Search

Search Results

• Semantic role labeling in natural texts

  164-171

  Zoltán Subecz

  Views:

  84

  The event extraction and semantic role labeling are important areas in information extraction from natural language texts. Semantic role labeling, sometimes also called shallow semantic parsing, is a task in natural language processing consisting of the detection of the semantic arguments associated with the predicate or verb of a sentence and their classification into their specific roles. We created a program for this task in Java language, in which we applied data
  mining and artificial intelligence algorithms. We introduce in this article the principles and results of this application.

• Verbal Event Detection in Hungarian Texts

  101-106

  Zoltán Subecz

  Éva Nagyné Csák

  Views:

  93

  The task of event deteci on is to ideni fy meni ons of events in texts. For the purposes of this task, an event meni on is any expression denoi ng an event or state that can be assigned to a pari cular point, or interval, in i me. In texts most event meni ons correspond to verbs, and most verbs introduce events. However this is not always the case. Events can be introduced by noun phrases, and some verbs fail to introduce events. Events somei mes are expressed with muli word expressions. Muli word expressions (MWEs) are lexical units that consist of more than one orthographical word, i.e. a lexical unit that contains spaces. The task is the deteci on of events in texts. In this ari cle we paid at eni on with verbs and infi nii ves. We wrote Java program for this task. We introduce in this ari cle the principles and results of this applicai on.

• Examination of the Notes to the financial statement through the example of enterprises engaged in sports activities

  61-66

  Veronika Fenyves

  Dóra Kerezsi

  Éva Bácsné Bába

  Krisztina Dajnoki

  Zoltán Bács

  Tibor Tarnóczi

  Views:

  214

  In the globalized and accelerated world of our time, it is essential that enterprises have up-to-date information. Only a company that is in possession of the necessary information is able to meet the challenges of business life. One source of public information might be the annual statement prepared and published in accordance with the stipulations of Act C of 2000. The annual statement consists of three parts: the balance sheet, the profit and loss account and the notes to the financial statement. The text-based information of the notes to the financial statement are for the more accurate interpretation of numerical data. The notes to the financial statement of certain market participants (consumers, suppliers, competitors, etc.) might support efficient operation of companies, their proper decision-making and risk assessment. In present study, the notes to the financial statement of companies identifying sports activities (Hungarian NACE 931) as their primary activity were utilized. By means of text mining, it was examined what information these notes to the financial statement include from amongst the mandatory information required by the Accounting Act.