python 3.x - Context based Similarity between 2 words (Unstructured-Structured)

Question

I want to build a Similarity Matching model that must match/score based on the Context of the 2 words.

Eg:

Word 1       -       Word 2         - Similar

Data Science - Machine Learning     - 1

Data Science - Data Mining          - 1

HR Management - Talent Consultant   - 1

Data Science - UI Development       - 0

Front End Development - UI Development - 0

I have a Set of inputs like the above mentioned, without 0's and only 1's. Is there any way to approach this problem. I have tried using Similarity Measures (Fuzzy, Levenshtein, WMD etc.) between the 2 words, but did not yield expected Results. Will training Character RNN/LSTM model for the words yield good results. would random creation of (word 1, word 2, 0's) to train the model be better.

Is there any Ideas/Resources to help solve the Issue.

Thanks!

question from:https://stackoverflow.com/questions/65951876/context-based-similarity-between-2-words-unstructured-structured

Answer 1

Waitting for answers