Keyword Extraction
Pratik R
Data Scientist
ML Engineer
Flask
Jupyter
Python
PyTorch
TensorFlow
Problem Statement:
Search and identify keywords from a description based on an existing list of keywords. The list of keywords will be updated from time to time and the model should be able to identify those exact keywords with the relevant context. The list will include keywords that are not part of the description for training, but when a new paragraph with an existing pre-trained keyword is used, then the model should extract that keyword with the right context.
Solution Approach
Generated key phrases from the input text
Created small summaries(split para based on the different contexts being discussed in the paragraph) of the input text.
Utilized custom BERT Models (Transformer NLP) for the task.
The Spark ✨
With the help of this project, the client was able to get the relevant keywords for his input paragraph which he was using for his work.
The Stats 💯
Timeline: 1 week
Budget: 400
Deliverables:
• Keyword extraction using BERT.
• Flask API
2022
