Core Chunking Terms

Chunking: The process of grouping words in a sentence into syntactically related phrases (chunks), such as noun phrases (NP), verb phrases (VP), and prepositional phrases (PP), without building a full parse tree.

Chunk: A group of words that function together as a single syntactic unit. Example: "the big dog" (NP), "ran quickly" (VP).

IOB Tagging: A scheme for marking chunk boundaries:

• B- (Beginning): Marks the first word of a chunk
• I- (Inside): Marks subsequent words in the same chunk
• O (Outside): Marks words not part of any chunk

Noun Phrase (NP): A chunk containing a noun and its modifiers. Example: "the red ball"

Verb Phrase (VP): A chunk containing a verb and its auxiliaries or complements. Example: "is running fast"

Prepositional Phrase (PP): A chunk starting with a preposition and its object. Example: "in the park"

Adverbial Phrase (ADVP): A chunk containing adverbs and their modifiers. Example: "very quickly"

Adjectival Phrase (ADJP): A chunk containing adjectives and their modifiers. Example: "extremely happy"

Cross-Linguistic Terms

Postposition: A word that follows a noun to indicate grammatical relationships, common in Hindi. Example: "घर में" (in the house)

SOV Order: Subject-Object-Verb word order, typical in Hindi.

SVO Order: Subject-Verb-Object word order, typical in English.

Analysis Terms

Chunk Boundary: The point in a sentence where one chunk ends and another begins.

Gold Standard Annotation: The correct chunking of a sentence, used for evaluation.

Chunking Accuracy: The percentage of correctly identified chunks compared to the gold standard.

Phrase Structure: The hierarchical organization of words into chunks and phrases.

Shallow Parsing: Another term for chunking, focusing on phrase-level analysis rather than full syntactic parsing.