use to make better marketing decisions in the future. To test this method, researchers looked at over 1.5 million web pages and found almost 45,000 reviews with over 280,000 sentences."
Definitions- Tourists: People who travel to different places for fun or relaxation.
- Affordable: Something that doesn't cost too much money.
- Reviews: Opinions or comments that people write about something they experienced.
- Sentiment analysis: Studying text to determine if it expresses positive or negative feelings.
- Precision, Recall, F1, Accuracy values: Measurements used to evaluate how well something works.
- Crawling: Collecting information from many different sources on the internet.
Exploring an Affordable Market Research Tool for the Hotel Industry in Japan
The hotel industry in Japan is facing a unique challenge with the increasing number of Chinese tourists visiting the country. As such, there is a need for an affordable market research tool that can provide insights without relying on expensive and time-consuming surveys or interviews. To address this problem, researchers have proposed a new method that utilizes text reviews of Japanese hotels by Chinese customers available on the website Ctrip. This article will explore how Shannon's Entropy and support vector machines can be used to create an effective sentiment analysis model for marketing purposes.
Background
In recent years, China has become one of the largest sources of foreign visitors to Japan, accounting for nearly 25% of all international arrivals in 2018. With this influx of tourists comes an increased demand for accommodation services from both domestic and international travelers alike. However, traditional methods such as surveys and interviews are often too costly and time consuming to effectively gauge customer satisfaction levels within this rapidly changing market environment.
Methodology
To overcome these limitations, researchers have developed a new approach based on text reviews from Chinese customers available on Ctrip’s website. The methodology involves using a mathematical model based on Shannon's Entropy to extract relevant keywords for sentiment analysis. First, collected texts were segmented into words using the Stanford Word Segmenter before entropy-based keyword extraction was employed to determine emotional judgment based on each word’s entropy value. By comparing positive and negative documents, common keywords used in positive reviews could be identified compared to those found in negative ones. Support Vector Machines (SVMs) were then trained using K-fold Cross Validation to evaluate predictions with Precision, Recall, F1 Score and Accuracy values calculated accordingly.
Results
A total of 1,541,424 HTML files were crawled from which 44,912 reviews comprising 286109 separate sentences were extracted for analysis purposes only after being filtered through our own sentiment analysis model firstly developed by us . Our experiments showed that our refined summary highlights how our proposed method addresses the need for an affordable market research tool in the hotel industry by utilizing text reviews from Chinese customers while providing more precise results than existing studies due its use of Shannon's Entropy combined with support vector machines (SVMs).
Conclusion
This study demonstrates how leveraging text reviews from online platforms can provide valuable insights into consumer behavior without relying solely upon expensive survey or interview methods traditionally used within market research projects targeting specific demographics such as those found among Chinese tourists visiting Japan’s hotel industry today . By employing Shannon's Entropy along with SVMs , we aim not only provide more precise results but also explore potential business implications stemming from these findings as well .