Information-Theoretic method for classification of texts

B. Ya Ryabko; A. E. Gus’kov; I. V. Selivanova

doi:10.1134/S0032946017030115

Information-Theoretic method for classification of texts

B. Ya Ryabko, A. E. Gus’kov, I. V. Selivanova

Кафедра компьютерных систем ФИТ

Результат исследования: Научные публикации в периодических изданиях › статья

Аннотация

We consider a method for automatic (i.e., unmanned) text classification based on methods of universal source coding (or “data compression”). We show that under certain restrictions the proposed method is consistent, i.e., the classification error tends to zero with increasing text lengths. As an example of practical use of the method we consider the classification problem for scientific texts (research papers, books, etc.). The proposed method is experimentally shown to be highly efficient.

Язык оригинала	английский
Страницы (с-по)	294-304
Число страниц	11
Журнал	Problems of Information Transmission
Том	53
Номер выпуска	3
DOI	https://doi.org/10.1134/S0032946017030115
Состояние	Опубликовано - 1 июл 2017

Доступ к документу

10.1134/S0032946017030115

Link to publication in Scopus

Fingerprint Подробные сведения о темах исследования «Information-Theoretic method for classification of texts». Вместе они формируют уникальный семантический отпечаток (fingerprint).

Цитировать

Ryabko, B. Y., Gus’kov, A. E., & Selivanova, I. V. (2017). Information-Theoretic method for classification of texts. Problems of Information Transmission, 53(3), 294-304. https://doi.org/10.1134/S0032946017030115

@article{a7c0520104494552a51ee4acf6e06f80,

title = "Information-Theoretic method for classification of texts",

abstract = "We consider a method for automatic (i.e., unmanned) text classification based on methods of universal source coding (or “data compression”). We show that under certain restrictions the proposed method is consistent, i.e., the classification error tends to zero with increasing text lengths. As an example of practical use of the method we consider the classification problem for scientific texts (research papers, books, etc.). The proposed method is experimentally shown to be highly efficient.",

author = "Ryabko, {B. Ya} and Gus{\textquoteright}kov, {A. E.} and Selivanova, {I. V.}",

year = "2017",

month = jul,

day = "1",

doi = "10.1134/S0032946017030115",

language = "English",

volume = "53",

pages = "294--304",

journal = "Problems of Information Transmission",

issn = "0032-9460",

publisher = "Maik Nauka-Interperiodica Publishing",

number = "3",

}

TY - JOUR

T1 - Information-Theoretic method for classification of texts

AU - Ryabko, B. Ya

AU - Gus’kov, A. E.

AU - Selivanova, I. V.

PY - 2017/7/1

Y1 - 2017/7/1

N2 - We consider a method for automatic (i.e., unmanned) text classification based on methods of universal source coding (or “data compression”). We show that under certain restrictions the proposed method is consistent, i.e., the classification error tends to zero with increasing text lengths. As an example of practical use of the method we consider the classification problem for scientific texts (research papers, books, etc.). The proposed method is experimentally shown to be highly efficient.

AB - We consider a method for automatic (i.e., unmanned) text classification based on methods of universal source coding (or “data compression”). We show that under certain restrictions the proposed method is consistent, i.e., the classification error tends to zero with increasing text lengths. As an example of practical use of the method we consider the classification problem for scientific texts (research papers, books, etc.). The proposed method is experimentally shown to be highly efficient.

UR - http://www.scopus.com/inward/record.url?scp=85031754667&partnerID=8YFLogxK

U2 - 10.1134/S0032946017030115

DO - 10.1134/S0032946017030115

M3 - Article

AN - SCOPUS:85031754667

VL - 53

SP - 294

EP - 304

JO - Problems of Information Transmission

JF - Problems of Information Transmission

SN - 0032-9460

IS - 3

ER -