Search code examples
pythonnlpdata-analysis

Get the topics or the keywords of a sentence

Good morning all

Does anyone of you know a tool or an API or something that takes a sentence as input and as output, it gives the topics or keywords of this sentence?

I tried TextRazor in the online demo it works well like you can see in the screenshot screenshot but when I used as a library in my python code it always gives me a blank list even for the sentence used in the demo this is my code in python:

import textrazor
import ssl
textrazor.api_key ="bdd69bdc3f91045cdb6d4261d39df34d887278602cb8f60401b7eb0b"
client = textrazor.TextRazor(extractors=["entities", "topics"])
client.set_cleanup_mode("cleanHTML")
client.set_classifiers(["textrazor_newscodes"])
sentence = "Adam Hill,b It's Super Bowl Sunday  pastors. Get your Jesus Jukes ready! Guilt is an awesome motivator! #sarcasm"
response = client.analyze(sentence)
print(sentence)
print(len(response.topics()))
entities = list(response.entities())
print(len(entities))
for topic in response.topics():
    if topic.score > 0.3:
        print (topic.label)

It gives me zero for entities and topics length

Someone proposed for me to use OpenNlp but I didn't get how to extract topics and keywords if anyone of you has any tutorial or clarification please help me

And thank you in advance

Solution

  • You have to remove the line client.set_cleanup_mode("cleanHTML"). Then it should work just fine.

    As I understand the cleanup_mode, it treads your text as an html. As your example text is not html it won't find any raw text in between html tags.