Rancang Bangun Perangkat Lunak Teenstagram untuk Mengelompokkan Topik Caption Akun Instagram dengan Permodelan Author Topic Models (Studi Kasus: Siswa SMA di Surabaya)

Insani, Faiz Nur Fitrah (2018) Rancang Bangun Perangkat Lunak Teenstagram untuk Mengelompokkan Topik Caption Akun Instagram dengan Permodelan Author Topic Models (Studi Kasus: Siswa SMA di Surabaya). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5214100131-Undergraduate_Theses.pdf]
Preview
Text
5214100131-Undergraduate_Theses.pdf - Accepted Version

Download (2MB) | Preview

Abstract

Menurut survei Asosiasi Penyelenggara Jasa Internet Indonesia (APJII) [2] bahwa penetrasi pengguna internet di Indonesia adalah pelajar sebanyak 69.8 persen dari 132,7 juta pengguna, dan 97,4 persen menggunakannya untuk media sosial. Konten dari media sosial yang sering dikunjungi salah satunya adalah Instagram yaitu 15 persen dengan data sebanyak 19.9 juta. Hal tersebut menunjukkan bahwa pelajar yang memakai media sosial tidak lagi hanya untuk mengikuti tren, melainkan menjadi sebuah kebutuhan untuk menunjukkan eksistensi dirinya kepada publik. Tidak hanya menunjukkan foto ataupun lokasi kegiatan para pelajar, mereka juga menunjukkan beberapa ungkapan emosi dan perasaan yang dituang dalam caption Instagram. Dari fenomena tersebut dibutuhkan sebuah platform yang mampu memberikan informasi visual terhadap aktivitas pelajar dalam hal berekspresi di sosial media Instagram dengan melakukan analisis topic modeling terhadap perilaku dan kebiasaan pelajar ketika mengunggah gambar beserta caption menggunakan metode Author-Topic Models atau ATM. Penelitian ini dikhususkan untuk menganalisa data caption akun Instagram siswa SMA di Surabaya, setelah data dididapatkan serta dianalisa menggunakan Author-Topic Models kemudian dilakukan visualisasi terhadap topik dan atau kategori siswa berdasarkan caption dari masing – masing sekolah. Melalui proses pembuatan model, telah didapatkan hasil terbaik berupa 6 topik. Adapun 6 topik tersebut dapat dikatakan baik karena memiliki nilai perplexity yang kecil setelah dilakukan percobaan 30 kali. 6 Topik yang terbentuk dianalisis dan diterjemahkan ke dalam label kategori, yaitu perasaan, fotografi, fotografi dan artis, event Surabaya, liburan, dan agama dan musik. Masing – masing topik memiliki hasil topik perasaan dibahas oleh 3 sekolah, topik event surabaya dibahas oleh 2 sekolah, topik fotografi dibahas oleh 6 sekolah, topik fotografi dan artis dibahas oleh 3 sekolah, topik liburan dibahas oleh 3 sekolah, dan topik agama dan musik dibahas oleh 1 sekolah dengan jumlah data caption 3346 dari 18 sekolah.
=============================================================================================
According to a survey conducted by the Association of
Indonesian Internet Service Providers (APJII) [2] that internet
user penetration in Indonesia is 69.8 percent of 132.7 million
are students
, and 97.4 percent use it for social media. Content
from social
media that often visited one of them is Instagram
that have
15 percent with data as much as 19.9 million. It shows
that students who use social media are no longer just to follow
trends, but become a necessity to show their existence to the
public.
They n
ot only show photos or locations of
their
activities,
but
also show
their
emotional expressions and
feelings
in Instagram captions. From the phenomenon
requires
a platform that is able to provide visual information on the
activities of students in terms of
expression in social media
Instagram by doing topic
modeling, analysis
of student
behavior and habits when uploading images and captions using
the method of Author
-
Topic Models or ATM. This study was
devoted to analyze data caption of Instagram account of
high
school students in Surabaya, after the data was obtained and
analyzed using Author
-
Topic Models then visualization of the
topic or categories of students based on the caption of each
school. Through the modeling process, the best results have
been ob
tained in the form of
6 topics
.
The 6 topics can be said
good because it has a small perplexity value after the
experiment was done 30 times. 6 The topics formed are
analyzed and translated into category labels, namely
feelings
,
photography
,
photography and
celebrity
,
Surabaya events
,
holidays
, and
religion and music
. Each topic has the topic of
feelings
discussed by
3 schools
, the topic of the
Surabaya
events
discussed by
2 schools
,
photography
topics discussed by
6 schools
,
photography topic
s and artists
discussed by
3
schools
,
holiday topics
discussed by
3 schools
, and
religious
and music
topics discussed by
1 school
with
3346
data caption
data from 18 schools.

Item Type: Thesis (Undergraduate)
Additional Information: RSSI 005.1 Ins r-1 3100018074275
Uncontrolled Keywords: Instagram, Caption, Topic modeling, High School, Author-Topic Models, ATM
Subjects: Q Science > QA Mathematics > QA76 Computer software
Q Science > QA Mathematics > QA76.758 Software engineering
T Technology > T Technology (General)
Divisions: Faculty of Information and Communication Technology > Information Systems > 57201-(S1) Undergraduate Thesis
Depositing User: Faiz Nur Fitrah Insani
Date Deposited: 05 Feb 2018 04:09
Last Modified: 14 May 2020 05:35
URI: http://repository.its.ac.id/id/eprint/49514

Actions (login required)

View Item View Item