<![CDATA[Current transportation development system has giving easiness for society to moving from place to another places. One of quite new public transportation is Light Rail Transit (LRT). LRT or light railroad have an opportunity to held public trial access for free just by registering yourself in LRT Jakarta website. For improving and maximize LRT Jakarta services, they have social media account where people may give feedback and assessment. One of way that could be done is by sentiment analysis to find out whether the society likes the services provided by LRT Jakarta. This study is using the Improved KNN as a classification method to determine people sentiment coupled with Information Gain to select features used during the classification process. The process of sentiment analysis includes data collection, text preprocessing that produces clean data, then weighting the terms with tf idf followed by feature selection using Information Gain. The next step is classification with Improved KNN using the features of the previous selection. The data used are primary data sourced from three social media namely Youtube, Twitter and Facebook. The results of this study are the best f-measure obtained when k = 11 using a 100% threshold or the whole term used that is equal to 85.51% with an average computational time calculated from 5-fold of 0.4647 minutes.]]>
Copyrights © 2020
