Natural Language Processing: 1.

NLP is a Programming Languages Vs Natural

branch of AI that helps computers Languages: 1. Programming languages are
understand, interpret, and generate human formal, precise, and unambiguous. 2.
language. 2. Used in chatbots, language Natural languages are context-dependent,
translation, sentiment analysis, and voice ambiguous, and nuanced. 3. Programming
recognition. 3. Includes tokenization, languages follow strict syntax and grammar
stemming, lemmatization, and named entity rules. 4. Natural languages have flexible
recognition. 4. Ambiguity, context grammar and vocabulary usage. 5.
understanding, and language diversity pose Programming languages are designed for
significant challenges. 5. Breaking text into machine execution and automation. 6.
words, phrases, or symbols for analysis. 6. Natural languages serve human
Determines emotions expressed in text, communication and expression. 7.
such as positive, negative, or neutral. 7. Programming languages are primarily used
Identifies and categorizes named entities in for instructing computers. 8. Natural
text like names, places, or organizations. 8. languages convey emotions, intentions, and
Translates text from one language to abstract concepts. 9. Programming
another using algorithms and models. 9. languages prioritize efficiency, clarity, and
Labels words in a sentence with their logic. 10. Natural languages evolve
grammatical parts of speech. 10. organically and reflect cultural influences.
Advancements in deep learning, neural
Are natural languages regular? 1. Natural
networks, and unsupervised learning are
languages exhibit a mix of regular and
shaping NLP's future.
irregular patterns. 2. They don't strictly
Why is NLP hard? 1. Understanding adhere to regular grammatical or
context and ambiguity in language poses phonological rules. 3. While some aspects
challenges. 2. Different languages, dialects, follow predictable patterns, exceptions are
and variations complicate analysis and common. 4. Syntax, morphology, and
processing. 3. Cultural nuances and phonology can vary greatly within and
linguistic subtleties require nuanced across languages. 5. Irregular verbs,
interpretation. 4. Contextual understanding irregular plurals, and irregular word
requires knowledge beyond word-to-word formations defy regularity. 6. Idioms,
translation. 5. Handling sarcasm, irony, and colloquialisms, and slang further contribute
figurative language demands advanced to irregularity. 7. Regional dialects and
algorithms. 6. Ambiguous pronouns and language evolution add complexity to
references necessitate sophisticated regularity. 8. Despite irregularities,
disambiguation techniques. 7. Dealing with languages still maintain some regular
slang, informal language, and evolving structures and conventions. 9. Linguists
vocabulary is challenging. 8. Language study both regular and irregular aspects to
evolution and change demand constant understand language dynamics. 10. Overall,
updates to models and algorithms. 9. natural languages are characterized by a
Processing unstructured data like social rich blend of regular and irregular elements.
media posts requires adaptability. 10.
Integrating NLP with other AI tasks like
reasoning and decision-making adds
complexity.
Why are finite automata important in interpretations of words or phrases. 2.
NLP? 1. Finite automata are used to create Context Understanding: Grasping meaning
tokenizers, breaking text into tokens based in varying contexts. 3. Language Diversity:
on predefined rules. 2. Automata help Handling the multitude of languages and
search for patterns within text, aiding tasks dialects. 4. Data Quality: Ensuring accuracy
like named entity recognition. 3. Finite and reliability of training data. 5. Domain
automata can model word morphology, Adaptation: Adapting models to different
useful in stemming or lemmatization. 4. subject areas or industries. 6. Rare
Regular expressions underpin many NLP Language Phenomena: Addressing
tools for tasks such as text search or uncommon linguistic patterns.7. Ethical
information extraction. 5. Automata assist Considerations: Managing biases and
in analyzing the lexical structure of fairness in language processing.8.
sentences, crucial for parsing. 6. Finite Interdisciplinary Integration: Incorporating
automata help in identifying languages knowledge from linguistics, psychology,
based on linguistic features. 7. Although and other fields. 9. Interpretability: Making
primarily used in parsing, automata models transparent and understandable. 10.
concepts influence syntactic analysis in Continuous Learning: Updating models to
NLP. 8. Automata-based methods are keep pace with evolving language usage.
employed to suggest corrections for
Tokenization: 1. Breaks text into smaller
misspelled words. 9. Automata serve as
units like words, phrases, or symbols. 2.
foundational concepts in designing
Helps in preparing text data for further
conversational systems. 10. Finite automata
analysis. 3. Can involve splitting based on
aid in preprocessing tasks like noise
spaces, punctuation, or specific rules. 4.
removal or text normalization.
Handles complexities like contractions,
Stages of NLP: 1. Text Preprocessing: hyphenated words, and special characters.
Cleaning and organizing text data before 5. Different languages may require unique
analysis. 2. Tokenization: Breaking text tokenization techniques. 6. Customization
into smaller units like words or phrases. 3. possible for domain-specific tokenization
Part-of-Speech Tagging: Assigning needs. 7. Important preprocessing step in
grammatical labels to words in a sentence. natural language processing tasks. 8.
4. Parsing: Analyzing the grammatical Handles challenges like tokenizing URLs,
structure of sentences. 5. Named Entity email addresses, or social media handles. 9.
Recognition: Identifying and categorizing Considers context for tokenizing compound
named entities such as names, dates, or words or multi-word expressions. 10.
locations. 6. Semantic Analysis: Efficiency important for large-scale text
Understanding the meaning of words and processing tasks.
sentences. 7. Sentiment Analysis:
Determining the sentiment or emotional
tone of text. 8. Language Modeling:
Predicting the next word in a sequence
based on context. 9. Machine Translation:
Translating text from one language to
another.
Challenges and Issues in NLP :1.
Ambiguity: Resolving multiple
Stemming: 1. Reduces words to their root
or base form. 2. Helps in collapsing variants
of words to a common form. 3. Improves
efficiency by reducing the vocabulary size.
4. Uses heuristic algorithms to strip affixes
from words. 5. Simple and fast but may not
always produce valid root forms. 6.
Handles different morphological variations
of words. 7. Common stemming algorithms
include Porter, Snowball, and Lancaster. 8.
May result in over stemming or under
stemming issues. 9. Not language-specific
but effectiveness varies across languages.
Lemmatization: 1. Derives the canonical
form or lemma of a word. 2. Considers
word's meaning and context to produce
valid lemmas. 3. Yields more accurate
results compared to stemming. 4. Utilizes
language dictionaries and morphological
analysis. 5. Slower but produces
linguistically correct lemmas. 6. Handles
irregular words and exceptions effectively.
7. Useful for tasks requiring precise word
forms like language generation. 8. Context-
aware, providing different lemmas based on
usage. 9. Suitable for languages with
complex inflectional patterns.
Part-of-Speech Tagging 1. Labels words
in a sentence with their grammatical
categories. 2. Assigns tags like noun, verb,
adjective, etc., to each word. 3. Helps in
understanding sentence structure and
meaning. 4. Utilizes statistical models or
rule-based approaches. 5. Accuracy varies
based on the complexity of language and
context. 6. Handles ambiguities arising
from homographs or polysemous words. 7.
Can be language-dependent due to different
grammatical structures. 8. Important for
syntactic analysis, parsing, and machine
translation 9. Improves accuracy of
downstream NLP tasks like sentiment
analysis. 10. Annotated corpora serve as
training data for POS taggers.
 ****Unit -2 **** forms effectively. 7. Can be language-
specific, tailored to the morphological
Morphology: 1. Study of word formation patterns of each language. 8. Requires
and structure in language. 2. Deals with linguistic expertise to design accurate
morphemes, the smallest units of meaning. transducers. 9. Provides a foundation for
3. Crucial in understanding how words are building robust morphological analyzers.
built from morphemes. 4. Helps analyze 10. Integral in many language processing
inflectional and derivational processes. pipelines for tasks requiring morphological
Types of Morphemes: 1. Free Morphemes: analysis.
Can stand alone as words (e.g., "cat"). 2.
Syntactic Representations of Natural
Root Morphemes: Carry core meaning and
Language :1. Hierarchical structures
can stand alone (e.g., "play"). 3. Inflectional
showing how words combine into phrases
Morphemes: Alter grammatical function or
and sentences. 2. Represent relationships
form without changing the word's basic
between words based on dependencies. 3.
meaning (e.g., verb tense markers). 4.
Group words into constituents (e.g., noun
Derivational Morphemes: Create new
phrases, verb phrases). 4. Capture
words or change word class, often altering
Grammatical Structure. 5. Provide Insights
meaning significantly (e.g., "-er" in
into syntactic relationships. 6. Used in
"teacher").
Parsing algorithms. 7. Facilitate Language
Inflectional Morphology: 1. Involves Understanding. 8. Enable Grammar
adding inflectional morphemes to a word. Checking. 9. Applied in Machine
2. Indicates grammatical information like Translation. 10. Crucial in Natural
tense, number, or case. 3. Does not change Language Understanding (NLU).
the word's core meaning. 4. Commonly
Parsing Algorithms: 1. Start with the root
seen in verb conjugation and noun
of the parse tree and recursively expand
declension. 5. Examples include plural "-s,"
constituents. 2. Start with individual words
past tense "-ed," and comparative "-er."
and gradually build up constituents to form
Derivational Morphology: 1. Modifies the
the parse tree. 3. Store intermediate parsing
meaning or part of speech of a word. 2.
results in a chart for efficient processing
Often involves adding derivational
and recombination. 4. Determine
morphemes to a root. 3. Can create entirely
grammatical relationships between words
new words or word forms. 4. Examples
based on dependencies. 5. Use probabilistic
include "-ize" (e.g., "realize") and "-ful"
models to find the most likely parse for a
(e.g., "beautiful").
given sentence. 6. Incrementally build the
Morphological Parsing with Finite State parse tree using a sequence of parsing
Transducers (FST): 1. Utilizes finite state actions. 7. Recognize sentences in a
transducers to analyze morphological context-free grammar efficiently using
structure. 2. Represents morphological dynamic programming. 8. Employ dynamic
rules as finite state machines. 3. Allows for programming to parse sentences in
efficient analysis of complex Chomsky normal form context-free
morphological processes. 4. Commonly grammars. 9. Adjust parsing strategies
used in natural language processing tasks. based on the characteristics of the input
5. Enables tasks like tokenization, sentence for improved efficiency.
stemming, and part-of-speech tagging. 6.
Handles ambiguity and variability in word
Probabilistic Context-Free Grammars: linguistic theories and computational
1. Extend context-free grammars with models for analysis. 7. Enables computers
probabilities for each production rule. 2. to comprehend and generate human-like
Assign likelihoods to different syntactic language. 8. Forms the basis for tasks like
structures. 3. Enable parsing algorithms to word sense disambiguation and semantic
consider the likelihood of alternative analysis. 9. Explores the nuances and
parses. 4. Used in statistical parsing to complexities of word meanings. 10. Crucial
capture uncertainty in sentence structure. 5. for developing accurate and effective
Incorporate probabilities to determine the language technologies.
most likely parse for a given sentence. 6.
Relations Among Lexemes: 1.
Model syntactic ambiguity inherent in
Homonymy: Words with the same form but
natural language. 7. Allow for flexible
different meanings. 2. Polysemy: Words
representation of language structures. 8.
with multiple related meanings. 3.
Facilitate robust parsing in the presence of
Synonymy: Words with similar meanings.
noise or incomplete data. 9. Useful in
4. Hyponymy: Relationship between a
applications like machine translation and
general term and its specific instances. 5.
speech recognition. 10. Play a key role in
WordNet: Lexical database organizing
probabilistic models for natural language
words based on semantic relations. 6. Word
understanding.
Sense Disambiguation (WSD): Identifying
Statistical Parsing: 1. Parsing approach correct meanings of ambiguous words. 7.
that utilizes statistical models and data- Dictionary-based approach: Resolving
driven methods. 2. Learns syntactic patterns word meanings using dictionary
and structures from annotated corpora. 3. definitions. 8. Latent Semantic Analysis:
Often based on probabilistic context-free Statistical method for capturing word
grammars (PCFGs). 4. Enables parsing semantics. 9. Enhances natural language
algorithms to find the most likely parse for understanding and information retrieval
a given sentence. 5. Accounts for variability systems.
and ambiguity in language use. 6.
Homonymy: 1. Words with the same form
Incorporates frequency-based information
but different meanings. 2. Example: "bat"
to guide parsing decisions. 7. Balances
(flying mammal) vs. "bat" (sports
between different parses based on their
equipment). 3. Often leads to ambiguity in
probabilities. 8. Can handle complex
language understanding. 4. Requires
syntactic constructions and long-distance
context to distinguish between meanings. 5.
dependencies. 9. Adapts to different
Common in languages with rich
domains or languages through training on
vocabularies. 6. Can create challenges in
relevant data.
natural language processing tasks. 7.
Lexical Semantic: 1. Study of word Examples include "bank" (financial
meanings and relationships in language. 2. institution) and "bank" (river edge). 8.
Focuses on understanding how words Homonyms may have unrelated or distantly
convey meaning in context. 3. Essential for related meanings. 9. Differentiates between
natural language understanding and homonyms and other lexical relationships.
processing tasks. 4. Analyzes relationships 10. Addressed through techniques like word
among words at the lexical level. 5. sense disambiguation (WSD).
Includes concepts like synonymy,
antonymy, hyponymy, and more. 6. Utilizes
Polysemy: 1. Words with multiple related Hyponymy, meronymy, antonymy. 7.
meanings. 2. Example: "book" (physical Widely used in research and applications. 8.
object) vs. "book" (to reserve). 3. Related Facilitates word sense disambiguation and
senses often share a common semantic semantic analysis.
thread. 4. May involve metaphorical or
Word Sense Disambiguation (WSD): 1.
extended meanings. 5. Can enrich language
Identifying correct meanings of ambiguous
by providing layers of nuance. 6. Requires
words. 2. Resolves ambiguity in natural
understanding of context for
language processing tasks. 3. Relies on
disambiguation. 7. Example: "arm" (body
context, knowledge sources, and
part) vs. "arm" (military branch). 8.
algorithms. 4. Enables accurate
Polysemous words contribute to linguistic
understanding and interpretation of text. 5.
richness.
Essential for applications like machine
Synonymy: 1. Words with similar translation and information retrieval. 6.
meanings. 2. Example: "happy" and Addresses challenges posed by polysemy
"joyful". 3. Provides flexibility in language and homonymy. 7. Utilizes techniques such
use. 4. Varies in degree of similarity and as supervised learning and knowledge-
context dependency. 5. Enhances language based methods. 8. Evaluates performance
richness and expressiveness. 6. Synonyms through metrics like precision and recall.
may differ in connotation or register. 7.
Dictionary-based Approach: 1. Resolving
Example: "big" and "large". 8. Important
word meanings using dictionary
for lexical choice in writing and
definitions. 2. Relies on lexical resources
communication. 9. Distinguished from
such as dictionaries or lexicons. 3. Matches
other lexical relationships like antonymy.
word senses with definitions or glosses. 4.
10. Utilized in lexical resources and natural
Provides a straightforward method for word
language processing tasks.
sense disambiguation. 5. Suitable for
Hyponymy: 1. Relationship between a applications requiring precise definitions.
general term and its specific instances. 2. 6. May not capture contextual nuances or
Example: "fruit" (general) and "apple" domain-specific meanings. 7. Examples
(specific). 3. Forms a hierarchy of include WordNet and Oxford English
inclusiveness. 4. Hierarchical structure Dictionary. 8. Offers a baseline approach
reflects semantic relationships. 5. Allows for sense disambiguation tasks.
for categorization and organization of
Latent Semantic Analysis : 1. Statistical
concepts. 6. Provides precision in language
method for capturing word semantics. 2.
use. 7. Example: "animal" (general) and
Represents words and documents in a high-
"cat" (specific). 8. Enables efficient
dimensional semantic space. 3. Identifies
communication by specifying details.
latent patterns of word co-occurrence. 4.
WordNet: 1. Lexical database organizing Utilizes techniques like singular value
words based on semantic relations. 2. decomposition. 5. Enables automatic
Contains synsets (sets of synonymous indexing and retrieval of text documents. 6.
words) and semantic relations. 3. Provides Discovers semantic relationships between
a structured representation of word words. 7. Addresses synonymy and
meanings. 4. Supports various natural polysemy through contextual analysis. 8.
language processing tasks. 5. Organizes Widely used in information retrieval and
words into hierarchical and non- text mining.
hierarchical relations. 6. Example: