Unit-I

1.Finding the Structure of Words

This section deals with words, its structure and its models
1. Words and Their Components
1. Tokens
2. Lexemes
3. Morphemes
4. Typology
2. Issues and Challenges
1. Irregularity
2. Ambiguity
3. Productivity
3. Morphological Models
1. Dictionary Lookup
2. Finite-State Morphology
3. Unification-Based Morphology
4. Functional Morphology
5. Morphology Induction
2.Finding the Structure of Documents
This chapter mainly deals with Sentence and topic detection or segmentation.
1. Introduction
1. Sentence Boundary Detection
2. Topic Boundary Detection
2. Methods
This section deals with statistical classical approaches (Generative and Discriminative
approaches)
1. Generative Sequence Classification Methods
2. Discriminative Local Classification Methods
2.3.3 Discriminative Sequence Classification Methods
4. Hybrid Approaches
5. Extensions for Global Modelling for
Sentence Segmentation
3. Complexity of the Approaches
4. Performance of the Approaches
 NATURAL LANGUAGE PROCESSING(NLP)

UNIT - I
i. Finding the Structure of Words:

 Words and Their Components 

 Issues and Challenges 
 Morphological Models 
ii. Finding the Structure of Documents:

 Introduction 
 Methods 
 Complexity of the Approaches 
 Performances of the Approaches 
 NATURAL LANGUAGE PROCESSING
 Humans communicate through some form of language either by text or speech. 
 To make interactions between computers and humans, computers need to understand natural languages used by
humans. 
 Natural language processing is all about making computers learn, understand, analyse, manipulate and interpret 
natural(human) languages.
 NLP stands for Natural Language Processing, which is a part of Computer Science, Human
language, and Artificial Intelligence.
 Processing of Natural Language is required when you want an intelligent system like robot to perform as per your
instructions, when you want to hear decision from a dialogue based clinical expert system, etc. 
 The ability of machines to interpret human language is now at the core of many applications that we use every day 
- chatbots, Email classification and spam filters, search engines, grammar checkers, voice assistants, and social
language translators.
 The input and output of an NLP system can be Speech or Written Text 
Definition
 Branch of Artificial Intelligence that helps computers to understand, interpret and manipulate human language. 
 Components of NLP
 There are two components of NLP, Natural Language Understanding (NLU)
and Natural Language Generation (NLG).
 Natural Language Understanding (NLU) which involves transforming human 
language into a machine-readable format.
 It helps the machine to understand and analyse human language by extracting the
text from large data such as keywords, emotions, relations, and semantics. 
 Natural Language Generation (NLG) acts as a translator that converts the
computerized data into natural language representation. 
 It mainly involves Text planning, Sentence planning, and Text realization. 
 The NLU is harder than NLG. 
 NLP Terminology
 Phonology − It is study of organizing sound systematically. 
 Morphology: The study of the formation and internal structure of words. 
 Morpheme − It is primitive unit of meaning in a language. 
 Syntax: The study of the formation and internal structure of sentences. 
 Semantics: The study of the meaning of sentences. 
 Pragmatics − It deals with using and understanding sentences in different situations
and how the interpretation of the sentence is affected. 
 Discourse − It deals with how the immediately preceding sentence can affect the
interpretation of the next sentence. 
 World Knowledge − It includes the general knowledge about the world. 
Steps in NLP
 There are general five steps : 
1. Lexical Analysis
2. Syntactic Analysis (Parsing)
3. Semantic Analysis
4. Discourse Integration
5. Pragmatic Analysis
1 . LEXICAL ANALYSIS
 The first phase of NLP is the Lexical Analysis. 
 This phase scans the source code as a stream of characters and converts it into meaningful
lexemes. 
 It divides the whole text into paragraphs, sentences, and words. 
 For example, a word like “uneasy” can be broken into two sub-word tokens as “un-
easy”.
2. Syntactic Analysis (Parsing) –
 Syntactic Analysis is used to check grammar, word arrangements, and 

 shows the relationship among the words. 

 The sentence such as “The school goes to boy” is rejected by English syntactic analyzer.
3. Semantic Analysis –
 Semantic analysis is concerned with the meaning representation. 
 It mainly focuses on the literal meaning of words, phrases, and sentences. 
 The semantic analyzer disregards sentence such as “hot ice-cream”. 
4. Discourse Integration –
 Discourse Integration depends upon the sentences that proceeds it
and also invokes the
 meaning of the sentences that follow it.

5. Pragmatic Analysis –
 During this, what was said is re-interpreted on what it actually meant.
 It involves deriving those aspects of language which require real world knowledge.
 Example: "Open the door" is interpreted as a request instead of an
order.
 Finding the Structure of Words
 Human language is a complicated thing. 
 We use it to express our thoughts, and through language, we receive information and infer its
meaning. 
 Trying to understand language all together is not a viable approach. 
 Linguists have developed whole disciplines that look at language from different perspectives
and at different levels of detail. 
 The point of morphology, for instance, is to study the variable forms and functions of words, 
 The syntax is concerned with the arrangement of words into phrases, clauses, and sentences. 
 Word structure constraints due to pronunciation are described by phonology, 
 The conventions for writing constitute the orthography of a language. 
 The meaning of a linguistic expression is its semantics, and etymology and
lexicology cover especially the evolution of words and explain the semantic,
morphological, and other links among them.
 Words are perhaps the most intuitive units of language, yet they are in general
tricky to define.

 Knowing how to work with them allows, in particular, the development of

syntactic and semantic abstractions and simplifies other advanced views on
language.

 Here, first we explore how to identify words of distinct types in human

languages, and how the internal structure of words can be modelled in
connection with the grammatical properties and lexical concepts the words
should represent.
 The discovery of word structure is morphological parsing. 
 In many languages, words are delimited in the orthography by whitespace and
punctuation. 
 But in many other languages, the writing system leaves it up to the reader to tell words
apart or determine their exact phonological forms. 
Words and Their Components
 Words are defined in most languages as the smallest linguistic units that can form a
complete utterance by themselves. 
 The minimal parts of words that deliver aspects of meaning to them are called 
morphemes.
Tokens
 Suppose, for a moment, that words in English are delimited only by whitespace and
punctuation (the marks, such as full stop, comma, and brackets) 
 Example: Will you read the newspaper? Will you read it? I won’t read it. 
 If we confront our assumption with insights from syntax, we notice two here: words 
newspaper and won’t.

 Being a compound word, newspaper has an interesting derivational structure. 

 In writing, newspaper and the associated concept is distinguished from the
isolated news and paper. 

 For reasons of generality, linguists prefer to analyze won’t as two syntactic words, or tokens,
each of which has its independent role and can be reverted to its normalized form. 
 The structure of won’t could be parsed as will followed by not. 

 In English, this kind of tokenization and normalization may apply to just a limited set of
cases, but in other languages, these phenomena have to be treated in a less trivial manner. 
• In Arabic or Hebrew, certain tokens are concatenated in writing
with the preceding or the following ones, possibly changing
their forms as well.
• The underlying lexical or syntactic units are thereby blurred into one compact string
of letters and no longer appear as distinct words.

• Tokens behaving in this way can be found in various languages and are often
called clitics.

• In the writing systems of Chinese, Japanese, and Thai, whitespace is not used to
separate words.
 Lexemes
 By the term word, we often denote not just the one linguistic form in the given
context but also the concept behind the form and the set of alternative forms that
can express it. 
 Such sets are called lexemes or lexical items, and they constitute the lexicon of a
language. 
 Lexemes can be divided by their behaviour into the lexical categories of verbs, nouns,
adjectives, conjunctions, particles, or other parts of speech. 
 The citation form of a lexeme, by which it is commonly identified, is also called its 
lemma.
 When we convert a word into its other forms, such as turning the singular mouse into
the plural mice or mouses, we say we inflect the lexeme. 
 When we transform a lexeme into another one that is morphologically related,
regardless of its lexical category, we say we derive the lexeme: for instance, the
nouns receiver and reception are derived from the verb to receive. 
 Example: Did you see him? I didn’t see him. I didn’t see anyone. 
• Example presents the problem of tokenization of didn’t and the investigation of the
internal structure of anyone.
• In the paraphrase I saw no one, the lexeme to see would be inflected into the
form saw to reflect its grammatical function of expressing positive past tense.
• Likewise, him is the oblique case form of he or even of a more abstract lexeme
representing all personal pronouns.
• In the paraphrase, no one can be perceived as the minimal word
synonymous with nobody.
• The difficulty with the definition of what counts as a word need not pose a
problem for the syntactic descriM ptoiorn
phifew
meesunderstand no one as two
closely connected tokens treated as one fixed element
 Morphology
 Morphology is the domain of linguistics that
analyses the internal structure of words.
 Morphological analysis – exploring the structure of words
 Words are built up of minimal meaningful elements called morphemes:
played = play-ed
cats = cat-s
unfriendly = un-friend-ly
 Two types of morphemes:
i Stems: play, cat, friend ii
Affixes: -ed, -s, un-, -ly
 Two main types of affixes:
i Prefixes precede the stem: un-
ii Suffixes follow the stem: -ed, -s, un-, -ly
 Stemming = find the stem by stripping off affixes
play = play
replayed = re-play-ed
computerized = comput-er-ize-d
 Problems in morphological processing
 Inflectional morphology: inflected forms are constructed from base forms and inflectional
affixes.
 Inflection relates different forms of the same word
Lemma Singular Plural
cat cat cats
dog dog dogs
knife knife knives
sheep sheep sheep
mouse mouse mice
 Derivational morphology: words are constructed from roots (or stems) and derivational
affixes:
inter+national = international
international+ize = internationalize
internationalize+ation = internationalization
 The simplest morphological process concatenates morphs one by one, as in dis-
agree-ment-s, where agree is a free lexical morpheme and the other elements are
bound grammatical morphemes contributing some partial meaning to the whole word.
 in a more complex scheme, morphs can interact with each other, and their forms may
become subject to additional phonological and orthographic changes denoted as
morphophonemic.
The alternative forms of a morpheme are termed allomorphs.
Typology
 Morphological typology divides languages into groups by characterizing the prevalent
morphological phenomena in those languages.
 It can consider various criteria, and during the history of linguistics, different classifications
have been proposed.
 Let us outline the typology that is based on quantitative relations between words, their
morphemes, and their features:
 Isolating, or analytic, languages include no or relatively few words that would comprise more
than one morpheme (typical members are Chinese, Vietnamese, and Thai; analytic tendencies
are also found in English).
 Synthetic languages can combine more morphemes in one word and are further
divided into agglutinative and fusional languages. 
 Agglutinative languages have morphemes associated with only a single function at a 
time (as in Korean, Japanese, Finnish, and Tamil, etc.)
 Fusional languages are defined by their feature-per-morpheme ratio higher than one
(as in Arabic, Czech, Latin, Sanskrit, German, etc.). 
 In accordance with the notions about word formation processes mentioned earlier, we 
can also find out using concatenative and nonlinear:
 Concatenative languages linking morphs and morphemes one after another. 
 Nonlinear languages allowing structural components to merge nonsequentially to
apply tonal morphemes or change the consonantal or vocalic templates of words. 
 Morphological Typology
 Morphological typology is a way of classifying the languages of the world that groups
languages according to their common morphological structures.
 The field organizes languages on the basis of how those languages form words by
combining morphemes.
 The morphological typology classifies languages into two broad classes of synthetic languages
and analytical languages.
 The synthetic class is then further sub classified as either agglutinative languages or fusional
languages.
 Analytic languages contain very little inflection, instead relying on features like word order and
auxiliary words to convey meaning.
 Synthetic languages, ones that are not analytic, are divided into two
categories: agglutinative and fusional languages.
• Agglutinative languages rely primarily on discrete particles(prefixes, suffixes, and infixes) for
inflection, ex: inter+national = international, international+ize = internationalize.
• While fusional languages "fuse" inflectional categories together, often allowing one word
ending to contain several categories, such that the original root can be difficult to extract
(anybody, newspaper).
 Issues and Challenges
 Irregularity: word forms are not described by a prototypical linguistic model. 
 Ambiguity: word forms be understood in multiple ways out of the context of their
discourse. 
 Productivity: is the inventory of words in a language finite, or is it unlimited? 
 Morphological parsing tries to eliminate the variability of word forms to provide higher-
level linguistic units whose lexical and morphological properties are explicit and well
defined. 
 It attempts to remove unnecessary irregularity and give limits to ambiguity, both of
which are present inherently in human language. 
 By irregularity, we mean existence of such forms and structures that are not described 
appropriately by a prototypical linguistic model.
 Some irregularities can be understood by redesigning the model and improving its
rules, but other lexically dependent irregularities often cannot be generalized 
 Ambiguity is indeterminacy (not being interpreted) in interpretation of expressions of
language.
 Morphological modelling also faces the problem of productivity and creativity in language, by
which unconventional but perfectly meaningful new words or new senses are coined.
Irregularity
 Morphological parsing is motivated by the quest for generalization and abstraction in the
world of words.
 Immediate descriptions of given linguistic data may not be the ultimate ones, due to either
their inadequate accuracy or inappropriate complexity, and better formulations may be
needed.
 The design principles of the morphological model are therefore very important.
 In Arabic, the deeper study of the morphological processes that are in effect during inflection
and derivation, even for the so-called irregular words, is essential for mastering the whole
morphological and phonological system.
 With the proper abstractions made, irregular morphology can be seen as merely enforcing
some extended rules, the nature of which is phonological, over the underlying or prototypical
regular word forms.
 Morphophonemic templates capture morphological processes by just organizing stem
patterns and generic affixes without any context-dependent variation of the affixes or ad hoc
modification of the stems.
 The merge rules, indeed very neatly or effectively concise, then ensure that such structured
representations can be converted into exactly the surface forms, both orthographic and
phonological, used in the natural language.
 Applying the merge rules is independent of and irrespective of any grammatical parameters
or information other than that contained in a template.
 Most morphological irregularities are thus successfully removed.
Ambiguity
 Morphological ambiguity is the possibility that word forms be understood in multiple
ways out of the context of their discourse (communication in speech or writing). 
 Words forms that look the same but have distinct functions or meaning are called 
homonyms.
 Ambiguity is present in all aspects of morphological processing and language
processing at large. 
 Table arranges homonyms on the basis of their behaviour with different endings.
Systematic homonyms arise as verbs combined with endings in Korean

 Arabic is a language of rich morphology, both derivational and inflectional.

 Because Arabic script usually does not encode short vowels and omits yet some other
diacritical marks that would record the phonological form exactly, the degree of its
morphological ambiguity is considerably increased.
 When inflected syntactic words are combined in an utterance, additional phonological and
orthographic changes can take place, as shown in Figure.
 In Sanskrit, one such euphony rule is known as external sandhi.
 ‘ MY
• cases are expressed by the same word form with ‘my study’ and TEACHERS’,
but the original case endings are distinct.
 Productivity
 Is the inventory of words in a language finite, or is it unlimited?
 This question leads directly to discerning two fundamental approaches to language,
summarized in the distinction between langue and parole, or in the competence versus
performance duality by Noam Chomsky.
 In one view, language can be seen as simply a collection of utterances (parole) actually
pronounced or written (performance).
 This ideal data set can in practice be approximated by linguistic corpora, which are finite
collections of linguistic data that are studied with empirical(based on) methods and can be
used for comparison when linguistic models are developed.
 Yet, if we consider language as a system (langue), we discover in it structural devices
like recursion, iteration, or compounding(make up; constitute)that allow to produce
(competence) an infinite set of concrete linguistic utterances. 
 This general potential holds for morphological processes as well and is called
morphological productivity. 
 We denote the set of word forms found in a corpus of a language as its vocabulary. 
 The members of this set are word types, whereas every original instance of a word form is a
word token.
 The distribution of words or other elements of language follows the “80/20 rule,” also known
as the law of the vital few.
 It says that most of the word tokens in a given corpus can be identified with just a couple of
word types in its vocabulary, and words from the rest of the vocabulary occur much less
commonly if not rarely in the corpus.
 Furthermore, new, unexpected words will always appear as the collection of linguistic data is
enlarged.
 In Czech, negation is a productive morphological operation. Verbs, nouns, adjectives, and
adverbs can be prefixed with ne- to define the complementary lexical concept.
 Morphological Models
 There are many possible approaches to designing and implementing morphological models.
 Over time, computational linguistics has witnessed the development of a number of
formalisms and frameworks, in particular grammars of different kinds and expressive power,
with which to address whole classes of problems in processing natural as well as formal
languages.
 Let us now look at the most prominent types of computational approaches to morphology.

1.Dictionary Lookup
2.Finite state Morphology
3. Unification based Morphology
4.Functional Morphology
 1. Dictionary Lookup
 Morphological parsing is a process by which word forms of a language are
associated with
corresponding linguistic descriptions.
 Morphological systems that specify these associations by merely
enumerating(is the act or process of making or stating a list of things one
after another) them case by case do not offer any generalization means.
 Likewise for systems in which analyzing a word form is reduced to looking it up
verbatim in word lists, dictionaries, or databases, unless they are constructed
by and kept in sync with more sophisticated models of the language.
 In this context, a dictionary is understood as a data structure that
directly enables obtaining some precomputed results, in our case word
analyses.
 The data structure can be optimized for efficient lookup, and the
results can be shared. Lookup operations are relatively simple and
usually quick.
 Dictionaries can be implemented, for instance, as lists, binary search trees, tries, hash
tables, and so on. 
 Because the set of associations between word forms and their desired descriptions is
declared by plain enumeration, the coverage of the model is finite and the generative
potential of the language is not exploited. 
 Despite all that, an enumerative model is often sufficient for the given purpose, deals easily
with exceptions, and can implement even complex morphology.
 For instance, dictionary-based approaches to Korean depend on a large dictionary of all
possible combinations of allomorphs and morphological alternations.
 These approaches do not allow development of reusable morphological rules, though.
 2. Finite-State Morphology
 By finite-state morphological models, we mean those in which the specifications written by
human programmers are directly compiled into finite-state transducers.
 The two most popular tools supporting this approach, XFST (Xerox Finite-State Tool) and
LexTools.
 Finite-state transducers are computational devices extending the power of finite-state
automata.
 They consist of a finite set of nodes connected by directed edges labeled with pairs of input
and output symbols.
 In such a network or graph, nodes are also called states, while edges are called arcs.
 Traversing the network from the set of initial states to the set of final states along the arcs is
equivalent to reading the sequences of encountered input symbols and writing the sequences
of corresponding output symbols.
 The set of possible sequences accepted by the transducer defines the inpu
language; the set of possible sequences emitted by the transducer defines the outpu
language. 
Input Input Morphological parsed output
Cats cat +N +PL
Cat cat +N +SG
Cities city +N +PL
Geese goose +N +PL
Goose goose +N +SG) or (goose +V)
Gooses goose +V +3SG
mergin merge +V +PRES-PART
g
Caught (caught +V +PAST-PART) or (catch +V +PAST)
Finite state automata for cats,dogs,fox
 For example, a finite-state transducer could translate the infinite regular language consisting of the
words vnuk, pravnuk, prapravnuk, ... to the matching words in the infinite regular language defined
by grandson, great-grandson, great-great-grandson. 
 In finite-state computational morphology, it is common to refer to the input word forms as surface strings and to
the output descriptions as lexical strings, if the transducer is used for morphological analysis, or vice versa, if it is
used for morphological generation. 
• In English, a finite-state transducer could analyze the surface string children into the lexical
string child [+plural], for instance, or generate women from woman [+plural].
 Relations on languages can also be viewed as functions. Let us have a relation R, and let us denote by [Σ] the set
of all sequences over some set of symbols Σ, so that the domain and the range of R are subsets of [Σ]. 
 We can then consider R as a function mapping an input string into a set of output strings, formally denoted by this
type signature, where [Σ] equals String: 

 A theoretical limitation of finite-state models of morphology is the problem of capturing reduplication of words or
their elements (e.g., to express plurality) found in several human languages. 
 Finite-state technology can be applied to the morphological modeling of isolating and agglutinative languages in a 
quite straightforward manner. Korean finite-state models are discussed by Kim, Lee and Rim, and Han, to mention
a few.
3. Unification-Based Morphology

 The concepts and methods of these formalisms are often closely connected to those
of logic programming. 
 In finite-state morphological models, both surface and lexical forms are by themselves 
unstructured strings of atomic symbols.
 In higher-level approaches, linguistic information is expressed by more appropriate 
data structures that can include complex values or can be recursively nested if
needed.
 Morphological parsing P thus associates linear forms φ with alternatives of structured
content ψ, cf. 

 Erjavec argues that for morphological modelling, word forms are best captured by
regular expressions, while the linguistic content is best described through typed
feature structures. 
 Feature structures can be viewed as directed acyclic graphs. 
 A node in a feature structure comprises a set of attributes whose values can be 
 Nodes are associated with types, and atomic values are attributeless nodes
distinguished by their type. 
 Instead of unique instances of values everywhere, references can be used to establish 
value instance identity.
 Feature structures are usually displayed as attribute-value matrices or as nested
symbolic expressions. 
 Unification is the key operation by which feature structures can be merged into a more 
informative feature structure.
 Unification of feature structures can also fail, which means that the information in them
is mutually incompatible. 
 Morphological models of this kind are typically formulated as logic programs, and 
unification is used to solve the system of constraints imposed by the model.
 Advantages of this approach include better abstraction possibilities for developing a
morphological grammar as well as elimination of redundant information from it. 
 Unification-based models have been implemented for Russian, Czech, Slovene, 
Persian, Hebrew, Arabic, and other languages.
4. Functional Morphology
 Functional morphology defines its models using principles of functional programming
and type theory. 
 It treats morphological operations and processes as pure mathematical functions and 
organizes the linguistic as well as abstract elements of a model into distinct types of
values and type classes.
 Though functional morphology is not limited to modelling particular types of
morphologies in human languages, it is especially useful for fusional morphologies. 
 Linguistic notions like paradigms, rules and exceptions, grammatical categories and 
parameters, lexemes, morphemes, and morphs can be represented intuitively(without
conscious reasoning; instinctively) and succinctly(in a brief and clearly expressed
manner) in this approach.
 Functional morphology implementations are intended to be reused as programming 
libraries capable of handling the complete morphology of a language and to be
incorporated into various kinds of applications.
 Morphological parsing is just one usage of the system, the others being
morphological generation, lexicon browsing, and so on. 
 we can describe inflection I, derivation D, and lookup L as functions of these generic
type 

 Many functional morphology implementations are embedded in a general-purpose

programming language, which gives programmers more freedom with advanced 
programming techniques and allows them to develop full-featured, real-world
applications for their models.
 The Zen toolkit for Sanskrit morphology is written in OCaml. 
 It influenced the functional morphology framework in Haskell, with which
morphologies of Latin, Swedish, Spanish, Urdu, and other languages have been
implemented. 
 In Haskell, in particular, developers can take advantage of its syntactic flexibility and 
design their own notation for the functional constructs that model the given problem.
 The notation then constitutes a so-called domain-specific embedded language, which makes programming even
more fun. 
 Even without the options provided by general-purpose programming languages, functional morphology models
achieve high levels of abstraction. 
 Morphological grammars in Grammatical Framework can be extended with descriptions of the syntax  and
semantics of a language.
 Grammatical Framework itself supports multilinguality, and models of more than a dozen languages are available in
it as open-source software. 
 2.Finding Structure of documents
Introduction
 In human language, words and setences do not appear randomly but have structure.
 For example, combinations of words from sentences- meaningful grammatical units, such as
statements, requests, and commands.
 Automatic extraction of structure of documents helps subsequent NLP tasks: for example,
parsing, machine translation, and semantic role labelling use sentences as the basic
processing unit.
 Sentence boundary annotation(labelling) is also important for aiding human readability of
automatic speech recognition (ASR) systems.
 Task of deciding where sentences start and end given a sequence of characters(made of words
and typographical cues) sentences boundary detection.
 Topic segmentation as the task of determining when a topic starts and ends in a sequence of
sentences.
 The statistical classification approaches that try to find the presence of sentence and topic
boundaries given human-annotated training data, for segmentation.
 These methods base their predictions on features of the input: local characteristics that give
evidence toward the presence or absence of a sentence, such as a period(.), a question
mark(?), an exclamation mark(!), or another type of punctuation.
 Features are the core of classification approaches and require careful design and selection in
order to be successful and prevent overfitting and noise problem.
 Most statistical approaches described here are language independent, every language is a
challenging in itself.
 For example, for processing of Chinese documents, the processor may need to first segment
the character sequences into words, as the words usually are not separated by a space.
 Similarly, for morphological rich languages, the word structure may need to be analyzed to
extract additional features.
 Such processing is usually done in a pre-processing step, where a sequence of tokens is
determined.
 Tokens can be word or sub-word units, depending on the task and language.
 These algorithms are then applied on tokens.
 SENTENCE BOUNDARY DETECTION
 Sentence boundary detection (Sentence segmentation) deals with automatically segmenting
a sequence of word tokens into sentence units.
 In written text in English and some other languages, the beginning of a sentence is usually
marked with an uppercase letter, and the end of a sentence is explicitly marked with a
period(.), a question mark(?), an exclamation mark(!), or another type of punctuation.
 In addition to their role as sentence boundary markers, capitalized initial letters are used
distinguish proper nouns, periods are used in abbreviations, and numbers and punctuation
marks are used inside proper names.
 The period at the end of an abbreviation can mark a sentence boundary at the same time.
 Example: I spoke with Dr. Smith. and My house is on Mountain Dr.
 In the first sentence, the abbreviation Dr. does not end a sentence, and in the second it does.
 Especially quoted sentences are always problematic, as the speakers may have uttered
multiple sentences, and sentence boundaries inside the quotes are also marked with
punctuation marks.
 An automatic method that outputs word boundaries as ending sentences according to the
presence of such punctuation marks would result in cutting some sentences incorrectly.
 Ambiguous abbreviations and capitalizations are not only problem of sentence segmentation
in written text.
 Spontaneously written texts, such as short message service (SMS) texts or instant
messaging(IM) texts, tend to be nongrammatical and have poorly used or missing
punctuation, which makes sentence segmentation even more challenging.
 Similarly, if the text input to be segmented into sentences comes from an automatic system,
such as optical character recognition (OCR) or ASR, that aims to translate images of
handwritten, type written, or printed text or spoken utterances into machine editable text,the
finding of sentences boundaries must deal with the errors of those systems as well.
 On the other hand, for conversational speech or text or multiparty meetings with
ungrammatical sentences and disfluencies, in most cases it is not clear where the boundaries
are.
 Code switching -that is, the use of words, phrases, or sentences from multiple languages by
multilingual speakers- is another problem that can affect the characteristics of sentences.
 For example, when switching to a different language, the writer can either keep the
punctuation rules from the first language or resort to the code of the second language.
 Conventional rule-based sentence segmentation systems in well-formed texts rely on patterns
to identify potential ends of sentences and lists of abbreviations for disambiguating them.
 For example, if the word before the boundary is a known abbreviation, such as “Mr.” or “Gov.,”
the text is not segmented at that position even though some periods are exceptions.
 To improve on such a rule-based approach, sentence segmentation is stated as a classification
problem.
 Given the training data where all sentence boundaries are marked, we can train a classifier to
recognize them.
Topic Boundary Detection
 Segmentation(Discourse or text segmentation) is the task of automatically dividing a stream
of text or speech into topically homogenous blocks.
 This is, given a sequence of(written or spoken) words, the aim of topic segmentation is to
find the boundaries where topics change.
 Topic segmentation is an important task for various language understanding applications, such
as information extraction and retrieval and text summarization.
 For example, in information retrieval, if a long documents can be segmented into shorter,
topically coherent segments, then only the segment that is about the user’s query could be
retrieved.
 During the late1990s, the U.S defence advanced research project agency(DARPA) initiated the
topic detection and tracking program to further the state of the art in finding and following
new topic in a stream of broadcast news stories.
 One of the tasks in the TDT effort was segmenting a news stream into individual stories.
Methods
 Sentence segmentation and topic segmentation have been considered as a boundary
classification problem.
 Given a boundary candidate( between two word tokens for sentence segmentation and
between two sentences for topic segmentation), the goal is to predict whether or not c
andidate is an actual boundary (sentence or topic boundary). the
 Formally, let xƐX be the vector of features (the observation) associated with a candidate and y
ƐY be the label predicted for that candidate.
. ഥ
 The label y can be b for boundary and 𝒃 for nonboundary.
 Classification problem: given a set of training examples(x,y)train, find a function that will assign
the most accurate possible label y of unseen examples xunseen.
 Alternatively to the binary classification problem, it is possible to model boundary types using
finer-grained categories.
 For segmentation in text be framed as a three-class problem: sentence boundary ba, without
an abbreviation and abbreviation not as a boundary
 Similarly spoken language, a three way classification can be made between non-boundaries
statements bs, and question boundaries bq .
• For sentence or topic segmentation, the problem is defined as finding the most probable
sentence or topic boundaries.
• The natural unit of sentence segmentation is words and of topic segmentation is sentence, as
we can assume that topics typically do not change in the middle of a sentences.
 The words or sentences are then grouped into categories stretches belonging to one
sentences or topic- that is word or sentence boundaries are classified into sentences or topic
boundaries and -non-boundaries.
 The classification can be done at each potential boundary i (local modelling); then, the aim is
to estimate the most probable boundary type yෝ ª ªiforeach candidate xi
𝑎𝑟𝑔𝑚𝑎𝑥 𝑃 𝑦
𝑦ෝ ª s 𝑦𝑖 𝑖𝑛 𝑌
𝑖 𝑥𝑖
=
Here, the ^ is used to denote estimated categories, and a variable without a ^ is used to show
possible categories.
 In this formulation, a category is assigned to each example in isolation; hence, decision is
made locally.
 However, the consecutive types can be related to each other. For example, in broadcast news
speech, two consecutive sentences boundaries that form a single word sentence are very
infrequent.
 In local modelling, features can be extracted from surrounding example context of the
candidate boundary to model such dependencies.
• It is also possible to see the candidate boundaries as a sequence and search for the sequence of boundary types
that have the maximum probability given the candidate examples,

𝑌෠= 𝑎 𝑟𝑔 𝑚 𝑎 𝑥
𝑃 𝑌𝑋
𝑦

 We categorize the methods into local and sequence classification.

 Another categorization of methods is done according to the type of the machine learning algorithm: generative versus
discriminative.
 Generative sequence models estimate the joint distribution of the observations P(X,Y) (words, punctuation) and the
labels(sentence boundary, topic boundary).

 Discriminative sequence models, however, focus on features that categorize the differences between the labelling of that
examples.
 GENERATIVE SEQUENCE CLASSIFICATION METHODS

 Most commonly used generative sequence classification method for topic and
sentence is the hidden Markov model (HMM). 
 The probability in equation 2.2 is rewritten as the following, using the Bayes rule:

Here 𝑌෠= Predicted class(boundary) label

Y = (y1,y2,….yk )= Set of class(boundary) labels
X = (x1,x2,….xn)= set of feature vectors
P(Y|X) = the probability of given the X (feature vectors), what is the probability of X
belongs to the class(boundary) label.
P(x) = Probability of word sequence
P(Y) = Probability of the class(boundary)
 P(X) in the denominator is dropped because it is fixed for different Y and hence does not
change the argument of max.
 P(X|Y) and P(Y) can be estimated as
 Discriminative Local Classification Methods
 Discriminative classifiers aim to model P(yi |xi) equation 2.1 directly.
 The most important distinction is that whereas class densities P(x|y) are model assumptions
in generative approaches, such as naïve Bayes, in discriminative methods, discriminant
functions of the feature space define the model.
 A number of discriminative classification approaches, such as support vector machines,
boosting, maximum entropy, and regression. Are based on very different machine learning
algorithms.
 While discriminative approaches have been shown to outperform generative methods in
many speech and language processing tasks.
 For sentence segmentation, supervised learning methods have primarily been applied to
newspaper articles.
 Stamatatos, Fakotakis and Kokkinakis used transformation based learning (TBL) to infer rules
for finding sentence boundaries.
 Many classifiers have been tried for the task: regression trees, neural networks, classification
trees, maximum entropy classifiers, support vector machines, and naïve Bayes classifiers.
 The most Text tiling method Hearst for topic segmentation uses a lexical cohesion metric in a
word vector space as an indicator of topic similarity.
 Figure depicts a typical graph of similarity with respect to consecutive segmentation units.

 The document is chopped when the similarity is below some threshold.

 Originally, two methods for computing the similarity scores were proposed: block
comparison and vocabulary introduction.
 The first, block comparison, compares adjacent blocks of text to see how similar they are
according to how many words the adjacent blocks have in common.
 Given two blocks, b1 and b2, each having k tokens (sentences or paragraphs), the similarity
(or topical cohesion) score is computed by the formula:

 Where wt,b is the weight assigned to term t in block b.

 The weights can be binary or may be computed using other information retrieval- metrics
such as term frequency.
 The second, the vocabulary introduction method, assigns a score to a token-sequence gap
on the basis of how many new words are seen in the interval in which it is the midpoint.
 Similar to the block comparison formulation, given two consecutive blocks b1 and b2, of
equal number of words w, the topical cohesion score is computed with the following formula:

 Where NumNewTerms(b) returns the number of terms in block b seen the first time in text.
2.2.3Discriminative Sequence Classification Methods
 In segmentation tasks, the sentence or topic decision for a given example(word, sentence,
paragraph) highly depends on the decision for the examples in its vicinity.
 Discriminative sequence classification methods are in general extensions of local
discriminative models with additional decoding stages that find the best assignment of labels
by looking at neighbouring decisions to label.
 Conditional random fields(CRFs) are extension of maximum entropy, SVM struct is an
extension of SVM, and maximum margin Markov networks(M3N) are extensions of HMM.
 CRFs are a class of log-linear models for labelling structures.
 Contrary to local classifiers that predict sentences or topic boundaries independently, CRFs
can oversee the whole sequence of boundary hypotheses to make their decisions.
Complexity of the Approaches
 The approaches described here have advantages and disadvantages.
 In a given context and under a set of observation features, one approach may be better than
other.
 These approaches can be rated in terms of complexity (time and memory) of their training
and prediction algorithms and in terms of their performance on real-world datasets.
 In terms of complexity, training of discriminative approaches is more complex than training
of generative ones because they require multiple passes over the training data to adjust for
feature weights.
 However, generative models such as HELMs can handle multiple orders of magnitude larger
training sets and benefits, for instance, from decades of news wire transcripts.
 On the other hand, they work with only a few features (only words for HELM) and do not
cope well with unseen events.
1. List and explain the challenges of morphological models. Mar 2021 [7]
2. Discuss the importance and goals of Natural Language Processing. Mar 2021 [8]
3. List the applications and challenges in NLP. Sep 2021 [7]
4. Explain any one Morphological model. Sep 2021 [8]
5. Discuss about challenging issues of Morphological model. Sep 2021 [7]
6. Differentiate between surface and deep structure in NLP with suitable examples. Sep 2021
[8] 7.Give some examples for early NLP systems. Sep 2021 [7]
8. Explain the performance of approaches in structure of documents? Sep 2021 [15]
9. With the help of a neat diagram, explain the representation of syntactic structure. Mar 2021 [8]
10.Elobarate the models for ambiguity resolution in Parsing. Mar 2021 [7]
11.Explain various types of parsers in NLP? Sep 2021 [8]
12.Discuss multilingual issues in detail. Sep 2021 [7]
13. Given the grammar S->AB|BB, A->CC|AB|a, B->BB|CA|b, C->BA|AA|b, word w=‘aabb’. Applay top down parsing test, word
can be generated or not. Sep 2021 [8]
14. Explain Tree Banks and its role in parsing. Sep 2021 [7]
List the applications in NLP.
Applications of NLP:
• Information retrieval & web search
• Grammar correction & Question answering
• Sentiment Analysis.
• Text Classification.
• Chatbots & Virtual Assistants.
• Text Extraction.
• Machine Translation.
• Text Summarization.
• Market Intelligence.
• Auto-Correct.
Discuss the importance and goals of Natural Language Processing.