Abstract :
The number of internet users in Indonesia in the year 2017 is 132.7 million. That number rises
about 51% since the year 2016. In addition as much as 24.74 million internet users in Indonesia
do online shopping and spending money of US $5.6 billion (about Rp74,6 trillion) to shop at a
variety of e-commerce. In the year 2015 as much as 12.2% Indonesia Society purchase
smartphones via e-commerce. Based on the results of the questionnaire, a total of 93.55% stated
requires a summary of reviews of comments. The methods used in this research is the hierarchical
aspect sentiment model (HASM). The purpose of doing research with the use of this method is to
apply the method in the process of summarizing the comments so it can be known the value of
sentiment on the respective specifications. The method is applied to do a klasfikasi customer
comments based on positive, negative, and neutral a branded smartphone to a specification. There
are nine smartphone brands used on this research with each of its specifications. The
specifications used are memory, network, camera, simcard, and connectivity. The results of this
research in the form of classification of customer comments a certain brand of smartphone with
products that have a positive, negative, or neutral to a specification. The use of this method has an
average rating of precission or appropriateness of the application of the method to the study of
61%, recall or the success rate of the application of the method of 54%, and the level of accuracy
of the method of application of 96.05%.
Keywords : Accurate, E-commnerce, Hierarchical Aspect Sentiment Model, Precission, Recall,
Smartphone