Результаты исследований: Публикации в книгах, отчётах, сборниках, трудах конференций › статья в сборнике материалов конференции › научная › Рецензирование
Contextual Predictability of Texts for Texts Processing and Understanding. / Krutchenko, Olga; Pronoza, Ekaterina; Yagunova, Elena; Timokhov, Viktor; Ivanets, Alexander.
Mining Intelligence and Knowledge Exploration - 7th International Conference, MIKE 2019, Proceedings. ред. / P. B.R.; Veena Thenkanidiyoor; Rajendra Prasath; Odelu Vanga. Springer Nature, 2020. стр. 104-119 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Том 11987 LNAI).Результаты исследований: Публикации в книгах, отчётах, сборниках, трудах конференций › статья в сборнике материалов конференции › научная › Рецензирование
}
TY - GEN
T1 - Contextual Predictability of Texts for Texts Processing and Understanding
AU - Krutchenko, Olga
AU - Pronoza, Ekaterina
AU - Yagunova, Elena
AU - Timokhov, Viktor
AU - Ivanets, Alexander
N1 - Funding Information: The authors acknowledge the RSF for the research grant 18-18-00114. Publisher Copyright: © 2020, Springer Nature Switzerland AG. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2020
Y1 - 2020
N2 - This paper is the first part of contextual predictability model investigation for Russian, it is focused on linguistic and psychology interpretation of models, features, metrics and sets of features. The aim of this paper is to identify the dependence of the implementation of contextual predictability procedures on the genre characteristics of the text (or text collection): scientific vs. fictional. We construct a model predicting text elements and designate its features for texts of different genres and domains. We analyze various methods for studying contextual predictability, carry out a computational experiment against scientific and fictional texts, and verify its results by the experiment with informants (cloze-tests) and word embeddings (word2vec CBOW model). As a result, text processing model is built. It is evaluated based on the selected contextual predictability features and experiments with informants.
AB - This paper is the first part of contextual predictability model investigation for Russian, it is focused on linguistic and psychology interpretation of models, features, metrics and sets of features. The aim of this paper is to identify the dependence of the implementation of contextual predictability procedures on the genre characteristics of the text (or text collection): scientific vs. fictional. We construct a model predicting text elements and designate its features for texts of different genres and domains. We analyze various methods for studying contextual predictability, carry out a computational experiment against scientific and fictional texts, and verify its results by the experiment with informants (cloze-tests) and word embeddings (word2vec CBOW model). As a result, text processing model is built. It is evaluated based on the selected contextual predictability features and experiments with informants.
KW - Cloze test
KW - Conditional probability
KW - Contextual predictability
KW - Dice
KW - Fiction texts
KW - Informational entropy
KW - Language model
KW - Scientific corpora
KW - Surprisal
UR - http://www.scopus.com/inward/record.url?scp=85098289150&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-66187-8_11
DO - 10.1007/978-3-030-66187-8_11
M3 - Conference contribution
AN - SCOPUS:85098289150
SN - 9783030661861
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 104
EP - 119
BT - Mining Intelligence and Knowledge Exploration - 7th International Conference, MIKE 2019, Proceedings
A2 - B.R., P.
A2 - Thenkanidiyoor, Veena
A2 - Prasath, Rajendra
A2 - Vanga, Odelu
PB - Springer Nature
T2 - 7th International Conference on Mining Intelligence and Knowledge Exploration, MIKE 2019
Y2 - 19 December 2019 through 22 December 2019
ER -
ID: 73342010