Wardhani, Fradila Octa Kusuma (2016) Aplikasi Penyaringan Status Dan Komentar Negatif Di Facebook. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5109100148-Undergraduate_Thesis.pdf - Accepted Version Download (1MB) | Preview |
Abstract
Kemudahan akses internet menyebabkan Facebook bisa diakses
oleh hampir semua orang, bahkan anak-anak sekalipun. Sementara,
tidak semua pengguna Facebook menyebarkan konten yang positif di
dalam Facebook. Dengan semakin banyaknya konten negatif yang
berkeliaran di Facebook, akan sangat membahayakan apabila anakanak yang belum bisa membedakan mana yang baik dan buruk bertemu
dengan konten negatif tersebut. Sehingga perlu adanya aplikasi yang
mampu meminimalisir jumlah konten negatif yang mungkin dapat
mengenai anak-anak.
Di dalam Tugas Akhir Penyaringan Status dan Komentar Negatif
di Facebook ini, dibuatlah suatu aplikasi yang mampu menyaring dan
menyembunyikan status dan komentar yang dinilai negatif.
Pengklasifikasian negatif dan tidaknya suatu status dan komentar dalam
aplikasi ini menggunakan metode Naive Bayes. Metode Naive Bayes
akan memecah status dan komentar menjadi kata-kata dan kemudian
mencari probabilitas negatif dari status tersebut terhadap data latih
yang telah dibuat sebelumnya. Aplikasi ini dibangun pada situs web
tersendiri dan bukan di dalam Facebook.
Aplikasi yang telah dibuat kemudian diuji berdasarkan
fungsionalitas dan validitas. Pengujian fungsionalitas dilakukan melalui
skenario yang mencerminkan fitur aplikasi. Hasil pengujian
fungsionalitas menunjukkan aplikasi dapat berjalan dengan cukup baik.
Pada pengujian validitas, didapatkan precision sistem adalah 46,67%
dan recall sistem adalah 93,33%.
==================================================================================================================
Ease of access to the Internet made Facebook can be accessed
by almost everyone, even for the kids. Meanwhile, not all Facebook
users post positive contents in Facebook. With the increasing number of
negative contents spreading on Facebook, will be very dangerous if
children who can not distinguish between what is good and bad met with
negative contents. Thus there is a need for an application that is able to
minimize the amount of negative contents that might be on children.
In this Final Project about Facebook Negative Status and
Comments Filtering, an application was made which is able to filter and
hide the status and comments that considered negative. Classification in
this application which conclude whether a status and comment is
negatives is using Naive Bayes method. Naive Bayes methods will break
status and comments into words and then look for a negative probability
of the status based on the training data that was created earlier. This
application is built on its own web site and not in Facebook.
Applications that have been created are then tested by the
functionality and validity. Functionality testing was done based on
testing scenarios which reflect the application's features. The test results
showed that functionality of the application can run pretty well. In
testing the validity, system precision was 46.67% and system recall is
93.33%
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSIf 006.312 War a |
| Uncontrolled Keywords: | Facebook; kalimat negatif; komentar; naive bayes; status |
| Subjects: | Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) |
| Divisions: | Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | ansi aflacha |
| Date Deposited: | 07 Aug 2020 04:42 |
| Last Modified: | 07 Aug 2020 04:42 |
| URI: | http://repository.its.ac.id/id/eprint/77180 |
Actions (login required)
 |
View Item |