Difference between revisions of "Language-Agnostic Relation Extraction from Wikipedia Abstracts"
(+ Infobox work) |
(+ Embed) |
||
Line 10: | Line 10: | ||
== Overview == | == Overview == | ||
Large-scale knowledge graphs, such as [[DBpedia]], [[Wikidata]], or YAGO, can be enhanced by relation extraction from text, using the data in the knowledge graph as training data, i.e., using distant supervision. While most existing approaches use language-specific methods (usually for English), authors present a language-agnostic approach that exploits background knowledge from the graph instead of language-specific techniques and builds machine learning models only from language-independent [[features]]. Authors demonstrate the extraction of relations from [[Wikipedia]] abstracts, using the twelve largest language editions of Wikipedia. From those, authors can extract 1.6M new relations in DBpedia at a level of precision of 95%, using a RandomForest classifier trained only on language-independent features. Furthermore, authors show an exemplary geographical breakdown of the information extracted. | Large-scale knowledge graphs, such as [[DBpedia]], [[Wikidata]], or YAGO, can be enhanced by relation extraction from text, using the data in the knowledge graph as training data, i.e., using distant supervision. While most existing approaches use language-specific methods (usually for English), authors present a language-agnostic approach that exploits background knowledge from the graph instead of language-specific techniques and builds machine learning models only from language-independent [[features]]. Authors demonstrate the extraction of relations from [[Wikipedia]] abstracts, using the twelve largest language editions of Wikipedia. From those, authors can extract 1.6M new relations in DBpedia at a level of precision of 95%, using a RandomForest classifier trained only on language-independent features. Furthermore, authors show an exemplary geographical breakdown of the information extracted. | ||
+ | |||
+ | == Embed == | ||
+ | === Wikipedia Quality === | ||
+ | <code> | ||
+ | <nowiki> | ||
+ | Heist, Nicolas; Paulheim, Heiko. (2017). "[[Language-Agnostic Relation Extraction from Wikipedia Abstracts]]". Springer, Cham. DOI: 10.1007/978-3-319-68288-4_23. | ||
+ | </nowiki> | ||
+ | </code> | ||
+ | |||
+ | === English Wikipedia === | ||
+ | <code> | ||
+ | <nowiki> | ||
+ | {{cite journal |last1=Heist |first1=Nicolas |last2=Paulheim |first2=Heiko |title=Language-Agnostic Relation Extraction from Wikipedia Abstracts |date=2017 |doi=10.1007/978-3-319-68288-4_23 |url=https://wikipediaquality.com/wiki/Language-Agnostic_Relation_Extraction_from_Wikipedia_Abstracts |journal=Springer, Cham}} | ||
+ | </nowiki> | ||
+ | </code> | ||
+ | |||
+ | === HTML === | ||
+ | <code> | ||
+ | <nowiki> | ||
+ | Heist, Nicolas; Paulheim, Heiko. (2017). &quot;<a href="https://wikipediaquality.com/wiki/Language-Agnostic_Relation_Extraction_from_Wikipedia_Abstracts">Language-Agnostic Relation Extraction from Wikipedia Abstracts</a>&quot;. Springer, Cham. DOI: 10.1007/978-3-319-68288-4_23. | ||
+ | </nowiki> | ||
+ | </code> |
Revision as of 11:11, 27 March 2020
Authors | Nicolas Heist Heiko Paulheim |
---|---|
Publication date | 2017 |
DOI | 10.1007/978-3-319-68288-4_23 |
Links | Original |
Language-Agnostic Relation Extraction from Wikipedia Abstracts - scientific work related to Wikipedia quality published in 2017, written by Nicolas Heist and Heiko Paulheim.
Overview
Large-scale knowledge graphs, such as DBpedia, Wikidata, or YAGO, can be enhanced by relation extraction from text, using the data in the knowledge graph as training data, i.e., using distant supervision. While most existing approaches use language-specific methods (usually for English), authors present a language-agnostic approach that exploits background knowledge from the graph instead of language-specific techniques and builds machine learning models only from language-independent features. Authors demonstrate the extraction of relations from Wikipedia abstracts, using the twelve largest language editions of Wikipedia. From those, authors can extract 1.6M new relations in DBpedia at a level of precision of 95%, using a RandomForest classifier trained only on language-independent features. Furthermore, authors show an exemplary geographical breakdown of the information extracted.
Embed
Wikipedia Quality
Heist, Nicolas; Paulheim, Heiko. (2017). "[[Language-Agnostic Relation Extraction from Wikipedia Abstracts]]". Springer, Cham. DOI: 10.1007/978-3-319-68288-4_23.
English Wikipedia
{{cite journal |last1=Heist |first1=Nicolas |last2=Paulheim |first2=Heiko |title=Language-Agnostic Relation Extraction from Wikipedia Abstracts |date=2017 |doi=10.1007/978-3-319-68288-4_23 |url=https://wikipediaquality.com/wiki/Language-Agnostic_Relation_Extraction_from_Wikipedia_Abstracts |journal=Springer, Cham}}
HTML
Heist, Nicolas; Paulheim, Heiko. (2017). "<a href="https://wikipediaquality.com/wiki/Language-Agnostic_Relation_Extraction_from_Wikipedia_Abstracts">Language-Agnostic Relation Extraction from Wikipedia Abstracts</a>". Springer, Cham. DOI: 10.1007/978-3-319-68288-4_23.