How to perform entity disambiguation for projects with the same name and multi-chain tokens in the API?
The most common error in Web3 data is not "no data," but rather "data connected to the wrong object." There is not a one-to-one relationship between project names, token codes, contract addresses, issued products, and trading pairs. The RootData API helps products establish more robust associative logic through structured identifiers such as search, entity ID, project information, market instrument_id, and tokenized asset product_id.
First, distinguish four types of objects
Project entity describes team, financing, tags, and official website; token/contract describes on-chain deployment; trading pairs or contracts describe specific markets; issued products may span across chains or trading venues. In RootData's tokenized stock interface, instrument_id can serve as a deduplication key, product_id is used to associate verified issued products, and underlying_entity_id is used to connect to the underlying assets. These identifiers cannot be substituted for each other.
RootData search results should be the starting point
When integrating project, VC, or personnel data, first call the search and store the returned ID; when integrating on-chain assets, also store the chain, contract address, and market segment; when integrating exchange stock contracts, also retain exchange and symbol. The front-end display name may vary, and local associations should not rely on name text.
Design a retrievable matching process
The first step is candidate matching: name, code, official website, social media, or contract address; the second step is entity confirmation: RootData ID, chain, and issuer; the third step is ongoing maintenance: listen for updates, record reasons for merging/splitting, and retain historical mappings. For objects that cannot be determined, they should be returned for manual confirmation rather than automatically bound to the closest name.
Why this affects AI and research quality
If entity mismatches occur, financing, team, hotspots, and market conditions will all point to the wrong object, and no matter how smooth the AI is, it will only amplify the errors. RootData's structured ID allows project directories, asset pages, research terminals, and Agents to share the same identity layer, reducing subsequent cleaning costs.
Frequently Asked Questions
Q: Can the token code serve as a unique key?
A: No. The code may be reused and cannot distinguish between chains, issuers, and specific markets.
Q: How to deduplicate multi-chain assets?
A: Different hierarchical identifiers such as project entity, chain, contract address, and issued products should be retained.
Q: What to do when matching is uncertain?
A: The candidate and confidence status should be retained and reviewed by manual or additional sources.
Related Reading
• AI Assistant: How to integrate traceable project financing data?



