An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification
Pengelasan teks automatik adalah penting kerana peningkatan bilangan dokumen digital dan oleh itu ia perlu diurus. Kaedah pemodelan statistik terkini tidak memberi maklumat berguna yang mencukupi tentang topik untuk setiap ciri dan kategori. Tambahan pula, penyarian sifat menggunakan frekuensi kata-...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2016
|
Subjects: | |
Online Access: | http://eprints.usm.my/31479/1/AMMAR_ISMAEL_KADHIM_24.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my-usm-ep.31479 |
---|---|
record_format |
uketd_dc |
spelling |
my-usm-ep.314792019-04-12T05:25:22Z An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification 2016-03 Kadhim, Ammar Ismael QA75.5-76.95 Electronic computers. Computer science Pengelasan teks automatik adalah penting kerana peningkatan bilangan dokumen digital dan oleh itu ia perlu diurus. Kaedah pemodelan statistik terkini tidak memberi maklumat berguna yang mencukupi tentang topik untuk setiap ciri dan kategori. Tambahan pula, penyarian sifat menggunakan frekuensi kata-frekuensi dokumen songsang (TF-IDF) tradisional menghasilkan pengenalan kategori yang terlalu banyak untuk sesuatu dokumen. Dalam usaha pengelasan pula, kaedah k-jiran terdekat (k-NN) sedia ada dengan jarak Euclid dan skor keserupaan kosinus menghasilkan julat varians yang besar dalam prestasinya. Untuk menangani isu ini, kajian ini mengelaskan topik untuk teks pendek dan panjang dengan menggunakan pendekatan baharu untuk tahap-tahap utama pengelasan teks (iaitu penyarian sifat dan pengelasan teks). Kajian ini juga memperkenalkan TD-IDF dengan logaritma dan k-NN dengan skor keserupaan kosinus yang baharu untuk penyarian sifat dan pengelasan masing-masing. Lagipun, faktor yang memberi kesan terhadap prestasi pembelajaran mesin berselia juga dikenalpasti. Automatic text classification is important because of the increased availability of digital documents and therefore the need to organize them. The current state-of-the-art statistical modeling approaches do not provide sufficient useful information on the topics for each feature and category. Furthermore, feature extraction using traditional term frequency-inverse document frequency (TF-IDF) results in the identification of too many categories for a particular document. In terms of classification, current k-NN approaches with Euclidean distance and cosine similarity score produce a wide range of variance in performance. To address these issues, this study classifies topics for short and long texts using a new method for the main stage (i.e., feature extraction and text classification). The study also introduces TF-IDF with logarithm and k-NN with a new cosine similarity score for feature extraction and classification, respectively. 2016-03 Thesis http://eprints.usm.my/31479/ http://eprints.usm.my/31479/1/AMMAR_ISMAEL_KADHIM_24.pdf application/pdf en public phd doctoral Universiti Sains Malaysia Pusat Pengajian Sains Komputer (School of Computer Sciences) |
institution |
Universiti Sains Malaysia |
collection |
USM Institutional Repository |
language |
English |
topic |
QA75.5-76.95 Electronic computers Computer science |
spellingShingle |
QA75.5-76.95 Electronic computers Computer science Kadhim, Ammar Ismael An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
description |
Pengelasan teks automatik adalah penting kerana peningkatan bilangan dokumen digital dan oleh itu ia perlu diurus. Kaedah pemodelan statistik terkini tidak memberi maklumat berguna yang mencukupi tentang topik untuk setiap ciri dan kategori. Tambahan pula, penyarian sifat menggunakan frekuensi kata-frekuensi dokumen songsang (TF-IDF) tradisional menghasilkan pengenalan kategori yang terlalu banyak untuk sesuatu dokumen. Dalam usaha pengelasan pula, kaedah k-jiran terdekat (k-NN) sedia ada dengan jarak Euclid dan skor keserupaan kosinus menghasilkan julat varians yang besar dalam prestasinya. Untuk menangani isu ini, kajian ini mengelaskan topik untuk teks pendek dan panjang dengan menggunakan pendekatan baharu untuk tahap-tahap utama pengelasan teks (iaitu penyarian sifat dan pengelasan teks). Kajian ini juga memperkenalkan TD-IDF dengan logaritma dan k-NN dengan skor keserupaan kosinus yang baharu untuk penyarian sifat dan pengelasan masing-masing. Lagipun, faktor yang memberi kesan terhadap prestasi pembelajaran mesin berselia juga dikenalpasti.
Automatic text classification is important because of the increased availability of digital documents and therefore the need to organize them. The current state-of-the-art statistical modeling approaches do not provide sufficient useful information on the topics for each feature and category. Furthermore, feature extraction using traditional term frequency-inverse document frequency (TF-IDF) results in the identification of too many categories for a particular document. In terms of classification, current k-NN approaches with Euclidean distance and cosine similarity score produce a wide range of variance in performance. To address these issues, this study classifies topics for short and long texts using a new method for the main stage (i.e., feature extraction and text classification). The study also introduces TF-IDF with logarithm and k-NN with a new cosine similarity score for feature extraction and classification, respectively. |
format |
Thesis |
qualification_name |
Doctor of Philosophy (PhD.) |
qualification_level |
Doctorate |
author |
Kadhim, Ammar Ismael |
author_facet |
Kadhim, Ammar Ismael |
author_sort |
Kadhim, Ammar Ismael |
title |
An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
title_short |
An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
title_full |
An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
title_fullStr |
An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
title_full_unstemmed |
An Improved K-Nearest Neighbors Approach Using Modified Term Weighting And Similarity Coefficient For Text Classification |
title_sort |
improved k-nearest neighbors approach using modified term weighting and similarity coefficient for text classification |
granting_institution |
Universiti Sains Malaysia |
granting_department |
Pusat Pengajian Sains Komputer (School of Computer Sciences) |
publishDate |
2016 |
url |
http://eprints.usm.my/31479/1/AMMAR_ISMAEL_KADHIM_24.pdf |
_version_ |
1747820433743282176 |