Ostyk / sensEmbedd

Learning sense embeddings for word similarity

senseEmbedd: Learning sense embeddings for word similarity

for a detailed overview please refer to the report

Code structure inside code folder

• preprocess.py
  • preprocessing of the Eurosense corpus
• grid_search.py
  • Grid search of parameter for the best models, and also training of the best one
• gridSearchVisulisation.py
  • visualization of the grid search performance and some other statistics
  • .py version of gridSearch_visualisation.ipynb
• tsne.py
  • TSNE visualization plots
  • .py version of tsne.ipynb