PERBANDINGAN METODE RANDOM FOREST DAN K-NEAREST NEIGHBORS PADA ANALISIS SENTIMEN PENGGUNA TWITTER MENGENAI PROMO GOJEK
Online transportation services are one of the hotly discussed topics. Many people rely on online traffic, or drivers and customers. Online transportation is a topic that is often discussed because public transportation needs in some areas are difficult to meet. Public responses to the services provi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Book |
| Published: |
2023-01-11.
|
| Subjects: | |
| Online Access: | Link Metadata |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Online transportation services are one of the hotly discussed topics. Many people rely on online traffic, or drivers and customers. Online transportation is a topic that is often discussed because public transportation needs in some areas are difficult to meet. Public responses to the services provided by online transportation service providers vary, some giving positive and negative results. Sentiment analysis is one way to understand the opinion of a person or group of people. Data in the form of tweets will be collected via Twitter as many as 453 tweets and divided into 80% training data and 20% test data. After that, the data will be preprocessed using case folding, data cleansing, normalization, stemming, stopword removal and tokenization. After the data is preprocessed, the data will be given term weighting using TF-IDF. After the data is weighted, the data will be processed using K-Nearest Neighbors and Random Forest. This research is expected to be able to obtain information on the sentiment of public opinion towards the Gojek promo and determine the performance of the K-Nearest Neighbors and Random Forest methods. In the first test, using the K-Nearest Neighbors method, the accuracy value is 75%, the precision value is 80%, the recall value is 64%, and the f-1 score is 71%. The second test is by applying the Random Forest method and obtained an accuracy value of 77%, a precision value of 75%, a recall value of 77%, and an f-1 score of 75%. Based on the evaluation results, the Random Forest method is better than the K-Nearest Neighbors method with an accuracy value of 77%. |
|---|---|
| Item Description: | http://repository.upnvj.ac.id/23265/1/ABSTRAK.pdf http://repository.upnvj.ac.id/23265/2/AWAL.pdf http://repository.upnvj.ac.id/23265/3/BAB%20I.pdf http://repository.upnvj.ac.id/23265/4/BAB%20II.pdf http://repository.upnvj.ac.id/23265/5/BAB%20III.pdf http://repository.upnvj.ac.id/23265/6/BAB%20IV.pdf http://repository.upnvj.ac.id/23265/7/BAB%20V.pdf http://repository.upnvj.ac.id/23265/8/DAFTAR%20PUSTAKA.pdf http://repository.upnvj.ac.id/23265/9/RIWAYAT%20HIDUP.pdf http://repository.upnvj.ac.id/23265/10/LAMPIRAN.pdf http://repository.upnvj.ac.id/23265/11/PLAGIARISME.pdf http://repository.upnvj.ac.id/23265/12/ARTIKEL%20KI.pdf |