trade-intelligence-graph is a Windows app for trade fraud analysis. It uses graph analysis to help find linked companies, repeated trade patterns, and risky networks.
The app helps you:
- Detect groups of connected entities
- Review central players in a network
- Track risk spread across related records
- Spot carousel fraud patterns
- Access results through a GraphQL API
It is built for users who need a clear view of complex trade links without reading raw data by hand.
Before you start, make sure your PC has:
- Windows 10 or Windows 11
- At least 8 GB of RAM
- 2 GB of free disk space
- Internet access for the first setup
- A modern browser for the web interface
- Access to the downloaded app package from the repository page
If your machine handles larger datasets, 16 GB of RAM works better.
Visit this page to download the app:
If the page contains a release file, download that file. If it contains the full project source, use the latest Windows build or packaged app listed there.
- Open the download link in your browser.
- Find the latest file or release package.
- Download it to your PC.
- If the file is in a ZIP folder, right-click it and choose Extract All.
- Open the extracted folder.
- Look for the main app file or setup file.
- Double-click the file to start the app.
- If Windows asks for permission, choose Yes.
If the app opens in a browser window, keep that window open while you work.
After setup, use the app like this:
- Start the app from the file you opened.
- Wait for the main screen to load.
- Load your trade data set.
- Run the analysis you need.
- Review the graph view and risk results.
For larger files, the first run may take a short time.
Find clusters of records that connect to each other more than to the rest of the network. This helps show possible fraud groups.
See which nodes matter most in the network. This helps identify traders, brokers, or firms that sit in the middle of many links.
Follow risk from one record to related records. This helps show how one bad link can affect a larger set of entities.
Look for repeated movement patterns that may point to circular trade activity or fake trade flow.
Use the API to query the network data in a structured way. This helps if you need to connect the app to another system.
Import the trade records you want to review. The app works best when your data includes names, dates, locations, amounts, and linked entities.
The app turns your records into a graph. Each entity becomes a node, and each connection becomes an edge.
Choose a method such as community detection, PageRank, or risk spread analysis.
Check the graph layout, node scores, and connected groups. Look for repeated paths, tight clusters, and high-risk nodes.
Use the output for review, case notes, or follow-up work.
The app works best with trade data that includes:
- Import and export records
- Company and trader names
- Country codes
- Product codes
- Invoice values
- Dates and time ranges
- Shipment routes
- Linked parties
Clean data gives better results. If your file has duplicate names or missing fields, review it before analysis.
Use community detection to show clusters that may belong to the same fraud ring.
Use centrality scores to see which people or firms carry the most weight in the network.
Use risk propagation to see how suspicious activity spreads across related records.
Use carousel fraud checks to look for loops, repeated handoffs, and trade paths that circle back.
The app includes a GraphQL API for structured queries.
Common uses:
- Pull node data
- Fetch connections
- Filter by score or label
- Query analysis results
- Feed outputs into another review tool
If you use the API, keep your queries focused so results stay easy to read.
- A customs team wants to find firms that trade through the same hidden network
- An analyst wants to rank the most connected importers in a case
- A review team wants to see where risk spreads after one suspicious shipment
- An investigator wants to check for repeated circular trade routes
- A data team wants a graph view instead of flat tables
- Start with one case or one trade set
- Use clear naming in your source data
- Remove obvious duplicates before import
- Check both direct links and second-degree links
- Compare graph results with known case facts
- Save your analysis after each run
- Check that the file finished downloading
- Extract the ZIP file if needed
- Right-click the app and choose Run as administrator
- Make sure Windows did not block the file
- Wait for large data to load
- Refresh the page if the app uses a browser
- Close and reopen the app
- Try a smaller data set first
- Use a smaller set of records
- Close other heavy apps
- Make sure you have enough memory free
- Check column names in your source file
- Make sure key fields are filled in
- Confirm that linked records use the same format
Repository: trade-intelligence-graph
Description: Graph-based network analysis for trade fraud ring detection. Community detection, centrality analysis, risk propagation, carousel fraud identification. GraphQL API.
Topics: community-detection, customs, fraud-detection, graph-analysis, graphql, neo4j, networkx, pagerank, python, trade-intelligence
If you need the app file again, use this link: