Twitter Text Mining for Sentiment Analysis on People’s Feedback about Oman Tourism









Abstract

Sentiment analysis plays vital role in the internet era due to extensive range of business applications and social media. Inspiration behind sentiment analysis is that it provides people`s opinion about the product, which helps to improve the product quality. It also supports to take purchase/manufacturing decisions. In this paper we apply sentiment analysis to catch people feedback about Oman tourism using social media messages. For this we use tweeter data set to analyze tourist opinion about this country. In this paper, we recommend innovative sentiment analysis method based on common sense knowledge (Domain Specific Ontology). We created our own Oman tourism ontology based on ConceptNet. Entities are identified from the tweets using POS tagger and entities are compared with concepts in the domain specific ontology. Further the sentiment of the extracted entities are determined by the combined sentiment lexicon approach. Finally semantic orientations of domain specific features are combined with respect to the domain. We deliberate conceptual semantic as feature which can be combined with machine learning algorithm to enhance the performance of sentiment analysis of Oman tourism.


Modules


Algorithms

Text mining,Sentiment analysis ,Feature extraction , Machine learning


Software And Hardware

python,anaconda,jupyter,mysql,pandas