Aplikasi Penilaian Atribut Kualitas Perangkat Lunak Berdasarkan Sentimen Pada User Review Menggunakan Stochastic Gradient Descent Classifier

Nawangsih, Theresia (2024) Aplikasi Penilaian Atribut Kualitas Perangkat Lunak Berdasarkan Sentimen Pada User Review Menggunakan Stochastic Gradient Descent Classifier. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025201144-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (10MB) | Request a copy

Abstract

Semakin berkembangnya zaman salah satu kunci berkembangnya teknologi dalam kehidupan sehari-hari adalah perangkat lunak. Kualitas perangkat lunak yang baik dapat menjadikan tolak ukur keberhasilan suatu perangkat lunak tersebut dan meningkatkan kualitas perangkat lunak itu sendiri. Penilaian kualitas perangkat lunak dapat dinilai di antaranya dengan standar ISO/IEC 25010.
Salah satu tantangan dalam penilaian kualitas perangkat lunak adalah menentukan bobot relatif dari setiap atribut kualitas, seperti Functional Suitability dan Usability. Penelitian sebelumnya menggunakan metode Analytical Hierarchy Process (AHP) salah satu kelemahannya bergantung pada evaluasi subyektif dari para pakar . Tugas Akhir ini memberikan solusi dengan mengimplementasikan pembelajaran mesin algoritma model Stochastic Gradient Descent Classifier (SGDC) dan Natural Language Processing (NLP). Data dari ulasan pengguna digunakan untuk analisis sentimen. Hasil penelitian ini menunjukkan bahwa perhitungan menggunakan algoritma Stochastic Gradient Descent Classifier (SGDC) mendapatkan hasil akurasi 0.90, dan aplikasi ini dapat menampilkan penilaian serta rekomendasi perbaikan untuk meningkatkan kualitas perangkat lunak.
Uji fungsionalitas telah dilakukan dan hasilnya sesuai dengan yang diharapkan. Dengan penelitian ini, diharapkan penilaian kualitas perangkat lunak menjadi lebih komprehensif dan objektif, memberikan kontribusi signifikan terhadap peningkatan kualitas perangkat lunak di masa depan.

=====================================================================================================================================
As times progress, one of the keys to technological advancement in daily life is software. Good software quality can serve as a benchmark for its success and improve its own quality. Software quality assessment can be evaluated, among others, by the ISO/IEC 25010 standards.
One of the challenges in assessing software quality is determining the relative weight of each quality attribute, such as Functional Suitability and Usability. Previous research using the Analytical Hierarchy Process (AHP) method had the drawback of relying on the subjective evaluation of experts. This thesis provides a solution by implementing machine learning algorithms, specifically Stochastic Gradient Descent Classifier (SGDC) , and Natural Language Processing (NLP). User review data is used for sentiment analysis. The results of this study show that calculations using the Stochastic Gradient Descent Classifier (SGDC) algorithm achieved an accuracy of 0.90, and the application can provide assessments and recommendations for improving software quality.
functionality testing has been conducted, and the results meet expectations. This research aims to make software quality assessment more comprehensive and objective, significantly contributing to the improvement of software quality in the future.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Functional Suitability, ISO/IEC 25010, kualitas perangkat lunak, Natural Language Processing (NLP), pembelajaran mesin, penilaian, Stochastic Gradient Descent Classifier, Usability, machine learning, software quality assessment
Subjects:	T Technology > T Technology (General)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Theresia Nawangsih
Date Deposited:	31 Jul 2024 13:18
Last Modified:	31 Jul 2024 13:18
URI:	http://repository.its.ac.id/id/eprint/110589

Actions (login required)

View Item