pdf

Fast text tendency classification method

  • 2013-09-20
  • 254.3KB
  • Points it Requires : 2

A fast text tendency classification method is proposed, which adopts the category space model to describe the tendency of words to categories, and realizes classification based on the statistical characteristics of words; in view of the complexity of tendency classification, a new secondary feature extraction method is proposed on the basis of comprehensive consideration of three statistical characteristics: word frequency, word text frequency, and word distribution: the first feature extraction adopts a combined feature extraction method to remove low-frequency words and noise words evenly distributed in each category; the second feature extraction removes words with unclear category tendency. Experiments show that this classification method not only has high classification performance, but also runs fast, and has certain practical value in information retrieval, information filtering, content security management, etc. Keyword category weight; Category space model; Text tendency classification; Secondary feature extraction

unfold

You Might Like

Uploader
justyouandmehr
 

Recommended ContentMore

Popular Components

Just Take a LookMore

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
×