In the modern age of rapid global information circulation, public opinion on social media reflects differences in people's opinions on a certain issue. These opinions can include emotions, likes and dislikes, etc. Once an incident happens, many social media gather a lot of discussions. Such situations let us to analyze the public's perception and preferences through public opinion analysis. Moreover, people cannot live without food, thus our team takes gourmet restaurants as an entry point to develop a public opinion analysis platform for gourmet restaurants. This study uses Python web crawler technology to capture data from public websites, such as PTT, Dcard, and Google Maps user comments. Then we extract public opinion data in the last year, clean and sort the data, and combine the sentiment analysis tool SnowNLP, analyze the positive and negative emotions of users’ comments in various restaurants. And we screen out the top popular restaurants in the form of scores, build a website entry function so that users can query and view the names, ratings of various restaurants, and emotional scores sorted out by our team. We hope this paper can be used as a reference to assist individual decision-making.