FNLP: Week 1: Overview, Ambiguity and Corpora

After presenting a summary of the course information, also available on this website, we begin our course with a brief introduction to Natural  Language Processing. We'll give an overview of what we'll cover in the course, and introduce some basic linguistic terminology that helps us to talk about the phenomena and the challenges that natural language processing technologies must handle.  One of the main challenges is linguistic ambiguity.  Ambiguity is pervasive in all human or natural languages, at all levels: there is ambiguity at the level of the word (both how it combines with other words in the sentence and what the word means), at the level of multi-word phrases and sentences (again, on both the form of the sentence and its meaning) and at the level of discourse or dialogue (in other words, what multi-sentence pieces of writing or speech mean).  Resolving ambiguities draws on rich knowledge sources, including (but not limited to) knowledge of the language itself, conventions of how language is used in conversation, and non-linguistic knowledge, like knowledge about the world or the domain that is being talked about. Humans are so good at identifying the speaker's intended meaning in spite of ambiguity that they often don't perceive they are doing the task at all!  Machines, on the other hand, need lots of clever tricks to do this task, and machine learning from corpora is a dominant method for doing NLP these days.  

This week is divided into three secitons:

These three sections include videos, the slides that were used in the videos, required readings, and a post-lecture quiz. The quiz is a chance for you to gauge your understanding of the material presented here, and so we strongly encourage you to review this content in the above order, and then complete the quiz. If there is anything you don't understand, then you have several options: 

    1. Post a question on piazza;
    2. Ask a question at the in person lectures; and/or
    3. Ask your tutor.