All the notebooks here are the part of the assignments of Natural Langauge Processing course at TU Dortmund. I have used Spherical KMeans clustering to create clusters, and I have also used Cross Tabulation Matrix, Pair counts, RAND index, adjusted RAND index, and Normalized Mutual Information to evaluate clusters. For topic modelling I have implemented Latent Semantic Indexing, Probabilistic Latent Semantic Indexing and Latent Dirichlet Allocation.