Please enable JavaScript.
Coggle requires JavaScript to display documents.
Natural Language Processing - Coggle Diagram
Natural Language Processing
Solved/Progress:
Mostly Solved: OCR, NER, POS tagging
Really Hard: Question Answering, Paraphrasing, Summarizing, Dialogue, Common Sense
Good Progress: Sentiment Analysis, Co-reference Resolution, WSD, Parsing, Information Extraction, Machine Translation
World Knowledge & Common Sense:
Implicit knowledge that is difficult to explain
Common Sense knowledge bases (e.g., Open Mind Common Sense)
Examples: understanding object permanence, making inferences about time and location, naive psychology
AI Capabilities:
NLP is an AI-COMPLETE problem
Solving it would make computers as intelligent as humans
Definition of NLP:
Computational techniques for analyzing and representing text
Achieving human-like language processing for applications
Interactions between computers and human languages
History:
1970s: Conceptual ontologies, chatterbots
1960s: Eliza, SHRDLU
1980s: Shift from hand-written rules to machine learning
1950s: Turing Test, Georgetown experiment
What is NLP
The goal of natural language processing is to create machines that can comprehend and react to text or voice input, as well as produce their own writing or speech in a manner similar to that of humans.
Challenges
Lexicon,
Morphology
Discourse Analysis
Pragmatics
N-gram
A continuous sequence of n items from a given text or speech sample is called an n-gram. Depending on the application, the items may be phonemes, syllables, letters, words, or base pairs.
Linguistics:
Semantics (meaning of sentences)
Phonetics (speech sounds)
Morphology (understanding words based on morphemes)
Lexicon (word meanings, relationships)
Pragmatics (extra meaning derived from context)
Definition of NLP:
Interactions between computers and human languages
Achieving human-like language processing for applications
Computational techniques for analyzing and representing text
Requirements for NLP:
Cognitive Psychology (language usage as a window to human cognition)
Linguistics (formal models, language universals)
Computer Science (data representations, efficient processing)