Skip to content

About BASA

What is BASA?

BASA is an open-source NLP toolkit for Indonesian and regional languages.

The project focuses on building practical, developer-friendly tools for Indonesian text processing, beginning with informal text normalization and gradually expanding toward regional language support, evaluation utilities, and dataset tooling.

The name BASA comes from the Indonesian and Javanese word bahasa, reflecting the project's long-term vision of supporting Indonesia's linguistic diversity.

Creator

BASA was created by Muanai Khalifah Revindo, an Informatics Engineering student at Universitas Sriwijaya.

His work primarily focuses on:

  • Artificial Intelligence
  • Machine Learning
  • Risk Intelligence Systems
  • Operational Analytics
  • Indonesian NLP
  • Low-resource language technologies

BASA represents an effort to build practical open-source infrastructure for Indonesian and regional-language NLP ecosystems.

Why BASA Exists

Many modern NLP tools are designed primarily around English-language workflows.

Indonesian developers often need to solve challenges such as:

  • Social media slang normalization
  • Repeated-character reduction
  • Informal text preprocessing
  • Regional language support
  • Low-resource dataset preparation

BASA aims to address these problems through simple APIs, conservative defaults, and strong developer experience.

Design Principles

BASA follows several core principles:

Conservative by Default

Potentially destructive operations, such as typo correction, are opt-in rather than automatic.

Indonesian First

The project prioritizes Indonesian NLP problems before expanding to broader multilingual use cases.

Incremental Growth

Features are introduced gradually, with an emphasis on maintainability, testing, and documentation.

Developer Experience

Most workflows should require only a few lines of code:

from basa import quick

quick("gw gk ngerti bngtttt")

Long-Term Vision

The long-term goal of BASA is to become a practical ecosystem for Indonesian and regional-language NLP, including:

  • Text normalization
  • Translation utilities
  • Evaluation frameworks
  • Dataset tooling
  • Synthetic data generation
  • Low-resource language support

The project values practical usability and sustainable open-source development over rapid feature expansion.

Open Source

BASA is an open-source project released under the MIT License.

Community contributions, feedback, and collaboration are always welcome.