اكتشاف الموضوعات المهيمنة باستخدام تقنية التنقيب في النصوص في تغريدات الناخبين و المرشحين خلال انتخابات مجلس الأمة الكويتي نموذجا

Abstract EN

This paper presents an applied study in which the text mining technique has been used to mine the texts of tweets that contained words related to the elections of the Kuwait national assembly (parliament) in order to discover the dominant topics (topic extraction) that the voters and candidates talked about during the period of the national assembly elections in the state of Kuwait in the year 2020.

furthermore, using text mining, the relationships that link certain words to certain topics have been discovered, in addition to discovering the important keywords within the context of the topics of the tweets.

the tweets were collected from the official twitter archive (twitter full archive), where I collected the tweets that contained keywords related to the national assembly elections during the mentioned period.

after that, I used SAS text miner system to mine the texts of the tweets and discover the topics they contain by using the latent semantic analysis (LSA) technique.

it was found that the general sentiment of voters and candidates towards government measures and projects is not generally positive, but rather tends to criticize the government’s shortcomings in aspects related to fighting corruption, developing education, achieving economic reform and development.

there is also a negative sentiment towards the lack of sufficient awareness among some voters when choosing candidates in a way that guarantees voting for the eligible candidates to reach the national assembly, in addition to the negative sentiment towards the phenomenon of buying the votes of some voters by some candidates.