Needing some NLP advice ...

Feb 17, 2021 5 Comments

Hi guys, posting for my brother .
Thought I’d take some advice from Blind fam who are into Natural Language Processing. Since this isn’t my Masters area of specialization.. I wanted to take advice from subject matter experts.

This is his query in his own words
.....
I recently started learning Natural Language Processing as a part of my academic project. My project involves understanding the existing techniques and methods used in estimating Semantic Similarity of short texts and see if there's any possibility to optimise or come up with a better approach. It's kind of like a research based project. So I need to implement my approach on a standard dataset for performance evaluation.
What would be the best sources to help me understand the concepts and implementation using Python. Are there any books, courses, YouTube channels and blogs that can be helpful?
......

My understanding is that he is looking for advice on the resources, a right starting point and approach .
Appreciate any help !

My tc : 187k
Yoe : 3

comments

Want to comment? LOG IN or SIGN UP
TOP 5 Comments
  • you can start from paperswithcode.com
    Feb 17, 2021 0
  • Sabre
    XxDd20

    Go to company page Sabre

    XxDd20
    Get the book "Natural Language Processing with Python" by Steven Bird
    Feb 17, 2021 0
  • New
    xgan24

    New

    xgan24
    I did something similar at the University and i will give the below suggestions-
    Dataset - Twitter (fits the short text stuff)
    Task - loads of possible tasks - easiest is sentiment analysis. More difficult - expansion of tweet to full sentence by expanding all hashtags to words, replaces jargons with full form, etc
    Methods/library : try Gensim, it's my fav library for text analysis. You can do LDA, W2v, d2v, and tons of other things!

    Hope it helps :)
    Feb 17, 2021 0
  • If u are in Microsoft there is an internal SharePoint site for all ML related learnings
    Feb 17, 2021 0
  • Uber
    bogeymn

    Go to company page Uber

    bogeymn
    Search for related papers on Google Scholar and check out their code on GitHub.
    Feb 17, 2021 0