Aka, Arinda (2025) Rekomendasi Keputusan Bisnis Pengembangan Aplikasi Melalui Analisa Segmentasi Ulasan Pelanggan Menggunakan Large Language Model. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6032232130-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6032232130-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy |
Abstract
Ulasan pelanggan merupakan aset strategis penting bagi perusahaan untuk memahami kebutuhan pelanggan dan mengidentifikasi peluang peningkatan layanan. Namun, studi terkait banyak memberikan rekomendasi bisnis yang kurang baik diakibatkan dari ulasan pelanggan yang kurang atau bahkan tidak relevan terhadap produk atau layanan yang diberikan. Sedangkan, analisis ulasan pelanggan dengan tepat berperan penting dalam meningkatkan kepuasan, mengidentifikasi masalah, memperkuat loyalitas, dan mendorong inovasi. Penelitian Tesis ini bertujuan untuk memberikan rekomendasi keputusan bisnis pengembangan aplikasi sesuai dengan analisis segmentasi ulasan pelanggan. Penelitian Tesis ini men- gusulkan sebuah metode untuk rekomendasi keputusan bisnis terkait pengembangan aplikasi berdasarkan Large Language Model yang mampu menganalisis segmentasi ulasan pelanggan. Lebih spesifik, metode yang diusulkan dapat mengekstraksi konteks, topik, sentimen dan emosi secara bersamaan yang tidak dipertimbangkan pada studi – studi terkait. Metode yang diusulkan memiliki dua tahap utama: (i) multi-stage LLM prompting segmentasi ulasan pelanggan untuk memberikan pengetahuan terkait ulasan pelanggan yang informatif atau dibutuhkan, dan (ii) retrieval-augmented generation informasi bisnis terkait untuk memberikan rekomendasi keputusan bisnis pengembangan aplikasi. Penelitian Tesis ini divalidasi berdasarkan data crawling ulasan pelanggan pada sebuah aplikasi dari suatu perusahaan di Indonesia. Hasil penelitian menunjukkan efektivitas metode ini, khususnya dalam menghasilkan klasi- fikasi dan segmentasi ulasan yang akurat, serta rekomendasi bisnis yang strategis dan relevan. Secara keseluruhan, penelitian ini memberikan kontribusi penting dalam pengembangan kerangka kerja analisis ulasan pelanggan, serta membuka peluang pengembangan sistem pendukung keputusan bisnis berbasis LLM dan RAG yang lebih efektif di masa depan.
=====================================================================================================================================
Customer reviews are a vital strategic asset for companies to understand customer needs and iden- tify opportunities for service improvement. Most related studies offerred poor quality business recom- mendation as they ignored the unrelated customer reviews towards the available product or services. Analysis on customer reviews are crucial to increase the customer satisfactory levels, identify the draw- backs, strength the customer loyalty, and encourage the innovation. Therefore, this Thesis aims to provide a recommendation of business decision on apps development inline with customer review segmentation analysis. This Thesis proposes a decision business recommendation method for apps development based on Large Language Model which can analyze the customer review segmentation. Specifically, the pro- posed method extract the context, topic, sentiment, and emotion altogether which were neglected by most studies. The proposed method contains two main stages: (i) multi-stage LLM customer review segmen- tation prompting to share the knowledge about informative customer reviews only, and (ii) retrieval- augmented generation-related business information to recommend the decision business for apps devel- opment. The validation of this Thesis is based on data crawling of customer reviews of an apps from a company in Indonesia. The findings of this study demonstrate the effectiveness of the proposed method, particularly in producing accurate review classification and segmentation, as well as strategic and rel- evant business recommendations. Overall, this research significantly contributes to developing a robust framework for analyzing customer reviews and opens avenues for enhancing future business decision support systems based on Large Language Models (LLM) and Retrieval-Augmented Generation (RAG).
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Sistem Rekomendasi, Pengembangan Aplikasi, Ulasan Pengguna, Large Language Model, Retrieval-Augmented Generation. Customer reviews are a vital strategic asset for companies to understand customer needs and iden- tify opportunities for service improvement. Most related studies offerred poor quality business recom- mendation as they ignored the unrelated customer reviews towards the available product or services. Analysis on customer reviews are crucial to increase the customer satisfactory levels, identify the draw- backs, strength the customer loyalty, and encourage the innovation. Therefore, this Thesis aims to provide a recommendation of business decision on apps development inline with customer review segmentation analysis. This Thesis proposes a decision business recommendation method for apps development based on Large Language Model which can analyze the customer review segmentation. Specifically, the pro- posed method extract the context, topic, sentiment, and emotion altogether which were neglected by most studies. The proposed method contains two main stages: (i) multi-stage LLM customer review segmen- tation prompting to share the knowledge about informative customer reviews only, and (ii) retrieval- augmented generation-related business information to recommend the decision business for apps devel- opment. The validation of this Thesis is based on data crawling of customer reviews of an apps from a company in Indonesia. The findings of this study demonstrate the effectiveness of the proposed method, particularly in producing accurate review classification and segmentation, as well as strategic and rel- evant business recommendations. Overall, this research significantly contributes to developing a robust framework for analyzing customer reviews and opens avenues for enhancing future business decision support systems based on Large Language Models (LLM) and Retrieval-Augmented Generation (RAG). |
| Subjects: | Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA278.55 Cluster analysis Q Science > QA Mathematics > QA336 Artificial Intelligence |
| Divisions: | Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT) |
| Depositing User: | Aka Arinda |
| Date Deposited: | 30 Jul 2025 02:56 |
| Last Modified: | 30 Jul 2025 02:56 |
| URI: | http://repository.its.ac.id/id/eprint/123126 |
Actions (login required)
 |
View Item |