Keywords
At a glance
The keywords model ranks noun phrases extracted from an input document, based on how relevant they are within the document. For example, in the text:
Anna went to school at University of California Santa Cruz.
Anna joined the university in 2015.
a Keywords model may decide that the noun phrases Anna and University of California Santa Cruz are more relevant than school and university.
| Class definition |
|---|
watson_nlp.workflows.keywords.text_rank.TextRank |
For language support, see Supported languages.
Pretrained models
Model names are listed below.
| Model ID | Container Image |
|---|---|
| keywords_text-rank-workflow_lang_ar_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_ar_stock:1.4.1 |
| keywords_text-rank-workflow_lang_cs_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_cs_stock:1.4.1 |
| keywords_text-rank-workflow_lang_da_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_da_stock:1.4.1 |
| keywords_text-rank-workflow_lang_de_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_de_stock:1.4.1 |
| keywords_text-rank-workflow_lang_en_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_en_stock:1.4.1 |
| keywords_text-rank-workflow_lang_es_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_es_stock:1.4.1 |
| keywords_text-rank-workflow_lang_fi_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_fi_stock:1.4.1 |
| keywords_text-rank-workflow_lang_fr_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_fr_stock:1.4.1 |
| keywords_text-rank-workflow_lang_he_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_he_stock:1.4.1 |
| keywords_text-rank-workflow_lang_hi_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_hi_stock:1.4.1 |
| keywords_text-rank-workflow_lang_it_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_it_stock:1.4.1 |
| keywords_text-rank-workflow_lang_ja_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_ja_stock:1.4.1 |
| keywords_text-rank-workflow_lang_ko_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_ko_stock:1.4.1 |
| keywords_text-rank-workflow_lang_nb_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_nb_stock:1.4.1 |
| keywords_text-rank-workflow_lang_nl_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_nl_stock:1.4.1 |
| keywords_text-rank-workflow_lang_nn_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_nn_stock:1.4.1 |
| keywords_text-rank-workflow_lang_pl_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_pl_stock:1.4.1 |
| keywords_text-rank-workflow_lang_pt_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_pt_stock:1.4.1 |
| keywords_text-rank-workflow_lang_ro_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_ro_stock:1.4.1 |
| keywords_text-rank-workflow_lang_ru_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_ru_stock:1.4.1 |
| keywords_text-rank-workflow_lang_sk_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_sk_stock:1.4.1 |
| keywords_text-rank-workflow_lang_sv_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_sv_stock:1.4.1 |
| keywords_text-rank-workflow_lang_tr_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_tr_stock:1.4.1 |
| keywords_text-rank-workflow_lang_zh-cn_stock | cp.icr.io/cp/ai/watson-nlp_keywords_text-rank-workflow_lang_zh-cn_stock:1.4.1 |
Running models
The Keywords model request accepts the following fields:
| Field | Type | Required Optional Repeated |
Description |
|---|---|---|---|
raw_document |
watson_core_data_model.nlp.RawDocument |
required | The input document on which to perform Keywords predictions |
limit |
int32 |
optional | The maximum number of predicted keywords. If not specified then the limit on the number of predicted keywords defaults to None |
Example requests
REST API
curl -s \
"http://localhost:8080/v1/watson.runtime.nlp.v1/NlpService/KeywordsPredict" \
-H "accept: application/json" \
-H "content-type: application/json" \
-H "Grpc-Metadata-mm-model-id: keywords_text-rank-workflow_lang_en_stock" \
-d '{ "raw_document": { "text": "Anna went to school at University of California Santa Cruz. Anna joined the university in 2015." }, "limit": 2 }'
Response
{"keywords":[
{"text":"University of California Santa Cruz","relevance":0.912354,"mentions":[
{"begin":23,"end":58,"text":"University of California Santa Cruz"}],"count":1},
{"text":"Anna","relevance":0.88892,"mentions":[
{"begin":0,"end":4,"text":"Anna"},{"begin":60,"end":64,"text":"Anna"}],"count":2}],
"producerId":{"name":"Keywords Text Rank Workflow","version":"0.0.1"}}
Python
import grpc
from watson_nlp_runtime_client import (
common_service_pb2,
common_service_pb2_grpc,
syntax_types_pb2,
)
channel = grpc.insecure_channel("localhost:8085")
stub = common_service_pb2_grpc.NlpServiceStub(channel)
request = common_service_pb2.KeywordsRequest(
raw_document=syntax_types_pb2.RawDocument(text="Anna went to school at University of California Santa Cruz. Anna joined the university in 2015."),
limit=2
)
response = stub.KeywordsPredict(
request, metadata=[("mm-model-id", "keywords_text-rank-workflow_lang_en_stock")]
)
print(response)
Response
keywords {
text: "University of California Santa Cruz"
relevance: 0.912354
mentions {
begin: 23
end: 58
text: "University of California Santa Cruz"
}
count: 1
}
keywords {
text: "Anna"
relevance: 0.88892
mentions {
end: 4
text: "Anna"
}
mentions {
begin: 60
end: 64
text: "Anna"
}
count: 2
}
producer_id {
name: "Keywords Text Rank Workflow"
version: "0.0.1"
}