Effect of term distributions on centroid-based text categorization

Lertnattee, Verayuth; Theeramunkong, Thanaruk

doi:10.1016/j.ins.2003.07.007

Cited by 66 publications

(58 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We conclude by highlighting another difference with respect to related papers by Joachims (1997), Han and Karypis (2000) and Lertnattee and Theeramunkong (2004), where all features are used in their experiments. 8 In this work, features are preliminarily filtered and only those deemed most discriminant actually contribute to the classification.…”

mentioning

confidence: 78%

“…The control parameters β and γ define the relative impact of positive and negative examples in the definition of the class prototype. Dumais, Platt, Heckerman andSahami (1998), Joachims (1997), Han and Karypis (2000), Lertnattee and Theeramunkong (2004) set β to 1 and γ to 0, so that the prototype of a class coincides with the centroid of its positive training examples. In this work we follow this mainstream and compute the classification score as the cosine similarity between the document vector and the centroid of a class.…”

Section: Centroid-based Classifiermentioning

confidence: 99%

See 1 more Smart Citation

Classifying web documents in a hierarchy of categories: a comprehensive study

Ceci

Malerba

2007

J Intell Inf Syst

View full text Add to dashboard Cite

mentioning

confidence: 78%

Section: Centroid-based Classifiermentioning

confidence: 99%

Classifying web documents in a hierarchy of categories: a comprehensive study

Ceci

Malerba

2007

J Intell Inf Syst

View full text Add to dashboard Cite

“…The part used in the test determines whether the classifier is assigning the document to the correct classes according to their characteristics (LERTNATTEE;THEERAMUNKONG, 2004;SHANKAR;KARYPIS, 2000). Thus, the document class is given by:…”

Section: Model Configurationsmentioning

confidence: 99%

Automatic classification of journalistic documents on the Internet1

Oliveira

Branquinho

2017

Transinformação

View full text Add to dashboard Cite

Online journalism is increasing every day. There are many news agencies, newspapers, and magazines using digital publication in the global network. Documents published online are available to users, who use search engines to find them. In order to deliver documents that are relevant to the search, they must be indexed and classified. Due to the vast number of documents published online every day, a lot of research has been carried out to find ways to facilitate automatic document classification. The objective of the present study is to describe an experimental approach for the automatic classification of journalistic documents published on the Internet using the Vector Space Model for document representation. The model was tested based on a real journalism database, using algorithms that have been widely reported in the literature. This article also describes the metrics used to assess the performance of these algorithms and their required configurations. The results obtained show the efficiency of the method used and justify further research to find ways to facilitate the automatic classification of documents.Keywords: Automatic classification. Internet. Vector model. Resumo As publicações de textos jornalísticos na Internet aumentam a cada dia. São muitas as agências de notícias, jornais e revistas com publicações digitais na grande rede. Os documentos publicados ficam disponíveis aos usuários, que, por sua vez, usam buscadores para encontrá-los. Para se encontrarem os documentos mais próximos da busca realizada, estes devem estar previamente indexados e classificados. Com o volume gigantesco de materiais publicados a cada dia, muitas pesquisas estão sendo realizadas para encontrar facilitadores para a classificação automática de documentos. Este artigo objetiva demonstrar uma experiência de classificação automática de documentos jornalísticos publicados na

show abstract

“…Dari aspek teknis, terdapat kemungkinan jika pengambilan cluster lebih dari satu, dengan mempertimbangkan adanya kemunculan sub topic lain dari cluster hasil pengelompokan tweets yang dapat dijadikan nilai informatif tambahan dalam penentuan pembobotan kalimat penting penyusunan ringkasan. Penggunaan lebih dari satu cluster inilah dilakukan sebuah teknik pembobotan term distribution on centroid based [6]. Metode ini menerapkan konsep pembobotan intra-cluster, inter-cluster dan keseluruhan dokumen sehingga dapat meningkatkan bobot term yang diskriminatif, dengan kata lain metode ini mampu membentuk representasi bobot menjadi lebih baik atau memiliki sense terhadap cluster trending issue yang terbentuk.…”

Section: Pendahuluanunclassified

“…Metode TDCB menggunakan konsep distribusi term berdasarkan intra-class, inter-class dan keseluruhan koleksi dokumen untuk meningkatkan bobot term yang diskriminatif [6]. Setiap term mempunyai bobot sesuai dengan frekuensi dokumennya (informasi intraclass) dan faktor diskriminatif yang berbanding terbalik dengan jumlah kelas atau cluster yang berisi term tersebut (informasi inter-class).…”

Section: A Term Distribution On Centroid Basedunclassified

Ekstraksi Trending Issue Dengan Pendekatan Distribusi Kata Pada Pembobotan Term Untuk Peringkasan Multi-Dokumen Berita

Aditya¹,

Fatichah²,

Purwitasari³

2016

JUTI

View full text Add to dashboard Cite

ABSTRAK I. PENDAHULUANUTOMATIC Text Summarization adalah proses peringkasan dokumen teks yang dilakukan secara otomatis melalui mesin komputer. Otomatisasi ringkasan dapat dikenakan terhadap satu dokumen (single document summarization) atau beberapa dokumen (multi-document summarization). Ringkasan yang baik merupakan ringkasan dengan cakupan pembahasan (coverage) yang luas pada dokumen sumber serta keterhubungan antarkalimat (coherence) yang tinggi [1]. Terdapat dua jenis ringkasan berdasarkan dari cara pembuatannya. Jenis pertama adalah generic summary, dimana perwakilan dari teks asli yang mencoba untuk mempresentasikan semua fitur penting dari sebuah teks asal, pemakai menginginkan segala informasi yang penting, jenis ini mengikuti pendekatan bottom-up. Untuk jenis yang kedua adalah query-driven summary, peringkasan bersandar pada spesifikasi kebutuhan informasi pemakai, seperti ekstraksi kata kunci, mengikuti pendekatan top-down [2]. Ekstraksi kata kunci (keyword extraction) adalah sebuah cara untuk mencari perwakilan kata yang menggambarkan isi sebuah teks.A

show abstract

Effect of term distributions on centroid-based text categorization

Cited by 66 publications

References 29 publications

Classifying web documents in a hierarchy of categories: a comprehensive study

Classifying web documents in a hierarchy of categories: a comprehensive study

Automatic classification of journalistic documents on the Internet1

Ekstraksi Trending Issue Dengan Pendekatan Distribusi Kata Pada Pembobotan Term Untuk Peringkasan Multi-Dokumen Berita

Contact Info

Product

Resources

About