Nonlinear transformation of term frequencies for term weighting in text categorization

Erenel, Zafer; Altınçay, Hakan

doi:10.1016/j.engappai.2012.06.013

Cited by 27 publications

(14 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…So reducing the effect of high TF may result in more reasonable term weighting. It has been found from some researches ( Erenel & Altinçay, 2012;Xuan & Quang, 2014 ) and our preliminary experiments that if the local TF factor is reduced properly, the accuracy of text classification is improved for some text corpora. The local TF factor is usually reduced by a logarithm operation, for example, replacing tf kd with log( tf kd + 1) ( Dumais, 1991 ).…”

Section: Term Weighting By Tf-igmmentioning

confidence: 95%

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

Chen

Zhang

Long

et al. 2016

Expert Systems with Applications

244

144

View full text Add to dashboard Cite

Section: Term Weighting By Tf-igmmentioning

confidence: 95%

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

Chen

Zhang

Long

et al. 2016

Expert Systems with Applications

244

144

View full text Add to dashboard Cite

“…The nttf is a nonlinear transformation of term frequencies given in Ref. [6]. Considering the case of two classes, we adopt…”

Section: Term Weighting Methodsmentioning

confidence: 99%

“…It helps to adjust for the fact that some words appear more frequently in general and is known as the global weight (or, the collection frequency factor). [4][5][6] However, tf-idf is not the best choice for a supervised learning task, in which the labels of the training documents contain important information. 4 Using these information, supervised term weighting methods can assign larger weights to the discriminative terms among different classes, and have gained increasing attention since the beginning of the new century.…”

Section: Introductionmentioning

confidence: 99%

“…13 After studying the nonlinear transformation of term frequency, Erenel and Altınçay suggested a supervised global weight factor which considers the term frequency and confirmed that the use of term frequency in the global weight factor is beneficial on tasks which does not involve highly repeated terms. 6 In addition, high frequency terms are expected to have larger weighs especially for the long documents. Transformation log 1 + tf and the direct use of tf have been studied in information retrieval and TC.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Term Frequency Based Weighting Scheme Using Naïve Bayes for Text Classification

Feng¹,

Wang²,

Sun³

et al. 2016

j comput theor nanosci

View full text Add to dashboard Cite

Term weighting is a strategy to assign weights to terms to improve the performance of many classifiers, such as kNN and SVM in text classification. Supervised term weighting methods have received increasing attention, in which information on the membership of training documents to classes is used. Most existing methods follow the local weight multiplies the global weight framework, but the contribution of term frequency for term weighting has not been fully investigated. In this paper, we propose a weighting scheme named term frequency-relevance term frequency based on a probabilistic model. After investigating two kinds of widely used naïve Bayes (NB) models, we employ the term event Multinomial NB model to capture the term frequency information. The matching score function based on the prediction probability ratio can then be factorized. Finally, we get the weight for each term by replacing the parameter by an estimator, term frequency is used in formulating not only the local weight factor but also the global weight factor. Numerical experiment results on two benchmark text datasets (Reuters-21578 and 20 Newsgroups) demonstrate that our proposed method outperforms the representative term weighting methods.

show abstract

“…Although the term frequency information within the documents is commonly employed in the local weighting factor, it rarely employed in the global weighting factor. Erenel and Altınçay confirmed that using term frequency in the global weight factor is beneficial for tasks which do not involve highly repeated terms [23]. …”

Section: Introductionmentioning

confidence: 99%

Relevance popularity: A term event model based feature selection scheme for text classification

Feng

Yang

et al. 2017

PLoS ONE

View full text Add to dashboard Cite

Feature selection is a practical approach for improving the performance of text classification methods by optimizing the feature subsets input to classifiers. In traditional feature selection methods such as information gain and chi-square, the number of documents that contain a particular term (i.e. the document frequency) is often used. However, the frequency of a given term appearing in each document has not been fully investigated, even though it is a promising feature to produce accurate classifications. In this paper, we propose a new feature selection scheme based on a term event Multinomial naive Bayes probabilistic model. According to the model assumptions, the matching score function, which is based on the prediction probability ratio, can be factorized. Finally, we derive a feature selection measurement for each term after replacing inner parameters by their estimators. On a benchmark English text datasets (20 Newsgroups) and a Chinese text dataset (MPH-20), our numerical experiment results obtained from using two widely used text classifiers (naive Bayes and support vector machine) demonstrate that our method outperformed the representative feature selection methods.

show abstract

Nonlinear transformation of term frequencies for term weighting in text categorization

Cited by 27 publications

References 26 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 Term Frequency Based Weighting Scheme Using Naïve Bayes for Text Classification

Relevance popularity: A term event model based feature selection scheme for text classification

Contact Info

Product

Resources

About