Title :
Measuring the Performance of Ontological Based Information Retrieval from a Social Media
Author :
Sediyono, Eko ; Suhartono ; Nivak, Christian
Author_Institution :
Satyawacana Christian Univ., Salatiga, Indonesia
Abstract :
Users of social media website now can search, share, or just browse for information with ease. In the field of travel and tourism, it can easily find and share information about travel activities, travel destination, and travel accommodation due to plentiful source of information, but there is also a drawback related to the quality of the information. Traditional search engine that uses keyword to search for information about the meaning of a word or a sentence often produce biased and irrelevant search results. Ontology has been developed to overcome this problem. In this paper an ontology framework is presented, which was specifically developed for travel and tourism field using data set from twitter. A crawler application also presented to get the data needed for analysis. Finally a performance measurement is done by using Average Precision and Mean Average Precision. A search result is calculated using the ontology data and crawler data. The goal of this research is to develop ontology for travel & tourism domain and to analyze the quality of the search result on all environments. The result of this research is ontology for travel & tourism domain from twitter with MAP of 91.5%.
Keywords :
data analysis; information retrieval; ontologies (artificial intelligence); search engines; social networking (online); travel industry; average precision; crawler application; data set; mean average precision; ontological based information retrieval; performance measurement; search engine; social media website; travel accommodation; travel activities; travel and tourism field; travel destination; twitter; Crawlers; Indexing; Ontologies; Resource description framework; Semantics; XML; Indexing; Ontology; Searching; Semantic; Social Media;
Conference_Titel :
Modelling Symposium (EMS), 2014 European
Print_ISBN :
978-1-4799-7411-5
DOI :
10.1109/EMS.2014.15