Mauluddi, Mochammad Reyhan (2025) Sistem Rekomendasi Strategi Evaluasi Top Brand Award sebagai Unit Bisnis Menggunakan Model OLLAMA untuk Meningkatkan Nilai Kompetitif. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6032222003-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6032222003-Master_Thesis.pdf Restricted to Repository staff only Download (9MB) | Request a copy |
Abstract
Top Brand Award (TBA) telah menjadi tolok ukur keberhasilan merek di Indonesia sejak 2000. Nilai kompetitif TBA terdapat pada indikator dalam Top Brand Index, namun belum sepenuhnya mencerminkan dinamika pasar yang berkembang pesat. iperlukan pendekatan adaptif berbasis data untuk meningkatkan nilai kompetitif TBA. Saat ini, strategi TBA masih bergantung pada survei dan wawancara manual yang kurang responsif terhadap perubahan pasar. Penelitian ini bertujuan memberikan rekomendasi strategi evaluasi TBA sebagai unit bisnis berbasis data dan analisis pasar. Sistem rekomendasi dikonstruksi menggunakan model Omni-Layer Learning Language Acquisition (OLLAMA) dengan data media sosial dan studi TBA. Sistem terdiri dari tiga modul: (1) analisis media sosial dengan multi-stage prompting (fase ekstraksi topik, insight, dan rekomendasi strategi); (2) analisis pasar dengan pendekatan serupa; dan (3) integrasi hasil kedua analisis untuk menghasilkan rekomendasi strategis berbasis data. Evaluasi dilakukan menggunakan word cloud, bobot kata, kesesuaian prompt-output dengan sentence transformer, dan diversity score. Hasil menunjukkan model Deepseek dan Llama3.2 menghasilkan rekomendasi sangat beragam dengan kesesuaian sedang hingga sangat tinggi. Deepseek lebih unggul pada analisis berbasis kategori media sosial, produk, dan jurnal. Rekomendasi akhir dianalisis dengan SWOT untuk mendukung keputusan manajerial. Penelitian ini membuktikan integrasi LLM dan multi-stage prompting dapat meningkatkan relevansi dan nilai strategis TBA agar lebih adaptif terhadap dinamika pasar.
===================================================================================================================================
Top Brand Award (TBA) has been a benchmark for brand success in Indonesia since 2000. The competitive value of TBA lies in the indicators of the Top Brand Index, but these indicators do not fully reflect the rapidly changing market dynamics. A more adaptive, datadriven approach is needed to enhance TBA’s competitiveness. Currently, TBA strategies still rely heavily on manual methods such as surveys and interviews, which are less responsive to market changes. This study aims to provide recommendations for TBA’s evaluation strategy as a business unit based on data and market analysis. The recommendation system was developed using the Omni-Layer Learning Language Acquisition (OLLAMA) model with social media data and TBA study analysis. The system consists of three modules: (1) social media analysis using multi-stage prompting (topic extraction, insights, and strategy recommendations); (2) market analysis with a similar approach; and (3) integration of both analyses to produce datadriven strategic recommendations. The evaluation was done using word clouds, word weighting, prompt-output alignment with a sentence transformer, and a diversity score. Results show that both Deepseek and Llama3.2 models produced highly diverse recommendations, with moderate to very high alignment between prompts and outputs. Deepseek performed better in social media category-based, product-based, and journal-based analyses. The final strategy recommendations were analyzed using SWOT to support managerial decision-making. This study demonstrates that integrating LLMs with multi-stage prompting can improve the relevance and strategic value of TBA in responding to market dynamics.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Sistem Rekomendasi, Top Brand Award, Large Language Model, Multi-stage Prompting |
| Subjects: | A General Works > AI Indexes (General) A General Works > AI Indexes (General) T Technology > T Technology (General) |
| Divisions: | Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT) |
| Depositing User: | Mochammad Reyhan Mauluddi |
| Date Deposited: | 30 Jul 2025 02:45 |
| Last Modified: | 30 Jul 2025 02:45 |
| URI: | http://repository.its.ac.id/id/eprint/122729 |
Actions (login required)
 |
View Item |