Improving Thai Academic Web Page Classification Using Inverse Class Frequency and Web Link Information

Lertnattee, Verayuth; Theeramunkong, Thanaruk

doi:10.1109/waina.2008.120

Cited by 6 publications

(4 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because the two factors in tf.idf model are both estimated on the document level, but the tf factor of tf.icf is estimated on the document level and the icf factor is estimated on the category level. Our tf.icf is different from previous tf.icf methods [25][26][27][28][29][30][31], which have the same name, but have different meanings. For example, Reed [25] use the abbreviation ICF (Inverse Corpus Frequency) in dealing with stream data, and ICF in Ref [25] is inverse document frequency of the whole corpus.…”

Section: Two Novel Term Weighting Schemes Based On Icfmentioning

confidence: 91%

“…Pei et al [16] proposed an improved tf.idf method, which combined tf.idf and information gain. Many variations of ICF have been used in document classification [26,27,30,31].…”

Section: Related Workmentioning

confidence: 99%

“…For example, Reed [25] use the abbreviation ICF (Inverse Corpus Frequency) in dealing with stream data, and ICF in Ref [25] is inverse document frequency of the whole corpus. Lertnatteed [26,27] also proposed TFICF, whose TF factor is estimated on category level. Kimura et al [30] proposed TF.ICF to weight terms in cross-language retrieval system, and the TF factor [30] is defined as the ratio between term frequency and the number of terms in category c j .…”

Section: Two Novel Term Weighting Schemes Based On Icfmentioning

confidence: 99%

See 2 more Smart Citations

On entropy-based term weighting schemes for text categorization

et al. 2021

View full text Add to dashboard Cite

Term weighting schemes often dominate the performance of many classifiers, such as kNN, centroid-based classifier and SVMs. The widely used term weighting scheme in text categorization, i.e., tf.idf, is originated from information retrieval (IR) field. The intuition behind idf for text categorization seems less reasonable than IR. In this paper, we introduce inverse category frequency (icf) into term weighting scheme and propose two novel approaches, i.e., tf.icf and icf-based supervised term weighting schemes. The tf.icf adopts icf to substitute idf factor and favors terms occurring in fewer categories, rather than fewer documents. And the icf-based approach combines icf and relevance frequency (rf) to weight terms in a supervised way. Our cross-classifier and cross-corpus experiments have shown that our proposed approaches are superior or comparable to six supervised term weighting schemes and three traditional schemes in terms of macro-F1 and micro-F1.

show abstract

Section: Two Novel Term Weighting Schemes Based On Icfmentioning

confidence: 91%

“…Pei et al [16] proposed an improved tf.idf method, which combined tf.idf and information gain. Many variations of ICF have been used in document classification [26,27,30,31].…”

Section: Related Workmentioning

confidence: 99%

Section: Two Novel Term Weighting Schemes Based On Icfmentioning

confidence: 99%

See 1 more Smart Citation

On entropy-based term weighting schemes for text categorization

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The results suggested that the bigram model of TFIDF with term distributions was a good model. [44] improved Thai-language academic web page classification by using inverse class frequency and web link information. They suggest that inverse class frequency should be used instead of inverse document frequency for centroid based text categorization.…”

Section: Literature Reviewmentioning

confidence: 99%

Modeling A Generic Web Classification System Using Design Patterns

Sukakanya¹,

Porkaew²

2011

JCP

View full text Add to dashboard Cite

In order to save time in extracting specific information from high volume of data in web documents, this paper proposes an architectural model of generic web document classification system using design patterns for classifying web documents. This work implements two classification techniques for classifying Thai web documents, namely centroid classification and neural network classification, based on the proposed model and compares their classification effectiveness empirically. The training data sets in this experiment consist of 500 web documents of the following five categories (100 documents for each category): mobile phone sales, book sales, travel sales, education information and company profile. Another two hundred and fifty web documents were then used to test the two classifiers. The experiment results showed that the centroid classifier outperforms the neural network classifier both in term of efficiency and effectiveness

show abstract

Automatic Term Extraction for Sentiment Classification of Dynamically Updated Text Collections into Three Classes

Rubtsova

2014

Knowledge Engineering and the Semantic Web

View full text Add to dashboard Cite

Improving Thai Academic Web Page Classification Using Inverse Class Frequency and Web Link Information

Cited by 6 publications

References 11 publications

On entropy-based term weighting schemes for text categorization

On entropy-based term weighting schemes for text categorization

Modeling A Generic Web Classification System Using Design Patterns

Automatic Term Extraction for Sentiment Classification of Dynamically Updated Text Collections into Three Classes

Contact Info

Product

Resources

About