ZIDAN MUBARAKBANDA ACEH, ID
Loading portfolio
Loading
ZIDAN MUBARAK
ENID
All work

04 / 06

Recommendation

NusantaraGo

A tourism recommender for Indonesia: TF-IDF content similarity plus collaborative filtering, served through Flask.

Source (opens in a new tab)
Role
Top contributor, machine learning
Year
2025
Stack
Python · Flask · scikit-learn · TensorFlow · NLTK · Pandas
Reading time
Project note
Team
Four-person capstone team; 107 commits by Zidan
Abstract typographic composition representing the NusantaraGo recommendation model.
Fig. 01 · Generated visualOwner-approved abstract typographic visual. It is not a screenshot of the application.

01 · Overview

What it is

A tourism recommender for Indonesia: TF-IDF content similarity plus collaborative filtering, served through Flask.

Built during the Coding Camp powered by DBS Foundation cohort. The team scraped attraction data province by province, cleaned it, and combined content-based, popularity and location signals into a hybrid recommender behind a Flask front end.

02 · Highlights

Highlights

  1. 01

    Content-based filtering with TF-IDF and cosine similarity, plus collaborative filtering.

  2. 02

    Hybrid ranking that also weighs popularity and location.

  3. 03

    Attraction data scraped and cleaned province by province, stored as CSV and JSON.

  4. 04

    107 commits, the largest contribution in the four-person team.

03 · Stack

Stack

  • Python
  • Flask
  • scikit-learn
  • TensorFlow
  • NLTK
  • Pandas

Sources