A parametric methodology for text classification

Abstract

Finding the correct category (class) a new unclassified document belongs to is an interesting and difficult problem, with a wide range of applications. Our methodology for narrative text classification is based on two techniques: we calculate the distance (similarity) between the new unclassified document and all the pre-classified documents of each class and also calculate the similarity of the new document to the ‘average class document’ of each class. In both cases we use key phrases (text phrases or key terms) as the distinctive features of our text classification methodology and eventually the proposed text classification method is based on the automatic extraction of an authority list of key phrases that is appropriate for discriminating between different classes. In this paper, we apply this methodology in handling Greek text and we present the key concepts, the algorithms, and some critical decisions. A number of parameters of the mining algorithm are also fine tuned. The actual text classification system, the adopted (embedded) ideas and the alternative values of parameters are evaluated using two training sets (test collections).

Conception And Development Of A System Used To Organize And Facilitate Access To Environmental Information

“…It is important to consider that the solution adopted was technically simple, but with a major impact on the management of documents. The BASGEO allows the decentralization and distribution of workflow functions (Pešović, Vidaković, Ivanović, Budimac, & Vidaković, 2011) despite the need for a better classification of documents (Rocha, et al, 2013;Yang, Lin, & Wei, 2010;Karanikolas & Skourlas, 2010).…”

Use of linguistic forms mining in the link analysis of legal documents

“…In addition, there are several studies exploring the frequency of occurrence of linguistic forms in documents, identification of key linguistic forms, identification of the "true quality" of linguistic forms and so forth. In their papers, Karanikolas & Skourlas [11,12] conduct research on automatic classification of documents. They focus on the issue of extracting key-phrases from a collection of texts in order to use them as attributes for text classification.…”

“…They look for sequences of words (key-phrases) that will be used as features for classification rules and not for extracting association rules. In their works Karanikolas & Skourlas [11,12] extracted key-phrases which are frequent within the documents of one or few classes but are not so frequent in the documents of the remaining classes of the training set. Furthermore, in [12] it is said that words that constitute key phrases must coexist in a specific window size.…”

A Methodology for Building Simple but Robust Stemmers without Language Knowledge: Stemmer Configuration

“…The Porter's stemmer (Porter, 1980) is more granular and it uses file levels. Except the abovepublished stemmers, there is another elder Greek stemmer (nnk's stemmer), incorporated to some commercial (Moumouris, 1995) and some research (Karanikolas, 2007;Karanikolas and Skourlas, 2010) applications. This stemmer uses a large number of suffixes and it elaborates the suffix removal (or replacement) in six iterations (levels).…”