Description
This course takes you from a beginner level to being able to understand NLP concepts, linguistic theory, and then practice these basic theories using Python - with very simple examples as you code along with me. Get experience doing a full real-world workflow from Collecting your own Data to NLP Sentiment Analysis using Big Datasets of over 50,000 Tweets. Data collection: Scrape Twitter using: OSINT - Open Source Intelligence Tools: Gather text data using real-world techniques. In the real world, in many instances you would have to create your own data set; i. e source your data instead of downloading a clean, ready-made file onlineUse Python to search relevant tweets for your study and NLP to analyze sentiment. Language Syntax: Most NLP courses ignore the core domain of Linguistics. This course explains the fundamentals of Language Syntax & Parse trees - the foundation of how a machine can interpret the structure of s sentence. New to Python: If you are new to Python or any computer programming, the course instructions make it easy for you to code together with me. I explain code line by line. No Installs, we go straight to coding - Code using Google Colab - to be up-to-date with what's being used in the Data Science world 2021! The gentle pace takes you gradually from these basics of NLP foundation to being able to understand Mathematical & Linguistic (English-Language-based, Non-Mathematical) theories of Deep Learning. Natural Language Processing FoundationLinguistics & Semantics - study the background theory on natural language to better understand the Computer Science applications Pre-processing Data (cleaning) Regex, Tokenization, Stemming, Lemmatization Name Entity Recognition (NER)Part-of-Speech TaggingSQuADSQuAD - Stanford Question Answer Dataset. Train your Q & A Model on this awesome SQuAD dataset. Libraries: NLTKSci-kit LearnHugging FaceTensorflowPytorchSpaCyTwintThe topics outlined below are taught using practical Python projects! Parse TreeMarkov ChainText Classification & Sentiment Analysis Company Name Generator Unsupervised Sentiment AnalysisTopic ModellingWord Embedding with Deep Learning ModelsClosed Domain Question Answering (Like asking questions on many different topics, from Beyonce to Iranian Cuisine)LSTM using TensorFlow, Keras Sequence ModelSpeech RecognitionConvert Speech to TextNeural NetworksThis is taught from first principles - comparing Biological Neurons in the Human Brain to Artificial Neurons. Practical project: Sentiment Analysis of Steam ReviewsWord Embedding: This topic is covered in detail, similar to an undergraduate course structure that includes the theory & practical examples of: TF-IDFWord2VecOne Hot EncodinggloVeDeep LearningRecurrent Neural NetworksLSTMs Get introduced to Long short-term memory and the recurrent neural network architecture used in the field of deep learning. Build models using LSTMs
