Analysis of inverse class frequency in centroid-based text classification

Lertnattee, Verayuth; Theeramunkong, Thanaruk

doi:10.1109/iscit.2004.1413903

Cited by 21 publications

(24 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike the above STW schemes which take into account only the term's coarse distribution on the positive and negative classes, the ICF-based weighting schemes, e.g., TF-ICF and TF-IDF-ICF ( Lertnattee & Theeramunkong, 2004 ), take the term distribution across multiple classes into account. As we know, the inverse document frequency (IDF) of a term reflects only the term's specificity or uniqueness with respect to a single document.…”

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

“…ICF means that the fewer classes a term occurs in, the greater its specificity to a class or its class distinguishing power is. Indeed, TF-ICF performs better than TF-IDF in some text classification tasks ( Lertnattee & Theeramunkong, 2004;Wang & Zhang, 2013 ). However, ICF adopts only the term distribution at the coarse class level instead of the document level.…”

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

“…However, ICF adopts only the term distribution at the coarse class level instead of the document level. In order to bear a term's distinguishing power among documents, ICF and IDF are combined in the TF-IDF-ICF scheme ( How & Narayanan, 2004;Lertnattee & Theeramunkong, 2004;Ren & Sohrab, 2013 ) which has been proved effective in text classification. Furthermore, in consideration of the fact that a term representative of a class may occasionally appear in other classes, so a new factor called inverse class-space-density-frequency (ICSDF) as a variant of ICF has recently been introduced in the TF-IDF-ICSDF scheme ( Ren & Sohrab, 2013 ).…”

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

“…So the effectiveness of some previous STW schemes for multiclass text classification remains unknown. In addition, although some STW schemes are based on inverse class frequency (ICF) over multiple classes ( Lertnattee & Theeramunkong, 2004 ;Wang & Zhang, 2013 ), they also have some deficiencies, e.g., class frequency is too coarser than document frequency.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Turning from TF-IDF to TF-IGM for term weighting in text classification

Chen

Zhang

Long

et al. 2016

Expert Systems with Applications

235

144

View full text Add to dashboard Cite

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

Section: Icf-based Term Weighting Schemesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Turning from TF-IDF to TF-IGM for term weighting in text classification

Chen

Zhang

Long

et al. 2016

Expert Systems with Applications

235

144

View full text Add to dashboard Cite

“…Feature Weighting: By comparing the main existing term weighting methods such as TF-IDF [5], TF-ICF [34], [36] and TF-IDF-ICF [35], [36], TF-IDF-ICF is chosen as the term weighting method of this paper.…”

Section: Experimental Settingmentioning

confidence: 99%

A Framework of Centroid-Based Methods for Text Categorization

Wang

Chen

Wang

2014

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYText Categorization (TC) is a task of classifying a set of documents into one or more predefined categories. Centroid-based method, a very popular TC method, aims to make classifiers simple and efficient by constructing one prototype vector for each class. It classifies a document into the class that owns the prototype vector nearest to the document. Many studies have been done on constructing prototype vectors. However, the basic philosophies of these methods are quite different from each other. It makes the comparison and selection of centroid-based TC methods very difficult. It also makes the further development of centroid-based TC methods more challenging. In this paper, based on the observation of its general procedure, the centroid-based text classification is treated as a kind of ranking task, and a unified framework for centroid-based TC methods is proposed. The goal of this unified framework is to classify a text via ranking all possible classes by document-class similarities. Prototype vectors are constructed based on various loss functions for ranking classes. Under this framework, three popular centroid-based methods: Rocchio, Hypothesis Margin Centroid and DragPushing are unified and their details are discussed. A novel centroid-based TC method called SLRCM that uses a smoothing ranking loss function is further proposed. Experiments conducted on several standard databases show that the proposed SLRCM method outperforms the compared centroid-based methods and reaches the same performance as the state-of-the-art TC methods. key words: text categorization, centroid-based methods, smoothing listwise ranking centroid method, a unified framework

show abstract

Prediction of Depression Using Machine Learning and NLP Approach

Mali

Sedamkar

2022

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Analysis of inverse class frequency in centroid-based text classification

Cited by 21 publications

References 11 publications

Turning from TF-IDF to TF-IGM for term weighting in text classification

Turning from TF-IDF to TF-IGM for term weighting in text classification

A Framework of Centroid-Based Methods for Text Categorization

Prediction of Depression Using Machine Learning and NLP Approach

Contact Info

Product

Resources

About