OsinachiEzemba/South-African-Language-Identification

Identifying South-African Languages using Natural Language Processing # South-African-Language-Identification **Identifying South-African Languages using Natural Language Processing** **NOTE:** Due to the nature of this project, code cannot be shared publicly. South African Language Identification was a project aimed at identifying which of the South Africa's 11 official language a text is in. The project was conducted to provide valuable insights into the trends and patterns of energy generation, with the ultimate goal of supporting the government's decision-making process regarding renewable energy infrastructure investments. # Project Scope **The project involved the following key objectives:** 1. Data Analysis: A comprehensive analysis of the supplied data was performed to understand its structure, identify potential errors or inconsistencies, and explore the relationships between variables. This step ensured the accuracy and reliability of the subsequent analysis and modeling stages. 2. Feature Engineering: Changing My Target to Labels and changing the texts in our dataset to numbers 3. Model Building: Various classification models were developed and evaluated to classify language texts accurately. In this project, a Naive Bayes model was trained and fine-tuned to achieve optimal performance and predictive accuracy. 5. Model Evaluation: The performance of the developed model, along with other candidate models, was rigorously evaluated using appropriate evaluation metrics and techniques. This step allowed for the selection of the best-performing model for the project. 6. Model Explanation: The inner workings of the selected model were explained in a clear and concise manner to a non-technical audience. The explanation included insights into how the model made predictions and the importance of various features. # Conclusion The South African Language Identification Hack 2023 project successfully achieved its objectives, having been able to successfuly classified texts into different SouthAfrica's 11 distinct languages.