chapter 1 Flashcards

1
Q

What is NLP

A

NLP stands for Natural Language Processing. It is described as a branch of Computer Science that combines linguistics and computer science to enable machines to read, understand, and process human languages.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the primary goal of NLP regarding communication for the speaker (machine)?

A

A3: The goal is to facilitate communication for the speaker (machine) by addressing intention, content selection, generation (translating information from the language of thought into words), and synthesis (outputting the string in the desired mode, such as text or speech).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Outline the steps involved in communication for the listener (human) in NLP.

A

A4: Communication for the listener involves perception (mapping input through optical character recognition or speech recognition), analysis (syntactic interpretation, semantic interpretation, pragmatic interpretation), and incorporation (deciding whether to believe the string or not).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the purpose of word segmentation in NLP, and why is it particularly relevant in the Chinese language?

A

Word segmentation involves breaking a string of characters into words. In the Chinese language, this task is crucial because Chinese does not use white spaces to separate words. For example, “myspace.com/wings” is segmented into “myspace .com wings.”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is morphological analysis, and how does it relate to Parts Of Speech (POS)?

A

Morphological analysis involves segmenting words into morphemes. Morphemes are the smallest linguistic units with semantic meaning. This task is related to POS, where examples like “Carried” are segmented into “carry + ed” to identify verb forms.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Explain the purpose of POS tagging in syntactic tasks.

A

POS tagging annotates each word with its Part Of Speech, such as noun, verb, adjective, etc. It helps identify the grammatical category of each word, aiding in syntactic analysis.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is phrase chunking, and how does it build upon POS tagging in NLP?

A

Phrase chunking, building upon POS tagging, provides chunks as output. It groups words into phrases like noun phrases (NP) or verb phrases (VP). For instance, “South Africa” can be chunked as a single word instead of separate words.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Describe the goal of syntactic parsing in NLP.

A

Syntactic parsing produces the correct syntax tree for a sentence. It determines how words group together as phrases and identifies subjects or objects of verbs. The typical sentence structure is Noun Phrase + Verb Phrase + Prepositional Phrase.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the purpose of Word Sense Disambiguation (WSD) in semantic tasks?

A

WSD determines the proper sense of ambiguous words in a sentence. For example, it discerns whether “interest” in “Ellen has a strong interest” refers to curiosity or financial investment.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Explain Semantic Role Labeling (SRL) and its role in NLP.

A

SRL determines the semantic role played by a noun phrase as an argument to the verb. It provides insights into the relationships between words and their functions in a sentence.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

How does semantic parsing contribute to natural language understanding?

A

Semantic parsing maps natural language text to formal logical representations based on the task’s inference requirements. For example, it transforms a question like “Who was the first person to walk on the moon?” into a SQL query.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is the goal of Anaphora Resolution in pragmatic/discourse tasks?

A

Anaphora Resolution identifies which phrases in a document refer to the same thing. It resolves references to maintain clarity and coherence in discourse.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What does Ellipses Resolution aim to achieve, and how does it leverage context?

A

Ellipses Resolution infers words omitted from a sentence using context. It helps complete the meaning by filling in the missing information based on the surrounding context.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

How does Information Extraction (IE) contribute to NLP tasks?

A

IE identifies entities and their relations from a text. Named Entity Recognition (NER) is a specific aspect that identifies names, people, places, organizations, etc.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is the primary goal of Relation Extraction in NLP?

A

Relation Extraction identifies relationships between entities. It focuses on extracting meaningful connections between different elements mentioned in the text.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

How does Question Answering differ from other NLP tasks, and what does it directly address?

A

Question Answering directly answers natural language questions based on information presented in corpora or textual documents. It involves comprehending queries and providing relevant responses.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

_______ deals with linguistic sounds

A

Phonology

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

________ explores the components of a word, such as roots, prefixes, and suffixes, and investigates how they combine to form meaningful units.

A

Morphology

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

_______ is concerned with the structural relationships between words in a sentence,

A

Syntax

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Semantics focuses on the meaning of words,

A

t

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

_______ explores the meaning of a word depending on context.

A

Pragmatics

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

______ refers to the meaning of a coherent group of sentences.

A

Discourse

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

What characterizes ambiguity in linguistics, and when does it occur in language?

A

Ambiguity occurs when an input has multiple alternative linguistic structures, leading to various interpretations.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

_______ involves determining the meaning of every word in context.

A

Lexical disambiguation

25
Q

What are Models?

A

Models and theories in NLP are drawn from the standard toolkits of mathematics, computer science, and linguistics.

26
Q

What are the key components of state machines in the context of models in NLP?

A

State machines in NLP consist of states, transitions among states, and inputs. Examples include Deterministic Finite State Automata and Non-Deterministic Finite State Automata.

27
Q

How do rule systems contribute to linguistic analysis in NLP models?

A

Rule systems, as declarative counterparts to state machines, use predefined linguistic rules to analyze and process textual data. They can be probabilistic or non-probabilistic.

28
Q

example of SM?

A

Deterministic finite state automata, Non-Deterministic finite state automata …

29
Q

example of RS

A
  • Regular grammars and regular relations- Context free and feature-augmented grammars
30
Q

SMs and RSs are the main tools used for dealing with _______

A

syntax, morphologyand phonology.

31
Q

What characterizes logic-based models in the context of NLP?

A

Logic-based models use formal logic to represent and reason about language. They rely on explicit rules or axioms, making them transparent and interpretable compared to statistical models.

32
Q

example of logic models

A

● Predicated calculus● Lambda calculus● Feature structure● Semantic primitives

33
Q

Traditionally logic based models were used to model _____ and recently _______

A
  • semantics and pragmatics- non-logical lexical semantics
34
Q

Other models, such as state machines, formal rule systems, and logic, can be augmented with probabilities to enhance their capabilities.

A

true

35
Q

The state machine can be augmented with probabilities to become_______

A

a weighted automaton or Markov model.

36
Q

How do probabilistic models contribute to NLP?

A
  • by providing probabilities for different linguistic choices.- Probabilistic models address ambiguous input by choosing the most probable interpretation among N choices.
37
Q

Vector-space models are based on _____

A

linear algebra

38
Q

Vector-space models play a significant role in ________

A

information retrieval

39
Q

In speech recognition, we search through a space of ________

A

phone sequences.

40
Q

In parsing, we search through a space of ____

A

trees.

41
Q

In machine translation, we search through a space of _______

A

translation hypotheses.

42
Q

For non-probabilistic tasks, such as state machines, we use _______

A

graph algorithms such as depth-first search

43
Q

For probabilistic tasks, we use heuristic variants such as ______ and ______

A

best-first and A* search

44
Q

What are the key types of machine learning tools, and how do they operate in language processing?

A
  • classifiers- and sequence models,
45
Q

_______ attempts to assign a single object to a single class.

A

Classifer -

46
Q

How do classifiers operate in language processing?

A

Classifiers in language processing, such as decision trees, support vector machines, and logistic regression, assign a single object to a single class. For example, they could decide whether a word is spelled correctly or not, making a binary decision for each word.

47
Q

What is the goal of sequence models in language processing?

A

Sequence models attempt to jointly classify a sequence of objects into a sequence of classes. Examples include hidden Markov models, maximum entropy Markov models, and conditional random fields.

48
Q

_____ is the application of distinct training and test sets, statistical techniques like cross-validation, and rigorous evaluation of trained systems.

A

Methodological training

49
Q

Effective use of language is intertwined with our general cognitive abilities.

A

t

50
Q

How does the effective use of language intertwine with general cognitive abilities, and what is the significance of the Turing Test in this context?

A

Effective use of language is intertwined with our general cognitive abilities. The Turing Test, involves an interrogator interacting with a human and a computer (participant), aiming to determine if the computer can process language as skillfully as a human.

51
Q

________ is an early NLP system capable of carrying on a limited form of conversation with a user.

A

The ELIZA program

52
Q

How is advanced NLP technology incorporated into services provided by Amtrak, United Airlines, Mercedes-Benz, and others?

A

Travelers using services like Amtrak and United Airlines interact with conversational agents. Luxury car manufacturers, including Mercedes-Benz, incorporate advanced NLP features such as speech recognition and text-to-speech systems.

53
Q

What services does Blinkx enable?

A

Blinkx leverages speech recognition technology to enable video search services.

54
Q

What cross-language information retrieval and translation services are provided by Google?

A

enables users to search and retrieve information in languages other than their own, as well as translating content between languages.

55
Q

By the late 1950s and early 1960s, speech and language processing was divided into two paradigms: ______ and _______

A

symbolic and stochastic.

56
Q

_______ Mainly embraced by statistics and electrical engineering departments.

A

stochastic Paradigm:

57
Q

______ Emerged from Chomsky’s formal language theory and generative syntax.

A

Symbolic Paradigm:

58
Q

what are the Four Paradigms in NLP: 1970–1983?

A
  • Stochastic Paradigm:- Logic-Based Paradigm:- Natural Language Understanding Paradigm:- Discourse Modeling Paradigm: