MUHAMAD HANIF ZEIN, NIM. 17106050038 (2021) ANALISIS PERBANDINGAN METODE KLASIFIKASI TEMA PADA TEKS BERITA MENGGUNAKAN ALGORITMA KNEAREST NEIGHBOR DENGAN FUZZY K-NEAREST NEIGHBOR. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.
|
Text (ANALISIS PERBANDINGAN METODE KLASIFIKASI TEMA PADA TEKS BERITA MENGGUNAKAN ALGORITMA KNEAREST NEIGHBOR DENGAN FUZZY K-NEAREST NEIGHBOR)
17106050038_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview |
|
![]() |
Text (ANALISIS PERBANDINGAN METODE KLASIFIKASI TEMA PADA TEKS BERITA MENGGUNAKAN ALGORITMA KNEAREST NEIGHBOR DENGAN FUZZY K-NEAREST NEIGHBOR)
17106050038_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (4MB) |
Abstract
In general, news delivered in news portals consists of several categories such as political news, sports, economics, health, and others. However, in dividing the news into these categories, for now it is still done manually. Automation is needed to support this, in this case it is necessary to have an algorithm that is able to classify news themes automatically. In Data Mining, one of the two classification methods, namely KNN and its development, namely FKNN as an alternative, is expected to find the best solution. The classification is carried out with data from the web crawling of the kompas.com news site taken from 3 categories as a sample of 105 news items each. then do preprocessing text and weighting the TF IDF so that it is in the form of a term weigh data set. After that, the classification of the KNN and FKNN algorithms was carried out by searching for the value of k and comparing the best training and testing data. And the last test is with K-Fold Cross Validation to determine the results of the comparison of the performance of the two algorithms. Based on the research that has been done, it can be concluded that the highest accuracy performance value of the classification method using the KNearest Neighbor algorithm is the value of k = 4 with a comparison of training and testing data that is 9:1. The development of KNN with Fuzzy KNN did not provide a significant increase in accuracy because it could only provide 1% accuracy with a value of k and the highest comparison of training and testing data from KNN.
Item Type: | Thesis (Skripsi) |
---|---|
Additional Information: | Muhammad Didik Rohmad Wahyudi, S.T., MT |
Uncontrolled Keywords: | news teks, classification, KNN, Fuzzy KNN, TF IDF, K-Fold Cross Validation. |
Subjects: | Tehnik Informatika |
Divisions: | Fakultas Sains dan Teknologi > Teknik Informatika (S1) |
Depositing User: | Drs. Mochammad Tantowi, M.Si. |
Date Deposited: | 19 Nov 2021 13:33 |
Last Modified: | 19 Nov 2021 13:33 |
URI: | http://digilib.uin-suka.ac.id/id/eprint/47033 |
Share this knowledge with your friends :
Actions (login required)
![]() |
View Item |