Welcome to my NLP Portfolio, where I showcase a range of projects in Natural Language Processing (NLP). As an AI engineer passionate about Python and Rust programming languages, I strive to provide efficient and fundamental solutions in the NLP domain. Explore the projects below, each designed to tackle unique challenges in text analysis and language understanding.
1. Fake News Classification
Overview: Distinguishing between real and fake news articles using machine learning.
Aim: Enhancing information reliability and combating misinformation.
Technologies: Python, TensorFlow, PyTorch, scikit-learn.
2. AI Summit Challenge by IndabaXKenya and DeKUT
Overview: Classifying text from university students in Kenya towards a mental health chatbot.
Aim: Addressing mental health challenges through AI-powered chatbot solutions.
Technologies: Python, TensorFlow, PyTorch, scikit-learn.
3. Part-of-Speech (POS) Tagging Challenge
Overview: Creating a machine learning solution for POS tagging in African languages.
Aim: Preparing low-resource African languages for NLP tasks.
Technologies: Python, TensorFlow, PyTorch, scikit-learn.
4. Swahili News Classification Challenge
Overview: Developing a multi-class classification model for Swahili news.
Aim: Classifying news content into predefined categories for effective organization and accessibility.
Technologies: Python, Fastai Blurr, Transformers.
5. Modeling Fake Climate Change Data
Overview: Exploring a dataset adopting the FEVER methodology to model real-world claims related to climate change.
Aim: Developing models to discern the veracity of climate change-related claims.
Technologies: Python, Transformers.
6. COVID-19 Tweets Identification
- Overview: Identifying tweets related to coronavirus without relying on traditional keyword-based approaches.
- Aim: Enhancing accuracy and efficiency in identifying COVID-19-related content on Twitter.
- Technologies: Python, Transformers, scikit-learn.
7. To Vaccinate or Not to Vaccinate: It's not a Question
Overview: Analyzing social media sentiment towards vaccines, particularly focusing on COVID-19 vaccines.
Aim: Developing a machine learning model to categorize Twitter posts related to vaccinations as positive, neutral, or negative.
Technologies: Python, NLTK, scikit-learn, PyTorch, Transformers.
1. GPT Finetuning Template
Subdirectory: nlp_templates/GPTS
Description: Template for fine-tuning the GPT model for various NLP tasks.
2. Named Entity Recognition (NER) Template
Subdirectory: nlp_templates/NER
Description: Template for NER tasks, facilitating efficient model training for entity recognition.
3. Sequence Classification Template
Subdirectory: nlp_templates/SEQ_CLASSIFICATION
Description: Template for sequence classification tasks, streamlining the development of models for classifying sequences.
4. DeBERTa Multiple Choice Classification Template
Subdirectory: nlp_templates/DEBERTA
Description: Template for multiple choice classification using the DeBERTa model, providing a foundation for efficient model development.
As I continue to explore the frontiers of NLP, expect more innovative projects that leverage the power of language understanding. Stay tuned for updates and new developments in the realm of Natural Language Processing.
Thank you for exploring my expanded NLP Portfolio!