Authorship Link Retrieval Between Documents

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

In this paper we propose a method for automatic author clustering called Document Authoring Link Retriever, DALIR. Documents are represented using Doc2Vec, experimenting with several parameters; afterwards, vectors are clustered (or linked together) using K-means and Hierarchical Agglomerative Clustering. We experimented with different vector representation sizes, different fixed number of clusters, and clustering methods. We evaluated our method on the author clustering task of PAN @ CLEF 2017. We used the BCubed F-score evaluation scheme of this task, being able to overcome some of the reported results from the first places of this challenge, although our method requires to manually establish a number of clusters a priori.

Idioma originalInglés
Título de la publicación alojadaAdvances in Computational Intelligence - 19th Mexican International Conference on Artificial Intelligence, MICAI 2020, Proceedings
EditoresLourdes Martínez-Villaseñor, Hiram Ponce, Oscar Herrera-Alcántara, Félix A. Castro-Espinoza
EditorialSpringer Science and Business Media Deutschland GmbH
Páginas297-305
Número de páginas9
ISBN (versión impresa)9783030608866
DOI
EstadoPublicada - 2020
Evento19th Mexican International Conference on Artificial Intelligence, MICAI 2020 - Mexico City, México
Duración: 12 oct. 202017 oct. 2020

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen12469 LNAI
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia19th Mexican International Conference on Artificial Intelligence, MICAI 2020
País/TerritorioMéxico
CiudadMexico City
Período12/10/2017/10/20

Huella

Profundice en los temas de investigación de 'Authorship Link Retrieval Between Documents'. En conjunto forman una huella única.

Citar esto