(SEM VIII) THEORY EXAMINATION 2024-25 NATURAL LANGUAGE PROCESSING

SECTION A – (2 Marks Each) – Paragraph Style Answers

a) What is Natural Language Processing (NLP)? Mention its applications.

Natural Language Processing, commonly known as NLP, is a branch of Artificial Intelligence that enables computers to understand, analyze, and interact using human language. It allows machines to read text, hear speech, interpret meaning, and generate responses in a natural way. NLP is widely used in applications such as chatbots, voice assistants, machine translation, sentiment analysis, speech recognition systems, text summarization, and spam email detection.

b) Explain the phases of Natural Language Processing.

Natural Language Processing is carried out through several important phases. It begins with lexical analysis, which breaks text into words or tokens. This is followed by syntactic analysis, where sentence structure and grammar are analyzed. Semantic analysis then focuses on understanding the meaning of words and sentences. Discourse integration helps interpret meaning across sentences, while pragmatic analysis considers real-world context and intent behind language.

c) Explain the key steps involved in text normalization.

Text normalization is the process of converting raw text into a standardized form suitable for processing. It includes tokenization to split text into words, converting text to lowercase, removing unnecessary stop words, and reducing words to their base form using stemming or lemmatization. These steps help improve accuracy and efficiency in NLP tasks.

d) Describe techniques for text representation and embedding.

Text representation techniques convert textual data into numerical form so that machines can process it. Traditional methods include Bag of Words and TF-IDF, which represent text based on word frequency. Advanced embedding techniques such as Word2Vec, GloVe, and FastText capture semantic relationships between words. Modern contextual embeddings like BERT understand word meaning based on surrounding context.

e) Give an example of a regular expression used in NLP.

Regular expressions are patterns used to match and extract information from text. For example, the expression \\d{10} is used to identify and extract a 10-digit mobile number from a document. Regular expressions are commonly used for text cleaning, pattern matching, and information extraction in NLP.

f) Differentiate between NLP and NLU.

Natural Language Processing focuses on processing and structuring language data, such as tokenization and parsing. Natural Language Understanding goes a step further by interpreting meaning, intent, and context. While NLP handles how language is processed, NLU focuses on what the language actually means.

g) What is Natural Language Understanding (NLU)? Mention its applications.

Natural Language Understanding is a subfield of NLP that enables machines to comprehend human language in terms of meaning and intent. It helps systems understand what users are trying to say. NLU is used in applications such as virtual assistants, intent classification systems, question-answering systems, chatbots, and dialogue management systems.

h) Explain the challenges faced in Natural Language Understanding.

Natural Language Understanding faces several challenges due to the complexity of human language. These include ambiguity of words, understanding context, recognizing sarcasm and emotions, handling multiple meanings of sentences, and processing informal or noisy text. These challenges make accurate language understanding difficult for machines.

i) Explain pre-trained language models and deep learning techniques.

Pre-trained language models are trained on large text datasets and reused for various NLP tasks. Examples include BERT, GPT, and RoBERTa. These models use deep learning techniques such as Recurrent Neural Networks, LSTM, GRU, Convolutional Neural Networks, and Transformers to capture complex language patterns and improve performance.

j) Explain how NLU works.

Natural Language Understanding works by analyzing text to identify intent, entities, and context. It uses linguistic rules, machine learning models, and deep learning techniques to understand sentence structure and meaning. By combining syntax, semantics, and contextual information, NLU systems are able to respond accurately to user input.

 SECTION B – (10 Marks Each) – Descriptive Answers

a) Explain Knowledge Representation and Semantics in NLP.

Knowledge representation refers to the way information is structured so that machines can understand and reason with it. In NLP, semantics plays a key role by focusing on the meaning of words, phrases, and sentences. Techniques such as semantic networks, frames, ontologies, and logic-based representations are used to store and interpret meaning. These methods help NLP systems understand relationships between concepts, making language interpretation more accurate.

b) Explain lexical semantics with an example.

Lexical semantics deals with the study of word meanings and the relationships between words. It examines concepts such as synonymy, antonymy, homonymy, and polysemy. For example, the word “bank” can refer to a financial institution or the edge of a river. The correct meaning is determined by context. Lexical semantics helps machines resolve such ambiguities.

c) Explain semantics and the components of semantic analysis.

Semantics is the study of meaning in language. Semantic analysis focuses on understanding word meanings, sentence meanings, and relationships between entities. Its components include word sense disambiguation, semantic role labeling, and contextual interpretation. Semantic analysis ensures that machines correctly understand what a sentence is conveying.

d) Explain basic units of a semantic system and techniques used.

The basic units of a semantic system include words, concepts, and relationships between concepts. These units help represent meaning in a structured way. Techniques used in semantic systems include rule-based methods, statistical approaches, and neural semantic models. These techniques help machines interpret meaning efficiently.

e) Explain Machine Translation.

Machine Translation is the process of automatically converting text from one language to another using computer systems. It consists of three main components: analysis of the source language, transfer of meaning, and generation of the target language. Machine Translation is widely used in applications such as Google Translate, multilingual websites, and global communication platforms.

SECTION C – (10 Marks Each) – Long Answers

a) Explain Augmented Grammars.

Augmented grammars are an extension of context-free grammars that include additional semantic rules and attributes. These grammars attach meaning to syntactic structures, enabling better interpretation of language. They are widely used in NLP to integrate syntax and semantics, improving accuracy during parsing and interpretation.

b) Explain parsing and compare Top-Down and Bottom-Up parsing.

Parsing is the process of analyzing the grammatical structure of a sentence according to language rules. Top-down parsing starts from the root symbol and works toward the input sentence, predicting structures in advance. It is simple but inefficient due to backtracking. Bottom-up parsing starts with input words and builds the parse tree upward. It is more efficient and handles ambiguity better, though it is more complex to implement.