Webmining & NLP

Objectives

The webmining & natural language processing course has several objectives:
– to practice data collection, information extraction and source matching
– equip students with theoretical tools for the study of textual data
– to provide an understanding of the main approaches structuring the proliferation of language models
– present concrete examples of applications in students’ various fields of interest
– provide the ability to perform classic text study tasks: classification, sentiment analysis, entity detection, etc.

Course outline

– Introduction to natural language processing
Main categories of models: bag-of-words and tf-ifd; neural networks (LSTM, GRU, etc.); word plunging (word2vec, GLoVe, fasttext, Elmo, BERT, etc.); probabilistic models (HMM, CRF, LDA, etc.).

– Applications: classification, sentiment analysis, entity detection, etc.
Text data processing and information extraction.

– Web data collection and use of an API.

Prerequisites

Statistical learning (neural networks); large-scale statistical learning; Bayesian statistics; Markov chains.