Combining Syntactic and Semantic Information in Knowledge Extraction Pipeline
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3ANN7PJ9VY" target="_blank" >RIV/00216208:11320/26:NN7PJ9VY - isvavai.cz</a>
Result on the web
<a href="http://dx.doi.org/10.1007/978-3-031-85067-7_1" target="_blank" >http://dx.doi.org/10.1007/978-3-031-85067-7_1</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-031-85067-7_1" target="_blank" >10.1007/978-3-031-85067-7_1</a>
Alternative languages
Result language
angličtina
Original language name
Combining Syntactic and Semantic Information in Knowledge Extraction Pipeline
Original language description
This paper presents an experimental pipeline and an annotation scheme for enriching text with semantic information by leveraging implicit information extracted from Abstract Meaning Representation (AMR). AMR, as an unanchored semantic representation language, encodes a wide range of information that can be useful for downstream NLP tasks. This implicit knowledge encompasses frames, semantic roles, numeric value labels, concept grounding to Wikipedia entries, and fine-grained relationship types. The extracted information can improve the performance of existing logic-based end-to-end question-answering pipelines that rely on deterministic rule-driven semantic parsers for text-to-logic conversion based on Universal Dependencies (UD). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
—
Others
Publication year
2025
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Lect. Notes Networks Syst.
ISBN
978-3-031-85066-0
ISSN
—
e-ISSN
—
Number of pages
10
Pages from-to
3-12
Publisher name
Springer Science and Business Media Deutschland GmbH
Place of publication
—
Event location
Sfax
Event date
Jan 1, 2026
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—