Question similarity and grammatical correctness

Natural Language Processing (NLP) in Python

Fouad Trad

Machine Learning Engineer

Question similarity

Identifies when two questions are paraphrases
Useful for:
- Deduplication
- Clustering similar questions
- Improving search accuracy
Done with models trained on the Quora Question Pairs (QQP) dataset

Image showing three people asking questions.

QQP pipeline

from transformers import pipeline

qqp_pipeline = pipeline(
    task="text-classification", 
    model="textattack/bert-base-uncased-QQP"
    )

question1 = "How can I learn Python?"
question2 = "What is the best way to study Python?"

result = qqp_pipeline({"text": question1, "text_pair": question2})

print(result)

{'label': 'LABEL_1', 'score': 0.6853412985801697}

QQP pipeline

from transformers import pipeline
qqp_pipeline = pipeline(
    task="text-classification", 
    model="textattack/bert-base-uncased-QQP"
    )
question1 = "How can I learn Python?"
question2 = "What is the capital of France?"
result = qqp_pipeline({"text": question1, "text_pair": question2})
print(result)

{'label': 'LABEL_0', 'score': 0.9999338388442993}

Assessing grammatical correctness

Assess how much a text is grammatically correct
Useful for:
- Educational tools
- Grammar checkers
- Writing assistants
Done with models trained on the Corpus of Linguistic Acceptability (CoLA) dataset

Image showing a person assessing the correctness of a written text.

CoLA pipeline

from transformers import pipeline
cola_classifier = pipeline(
  task="text-classification", 
  model="textattack/distilbert-base-uncased-CoLA"
)

result = cola_classifier("The cat sat on the mat.")

print(result)

[{'label': 'LABEL_1', 'score': 0.9918296933174133}]

CoLA pipeline

from transformers import pipeline
cola_classifier = pipeline(
  task="text-classification", 
  model="textattack/distilbert-base-uncased-CoLA"
)
result = cola_classifier("The cat on sat mat the.")
print(result)

[{'label': 'LABEL_0', 'score': 0.9628171324729919}]

Let's practice!

Natural Language Processing (NLP) in Python