- If you plan to run each service separately, we recommend creating an Anaconda environment or Virtual environments per service to manager its dependencies separately
- Here we describe the implemented services
| Service Name | localhost Port | Purpose | Type | docker | Status |
|---|---|---|---|---|---|
| Solver | 5000 | Core node solving semantic table annotation tasks | Solver | Solver/5000 | |
| Autocorrect | 5005 | Fix typoes using different strategies that are enabled in config | Util | Autocorrect/5000 | |
| Wikidata_Proxy | 5007 | encapsulates the lookup up and SPARQL query endpoint for Wikidata | Util | Wikidata_Endpoint_API/5000 | |
| DBpedia_Proxy | 5003 | encapsulates the lookup up and SPARQL query endpoint for DBpedia | Util | Wikidata_Lookup_API/5000 | |
| Generic_Lookup | 5010 | Wikipedia_Lookup in the pre-computed lookup.db3 | Util | Generic_Lookup/5000 |
- All services handles server errors (500), return it to Manager. Manager is a centeral point of errors
- Stand-alone service with no dependencies on other services
- Reusable across Approaches
- Examples:
- Autocorrect
- Wikidata_Proxy
- DBpedia_Proxy
- Fat service
- Has Solve API, that recieved a table and solves it in terms of CEA, CTA and CPA
- Solve API is time scoped function using a configured SLA in seconds.
- Prepare a solution dictionary of the submission results, checkpoints and audit data (if any is enabled)