Emerging Topic Recommendation System
Recommendation system that surfaces emerging topics for Telkom's Digital Business Directorate.
The Challenge
Telkom's Digital Business Directorate needed to spot topics that were starting to gain traction early, rather than discovering them after they had already peaked.
Technical Approach
Built an Emerging Topic Recommendation system that extracts topics from text data and recommends the ones showing emerging trend signals.
Outcome & Impact
Gave the Digital Business Directorate a recommendation feed of emerging topics to support earlier, data-driven decisions.
How It Works
- 01Text Data
- 02Topic Extraction
- 03Trend Signal
- 04Emerging Topic Recommendation
More case studies
All projectsAspect-Based Sentiment Analysis (ABSA) Engine with IndoBERT
Sentiment scored per aspect, so "fast app, slow CS" becomes two actionable signals instead of one muddy score.
NLPNER & Statement Extraction with Entity Normalization
Answers "who said what about whom" by linking actors to their statements, including implicit references.
Data EngineeringReal-Time NLP Platform with Kafka & Event-Driven Architecture
Event-driven streaming platform that surfaces shifts in public conversation as they happen, not in tomorrow's report.