Natural Language Processing
Is a field of linguistics focused on understanding human language. There are classic ==NLP Tasks which are used to measure how well computers understand language. The most successful models for natural language understanding are called Transformers==.
Tasks
Classifying Sentences
There are many ways in which sentences can be classified including sentiment analysis, spam detection, grammar correctness, feature extraction, & ==zero-shot classification==.
Zero-shot Classification
Given an unseen set of labels & a candidate text, return a probability distribution over the labels given the candidate.
Classifying Words
Words can also be classified in tasks like parsing grammar & named entity recognition.
Generating Text
Is another very common task which includes summarization, translation, masked language modeling/fill in the blank, & causal language modeling/predict the next token.
Extracting Answers From Text
Is another complex & important task a.k.a. question answering.