Introducing
NLP
Concept
History
Applications
Approaches
Language Challenges
ChatGPT vs Human
Interaction using NLP
Human talks to the machine
Machine captures the audio
Audio to text conversion takes place
Processing of text data
Data to audio conversion takes place
Machine responds to human by playing audio file
Branch of Computer Science & AI that gives computers the ability
to understand text and spoken words in the same way as human
beings can.
Constituents of NLP
Natural Language Understanding
Lexical Ambiguity
He is looking for a match
The fish is ready to eat
Syntatical Ambiguity
They are cooking apples
NLP pipeline (pre-processing)
NLP pipeline
Segmentation
Tokenization
Removing Stop Words
Stemming
Lemmatization
Part of Speech Tagging
Named Entity Recognition
Information Retrieval vs Information Extraction
Information Retrieval vs Information Extraction
Phases of NLP
Phases of NLP
Lexical Analysis
Lexicon of a language – collection of words & phrases
Divide text – paragraphs, sentence, words – lexicon normalization required
Lexicon normalization – Stemming, Lemmatization
Stemming – reduce derived words to root form like ing, ly, es, s
Lemmatization – Reduce to lemma or dictionary form
Syntactic Analysis
Used to check grammar, interrelationship between words
“Mumbai goes to Sara” rejected by Syntactic analyzer
Dependency Grammar and POS used
Semantic Analysis
Retrieve meaningful insights from text
Process of extracting meaning from text
Combines machine learning and natural language processing to find the
text’s main ideas and connections
Utilize pre-established ontologies and structured databases of concepts and
relationships in a particular subject
Where does Semantic Analysis work?
Discourse Integration
Understand sense of context – use of proper nouns / pronouns
Ram wants it
“it” keyword does not make sense
Depends on previous sentence
Pragmatic Analysis
Study of meaning – extract insight from text
Understand context, who said to whom
“it” keyword does not make sense
Depends on previous sentence
"What do you want to eat?" and another responded, "Ice cream is good
this time of year."
What is NLP?
What is
NLP?
Applications of NLP
Applications of NLP
Alexa – understand words, OK Google, SIRI
We say Play a song
Speech – text – text – speech
Word processing
Language translation
Grammarly for correction of grammar
Interactive voice response in call centres
Question
What is the main challenge of NLP?
a)Handling Ambiguity of Sentences
b) Handling Tokenization
c) Handling POS-Tagging
d) All of the mentioned
Answer
What is the main challenge of NLP?
a)Handling Ambiguity of Sentences
b) Handling Tokenization
c) Handling POS-Tagging
d) All of the mentioned
Question
Choose form the following areas where NLP can be useful.
a) Automatic Text Summarization
b) Automatic Question-Answering Systems
c) Information Retrieval
d) All of the mentioned
Answer
Choose form the following areas where NLP can be useful.
a) Automatic Text Summarization
b) Automatic Question-Answering Systems
c) Information Retrieval
d) All of the mentioned
History of NLP
History of NLP
Chatbots
JABBER Dr. SmartC
ELIZA PARRY A.L.I.C.E
WACKY Sbaitso hild
1966 1972 1995
1988 1992 2001
Google
SIRI Cortana ALEXA ChatGPT
Assistant
2010 2014 2014 2021
2017
Why Tokenization?
Text Preprocessing : remove unnecessary words
Feature extraction : token used as a feature
Word-level analysis : POS, NER
Text Representation : using bag-of-words, word embeddings
Language Processing : Parsing, Stemming, Lemmatization
Efficient storage and retrieval : reduce dimensionality
Statistical Analysis: Calculate word frequency to extract
information
Stemming
Stemming
Stemming
Stemming
Morphology
Morphology
Morphology
Types of Morpheme
Free Morpheme
Lexical Morpheme
Lexical Morpheme
Functional Morpheme
Functional Morpheme
Bound Morpheme
Bound Morpheme
Bound Morpheme
Bound Morpheme
Derivational Morpheme
Derivational Morpheme
Derivational Morpheme
Derivational Morpheme
Inflexional Morpheme
Inflexional Morpheme
Morpheme
Classify Morpheme – Free, Bound,
Inflexional, Derivational
Token Morpheme Type
Happy
Book
Ness in happiness
Ed in walked
Classify Morpheme – Free, Bound
Token Morpheme Type
Happy Free - lexical
Book Free - lexical
Ness in happiness Bound - derivational
Ed in walked Bound - inflexional
Language Challenges
Main Approaches in NLP
Rule Based Methods
▪ Regular Expressions
▪ Context-free grammar
Probablistic modeling & Machine Learning
▪ Likelihood maximization
▪ Linear classifiers
Deep Learning
▪ Recurrent Neural Network
▪ Convolutional Neural Network
Main Approaches in NLP
Rule Based Methods
▪ Regular Expressions
▪ Context-free grammar
Probablistic modeling & Machine Learning
▪ Likelihood maximization
▪ Linear classifiers
Deep Learning
▪ Recurrent Neural Network
▪ Convolutional Neural Network
Main Approaches in NLP
Rule Based Methods
▪ Regular Expressions
▪ Context-free grammar
Probablistic modeling & Machine Learning
▪ Likelihood maximization
▪ Linear classifiers
Deep Learning
▪ Recurrent Neural Network
▪ Convolutional Neural Network
Main Approaches in NLP
Rule Based Methods
▪ Regular Expressions
▪ Context-free grammar
Probablistic modeling & Machine Learning
▪ Likelihood maximization
▪ Linear classifiers
Deep Learning
▪ Recurrent Neural Network
▪ Convolutional Neural Network
