EVALUATION OF THE EFFICIENCY OF THE USE OF SEMANTIC SEARCH ENGINES IN THE NEWSPAPER DISCOURSE (USING THE EXAMPLE OF THE SEMANTIC SEARCH SYSTEM DUCKDUCKGO)

Abstract

The article describes the study of the effectiveness of the use of semantic search systems in the newspaper discourse. The research was conducted on the example of the semantic search engine DuckDuckGo. The attention is paid to the notion of semantic search system, discourse, newspaper discourse and criteria for searching information on the Internet. The principles of extraction of linguistically marked results of semantic search engine DuckDuckGo are investigated. The evaluation of the received data was carried out in accordance with the received information to the criteria for the search of information, among which the relevance, pertinence, accuracy and completeness are distinguished. To carry out an analysis of the content of contemporary newspaper discourse, 5 actual directions were selected for the experiment. Three variations were selected in each of the directions, and thus, 15 different requests were examined. They are conventionally divided into single tokens, two-component tokens and multi-component tokens. For the effectiveness and objectivity of the study, the first 10 search results were analyzed for each query in the semantic search engine DuckDuckGo. The article describes in more detail the methodology of conducting the research, namely the formulation of the criteria for the search and evaluation of the conformity of information. The results of issuing of semantic search engine DuckDuckGo for each search query are evaluated. For ease of analysis of the obtained results, the criteria of relevance, pertinence and completeness were presented in percentage equivalents. Each search criterion for different query types is analyzed. The importance of the criterion of pertinency in the newspaper discourse is established. It is stated that with an increase in the length of the search query it is significantly increased. It was found that for the research of newspaper discourse the best option is to formulate a multicomponent query in the natural language. The effectiveness of using semantic search systems in the newspaper discourse is evaluated. The effectiveness of using the DuckDuckGo semantic search engine in the newspaper discourse is proved.