Skip to content

Jspano95/un_speeches_nlp_analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

UNITED NATIONS SPEECHES 1970-2015 ANALYSIS

GENERAL STRUCTURE

  • Data introduction & overview in notebook (1)
  • Data cleaning & tokenisation
  • TF-IDF weighting of document-term matrix
  • Initial exploratory analysis & visualisations (wordclouds etc.)
  • Topic modeling with SKLEARN
    • includes: Nonnegative Matrix Factorisation (NMF)
    • Singular Value Decomposition (SVD)
    • SpaCy-processed lemma extracted NMF model
  • Topic modeling with GENSIM
    • NMF unigrams and NMF bigrams
    • Latent Dirichlet Allocation (LDA)
    • Hierarchical Dirichlet Process (HDP) (experimental implementation)
    • Optimal topic counts with Coherence optimisation
  • Automated text summarisations of speeches
    • TF-IDF
    • LSA Algo
    • TextRank Algo

STILL A WORK IN PROGRESS - FUTURE WORKS:

  • Further refine models using different grammatical structures extracted with SpaCy in notebook (5)
    • E.G. noun phrase, NER mentions etc.
  • Re-add tuned Latent Dirichlet Allocation & interactive visual for SKLEARN
  • Re-add the SVD model (required fixing before pushing changes - temporarily removed as of 31.03.22)

Possible Future Expansions

  • Some sort of sentiment analysis if applicable - perhaps by country over time

About

NLP analysis of UN-speeches from 1970 - 2015

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published