Skip to content

Keyphrase extraction output

The relevants resource that performs keyphrase extraction returns a JSON object with this structure:

{
    "success": Boolean success flag,
    "data": {
        "content": analyzed text,
        "language": language code,
        "version": technology version info,      
        "knowledge": [],
        "topics": [],
        "mainSentences": [],
        "mainPhrases": [],
        "mainSyncons": [],
        "mainLemmas": []
    }
}

Tip

Use the live demo to see how API responses look like. Run an analysis, select Key Elements on the left menu then select the {...} json tab.

For the description of the contents, language and version properties see the API resources output overview.

Common properties

Items that can be directly mapped to the text have properties indicating their position.
Items that occur only once, such as a sentence, have a start and end properties while items that can occur multiple times, such as a main lemma, have a positions array containing the start and end positions of all the occurrences.

Items also have a score property which provides a measure of their relevance.

topics

The topics array contains references to Knowledge Graph topics pertinent with the text.
Each array item corresponds to a topic, for example:

{
    "id": 223,
    "label": "mechanics",
    "score": 3.5,
    "winner": true
}

Possible topics are listed in the reference section.
id is the identification number, winner is a Boolean flag set to true if the topic is considered particularly relevant.

mainSentences

The mainSentences array contains info about relevant sentences.
Each array item represents a sentence, for example:

{
    "value": "The machine is held until ready to start by a sort of trap to be sprung when all is ready; then with a tremendous flapping and snapping of the four-cylinder engine, the huge machine springs aloft.",
    "score": 13.3,
    "start": 740,
    "end": 936
}

mainPhrases

The mainPhrases array contains info about the phrases deemed particularly representative during the analysis.
Each array item represents a phrase, for example:

{
    "value": "four-cylinder engine",
    "score": 8,
    "positions": [
        {
            "start": 883,
            "end": 903
        }
    ]
}

mainSyncons

The mainSyncons array contains references to Knowledge Graph syncons corresponding to the concepts that were considered relevant.
Each array item represents a syncon, for example:

{
    "lemma": "experiment",
    "positions": [
        {
            "end": 224,
            "start": 213
        },
        {
            "end": 2830,
            "start": 2820
        }
    ],
    "score": 5.8,
    "syncon": 2496
}

The syncon and the lemma properties are the outcome of semantic analysis and lemmatization respectively. These are exactly the same processes carried out during deep linguistic analysis. syncon can be interpreted as a pointer to the knowledge array entry having its syncon property set to the same value.

mainLemmas

The mainLemmas array contains relevant lemmas.
Each array item represents a lemma, for example:

{
    "value": "locomotive",
    "score": 6.5,
    "positions": [
        {
            "start": 1152,
            "end": 1162
        },
        {
            "start": 1163,
            "end": 1167
        },
        {
            "start": 1239,
            "end": 1249
        },
        {
            "start": 1335,
            "end": 1345
        },
        {
            "start": 1394,
            "end": 1404
        }
    ]
}

knowledge

The knowledge array contains Knowledge Graph data for the items of the mainSyncons array. Its contents are described in the article about the output of full analysis.