SEMANTIC FEATURES AND SELECTION RESTRICTIONS

3. SEMANTIC FEATURES IN SYSTEMS OF NATURAL LANGUAGE PROCESSING (NLP)

Semantic Features in NLP

Crucial Role

Semantic features play a crucial role in natural language processing (NLP).

Predicate-Argument Relations

Contribute to revealing predicate-argument relations in parsing algorithms.

Disambiguation

Aid in disambiguating lexically homonymous words

Combinability

To assist in the combinability of verbs with

Adverbials

Coordinated constructions

Anaphoric Relations

Help identify anaphoric relations in text

Interpretations

A distinction is made between:

Literal interpretations

Deviant or metaphorical meanings

Overall Importance

Essential for an analysis either:

Syntactic

semantic analysis in NLP

4. SEMANTIC FEATURES AND SELECTION RESTRICTIONS IN LEXICON AND GRAMMAR

Semantic features were a primary tool for semantic analysis in the early 1960s

but took a backseat in the 1970s and 1980s with advancements in semantic theory.

The notion of semantic features is gaining prominence again

particularly in understanding selection restrictions in lexicon and grammar.

Anna Wierzbicka suggests that grammatical distinctions are motivated by

semantic distinctions

arguing that semantic features play a leading role in regulating selection restrictions.

Examples demonstrate how seemingly syntactic selection restrictions can be semantically motivated

such as predicates allowing Neg-Raising

or the distribution of Russian conjunctions "что" and "как" based on semantic components.

Semantic invariants are identified for predicates capable of introducing indirect questions

showing the influence of semantic components like "X knows" in linguisticstructures.

Examples:

Neg-Raising predicates

have semantic features like

[+Incompatibility of contraries]

[+Excluded neutrality].

Russian conjunctions что and как are used based on semantic components;

что follows verbs with 'know/believe',

while как follows words with 'perceive'.

Predicates introducing indirect questions or parameters are determined by the semantic component'Xknows'.

5. ON SEMANTIC IN-VARIANT OF THE CLASS OF WORDS WITH GENITIVE SUBJECT

Semantic features were crucial for semantic analysis in the 60s

but have since taken a secondary position in natural language processing.

Semantic features

can label components in lexical meanings and play a role in semantic decomposition.

Semantic features

may regulate selection restrictions in lexicon and grammar more than syntactic features.

Anna Wierzbicka

argues that all selection restrictions in grammar can be motivated by semantic features.

Examples demonstrate

how semantic components influence selection restrictions in Russian grammar.

Negation of existence and presence in the field of vision

are key semantic components influencing genitive subject constructions.

Different semantic components account for the choice of genitive subjects in negative sentences.

The semantic invariant of genitive verbs allows for the characterization of semantic classes.

There is a connection between selectional restrictions and semantic features of words.

The presence vs. absence of semantic components influences the grammatical structureofsentences.

Examples:

Examples of sentences with genitive subjects in Russian:

"Ответа не пришло" (No answer came), "Мороза не чувствуется" (The frost is not felt)

"Катастрофы не произошло" (No catastrophe happened).

Semantic explanation

for the choice of case in genitive subject construction.

Semantic components

determining the selection of genitive subjects: 'X exists' and 'X is present in the field of vision of an observer'.

Different meanings

conveyed by genitive and nominative subjects in negative sentences.

Factors influencing the choice of genitive subject

animate vs. inanimate subject, referentiality, topic-focus articulation, presence vs. absence oftheobserver.

1. LEXICAL DATABASE OF THE SYSTEM

Main characteristics

Lexical Database (LBD) and Bibliographical Database (BBD)

Are key components of the Lexicographer expert system for natural language processing.

LBD contains machine-readable vocabulary

Morphological

Syntactic

Semantic

Prosodic

and referential features of lexical items.

Semantic features

speech act verbs

performative verbs

verbs of motion

kinship terms

body parts

BBD includes bibliographic information

Individual lexemes

unlike existing bibliographic catalogs.

The vocabulary comprises approximately 12,500 words

with morphological information sourced from Zalizniak's dictionary.

Syntactic and semantic information is often absent in existingdictionaries.

2. SEMANTIC FEATURE ACCORDING TO U.WEINREICH

Semantic Features

Central Role

Semantic features play a central role in the paper.

U. Weinreich's Contributions

Proposed a distinction between paradigmatic and transfer semantic features

Semantic features serve multiple purposes.

Explain deviant and metaphorical readings in language.

Impose provisional semantic contents on potentially ambiguous words.

Types of Semantic Features

Categorial Features

Pertains to the inherent characteristics of a word

Transitive Features

Imposes semantic conditions on arguments in a sentence

Application to Verbs

Verbs of Emotional State

Have specific transitive features

Verbs of Motion

Require certain categorial features in their arguments