Sense Inventory Alignment Using Lexical Substitutions and Crowdsourcing / Ustalov Dmitry,Igushkin Sergey // . - 2016. - V. , l. . - P. 56-61.

ISSN/EISSN:

нет данных / нет данных

Type:

Proceedings Paper

Abstract:

Sense inventory induction is a topical problem of deriving a set of synsets representing concepts using various automatic or human-assisted methods. There might be, and actually are, mistakes in such synsets. In this paper, we are focused on the problem of eliminating potentially duplicate synsets having exactly two words in common as the broader intersection is known to be successfully addressed by heuristics. We exploit the phenomena of lexical substitutions and microtask-based crowdsourcing for aligning the synsets to the individual word senses. We also present an open source mobile application implementing our approach. Our experiments on the Russian language show that the approach scales well and dramatically reduces the number of duplicate synsets in the inventory.

Author keywords:

нет данных

DOI:

нет данных

Web of Science ID:

ISI:000386684500009

Соавторы в МНС:

—

Другие поля

Поле	Значение
Book-Group-Author	IEEE
Booktitle	PROCEEDINGS OF THE INTERNATIONAL FRUCT CONFERENCE ON INTELLIGENCE, SOCIAL MEDIA AND WEB (ISMW FRUCT 2016)
Note	International FRUCT Conference on Intelligence, Social Media and Web ISMW-FRUCT, St Petersburg, RUSSIA, AUG 28-SEP 04, 2016
Organization	St Petersburg State Univ; ITMO Univ; RUSSIN FDN BASIC RES; IEEE; IEEE COMMUN SOC; FRUCT Acad Ind; EMC2; ANTIPLAGIAT RES; EwerestMD LLC; FRUCT Oy; NLPub project; E-WeREST
Publisher	IEEE
Address	345 E 47TH ST, NEW YORK, NY 10017 USA
Language	English
ISBN	978-9-5268-3977-6
Research-Areas	Computer Science; Engineering
Web-of-Science-Categories	Computer Science, Theory \& Methods; Engineering, Electrical \& Electronic
Author-Email	dmitry.ustalov@urfu.ru hotkeytlt@gmail.com
ResearcherID-Numbers	Ustalov, Dmitry/P-6307-2014
ORCID-Numbers	Ustalov, Dmitry/0000-0002-9979-2188
Number-of-Cited-References	24
Usage-Count-Last-180-days	4
Usage-Count-Since-2013	4
Doc-Delivery-Number	BG1ED