question

Kirill Milinevskiy avatar image
1 0 0 4

Data Fustion API /entity/analize/connect response

Is there any way we can receive more info about some parameters used in data fusion API?

May be any document with the description or contact person.

"entities": [
  {
    ...
    "hasExternalLink": false,
    "internalDocument": true,
    "score": 5.911617279052734,
    "subType": "",
    "isImageMap": false,
    "contextFields": {},
    "timestamp": "2016-12-07T07:57:46Z"
  }
]
...
"paths": [
  {
    "predicates": [
      {
        "IN_http://rdf.entagen.com/context/rdf-connector/graph-feed-supply-chain```http://permid.org/sc/supplierOf_uri": [
          {
            ...  
            "score": 0.96
	    ...
          }
      }
    ]
  }
]
...
"matchingEntityUris": [],
"resumptionTokens": [],
...
"pathTimestamps": {
  "http://permid.org/1-4297034757|http://permid.org/1-4295861160|IN_http://rdf.entagen.com/context/rdf-connector/graph-feed-supply-chain```http://permid.org/sc/supplierOf_uri": "2016-12-03T06:26:20Z"
}

We get All of these parameters in response of /entity/analize/connect.

I don't have questions about entities, paths and predicates, but the other fields confuse me.

I have hypothesis but I would like to have a clear understanding of API.

Thank you in advance.

apidata-fusion
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

1 Answer

ola.bildtsen avatar image
21 0 0 1

Most of the fields are documented in the API documentation found at 'datafusion/api/#!/entity/analyzeConnect'. Click on the 'Model' link under the 'Response Class' heading. Please let us know if that documentation is insufficient.

I see 'matchingEntityUris' is not documented there -- I will enter a ticket for that. That response property is a convenience listing of the URIs of all the matching entities for the paths found on this response.

1 comment
10 |1500

Up to 2 attachments (including images) can be used with a maximum of 512.0 KiB each and 1.0 MiB total.

Do you have more information about the "score" parameter. According the description it means relevancy score for searched items (higher is more relevant).

When we request connections and limit them with 5 rows per level, does it mean score was analyzed among all connections for entities?

This question is also connected with prioritization. Is there any? Or do we receive connection in random order (which means most relevant connection can be returned after several requests)?