How to Scrape and Analyze Holiday Tourist Spot Data with Python

Goal

Use Python to analyze National Day holiday tourist attractions that are fun, cheap, and less crowded.

Data Acquisition

Official tourism data was hard to find, so we resorted to scraping ticket information from Qunar, which reflects attraction popularity.

1. Scrape Single Page Data

Search Qunar's ticket page for "National Day tourist attractions" to obtain JSON data containing name, region, popularity, sales, price, rating, coordinates, etc.

Open the browser's developer tools (F12) to locate the data URL; the response is already JSON.

Use the requests library to send a GET request and retrieve the page data.

2. Extract Relevant Information

Inspect the JSON structure and extract fields such as id, name, star rating, score, ticket price, sales volume, region, coordinates, and description.

3. Save to Excel

After extracting the needed data, save it with pandas.

pip install xlrd pip install openpyxl pip install numpy pip install pandas

4. Batch Scraping

Identify the pagination parameter (e.g., page ) from the network requests and loop over page numbers to collect all pages.

Continue fetching until the returned item count drops to zero.

Data Analysis

With the data saved, perform several analyses using the pyecharts visualization library:

Ticket sales ranking

Revenue ranking (price × sales)

Number of attractions per province and rating level (unfinished)

Sales heatmap on a map

Recommendation of attractions

1. Ticket Sales Ranking

Create a pivot table, sort by sales, and plot a bar chart. Disney ranks first.

2. Revenue Ranking

Calculate revenue as price × sales, add it to the dataframe, sort, and visualize. Disney again leads.

4. Sales Heatmap

Use Baidu Map's open API to draw a heatmap. Apply for a Baidu Map API key (browser type) and replace the key and JSON data in the demo HTML.

5. Recommendation of Attractions

Recommend spots that have high scores, low sales, and cheap prices. The simple recommendation coefficient is:

Recommendation Score = Score / (Sales * Price) * 1000

The top‑20 results include many foreign attractions, especially in Japan, showing that domestic spots are crowded during the holiday.

Feel free to improve the algorithm and experiment with the code.

Source code: https://github.com/pig6/qunar_spider