Are keywords really keywords?
Lengyelné dr. Molnár Tünde
<>
Eszterházy Károly Főiskola
The theory and practice of the extraction of the keywords are based on statistic and linguistic methods. The statistic methods are pointed to the structured representation of the document (1), the significant expressions are established by the frequency of the occurrence of the words and the assigned positive or negative weight on them (2). In my lecture I present statistical data analysing methods, which make possible to define the ‘average behaviour’ of the words, and grouping methods, that can be used well in the exploratory work, for words with similar traits.
In my previous researches I made several survey of how people see the main point of an article, and I asked the participants to explore the most relevant sentences in articles on various topics. I am working on the extension of the research at present, and beside the extraction of the articles I also collect data for the automation of defining keywords. In my lecture I review the result of 2014 survey in which the participants besides highlighting the main point of the article, they marked the keywords. To answer the question ‘Are keywords really keywords?’ the result of the survey will be compared with the keywords marked by the authors.
(1) PRIBE, Torsten – KOLTER, Jan – KISS, Christine: Semiautomatische Annotation von Textdokumenten mit semantischen Metadaten. In: Wirtschaftsinformatik 2005. Eeconomy, egovernment, esociety. Heidelberg, 2005., p. 1319.
(2) Wordsimth tools. Definition of Key-ness [elektronikus dokumentum]. [http://www.lexically.net/downloads/version4/html/keyness_definition.htm] [2014.01.22]