About Me

Hi, I’m Alleluia 👋. I’m a Computer Science student at BINUS University, majoring in Data Science. My main interests are Machine Learning, Artificial Intelligence, and Data Science applications.

I enjoy building models that can solve real-world problems, from detecting spam reviews to forecasting customer demand and predicting user behavior. I also like sharing knowledge through teaching and collaborating with others.

Currently, I’m pursuing the Fast Track program to continue my Master’s degree in Computer Science, while developing projects and research in applied machine learning.

Contact

• Email: ndolualleluia@gmail.com
• Linkedin: Alleluia Rehuellah Alefika Ndolu

Education

BINUS University | 2022 - Present

S.Kom. Computer Science (Data Science Major), Expected 2026.

M.Kom. Fast Track Master of Computer Science, Expected 2027.

Professional Experience

Demand Data Analyst Intern @ PT Arnott’s Indonesia (February 2025 – August 2025)

Built forecasting & NLP models (IndoBERT sentiment analysis), deployed dashboards with Streamlit.

Part-time Teacher - Coding and Programming for Children @ PT Timedoor Academy (December 2023 - September 2024)

Taught programming basics to students, designed simple coding projects, fostered interest in technology.

Projects

🔹 NLP Projects

Spam Detection for Tiktok App Reviews on Google Play Store (IEEE IAICT Publication) | 2025

Research Project See Repository! Read on IEEE Xplore

• Description: Developed a two-stage spam detection framework for TikTok reviews (spam vs non-spam; sensitive vs insensitive).
• Tech Stack: Python, Scikit-learn, TensorFlow/Keras, TF-IDF, Word2Vec
• Results: Achieved 96% accuracy using MLP with TF-IDF.
• Publication: Co-authored IEEE IAICT 2025 paper.

Multiclass Text Classification Using LLMs | 2025

Text Mining Project See Repository!

• Description: Classified scraped Indonesian sports news into 5 categories using IndoBERT (Hugging Face Transformers). Compared 1-stage vs 2-stage classification approaches.
• Tech Stack: Python, Hugging Face Transformers, Pandas, Scikit-learn
• Results: Improved performance with hierarchical 2-stage classification.

Public Opinion Analysis on Digital Transformation (Clustering & Topic Modeling | 2025

Text Mining Project See Repository!

• Description: Scraped YouTube comments to analyze public sentiment on digital transformation in Indonesia. Applied clustering (K-Means, silhouette evaluation) and topic modeling (LDA).
• Tech Stack: Python, Scikit-learn, Gensim, TF-IDF, Word2Vec
• Results: Identified distinct audience segments and extracted topics related to opportunities and challenges of digital transformation.

🔹 Other ML/DL Projects

Batik Image Classification using PyTorch | 2024

Deep Learning Project See Repository!

• Description: Built an image classification model to detect and predict batik motifs using PyTorch.
• Tech Stack: Python, PyTorch, NumPy, Matplotlib
• Results: Achieved high accuracy in motif recognition, demonstrating strong understanding of deep learning concepts.

Bank Customer Churn Prediction Model with Machine Learning | 2024

Model Deployment Project See repository!

• Description: Developed a machine learning model to predict customer churn using Random Forest and XGBoost.
• Tech Stack: Python, Pandas, Scikit-learn, XGBoost, FastAPI, Streamlit
• Results: Identified key churn drivers and deployed the model into an interactive app for business use.

Bank Loan Granting Prediction using Deep Learning | 2024

Deep Learning Mid-Exam Project See repository!

• Description: Built a deep learning model to predict loan granting decisions with TensorFlow Keras.
• Tech Stack: Python, TensorFlow Keras, Scikit-learn
• Results: Improved model accuracy by applying class weighting, feature scaling, and hyperparameter tuning.

Image Classification using Custom VGG16 Architecture | 2024

Deep Learning Project See repository!

• Description: Implemented a custom VGG16 architecture from scratch with batch normalization and dropout.
• Tech Stack: Python, PyTorch, NumPy, Matplotlib
• Results: Enhanced accuracy on CIFAR-10 dataset through optimized hyperparameters and regularization techniques.