Logo Lanfrica

sam23121/sentiment_analysis_amh

Domaine:

natural language processing

Type de record:

software
Créateur:
sam
Hôte:
A sentiment analysis tool for the Amharic language ### Amharic Sentiment Analysis This repository contains a Python-based tool for performing sentiment analysis on Amharic text. Amharic is the official language of Ethiopia and one of the Semitic languages spoken in the Horn of Africa region. #### Features - Sentiment classification: Classify Amharic text as either positive, negative, or neutral sentiment. - Polarity scoring: Assign a numerical sentiment score ranging from -1 (negative) to +1 (positive) for Amharic text. - fine-tuned model: The tool uses a pre-trained model and fine tunes it on specifically trained on Amharic data. which can be found in the data folder - for the word vector I used fasttext because it showed better results for the amharic langugage. I haven't uploaded it because it was large but can be found on the fasttext website #### Short cuts If someone doesn't want the witty griddy details on how to fine tune a model and just wants a short implementation, the hugging_face folder conssists of a translation model and zero shot classification model which uses the NLLB (No langugage left Behind) model by facebook and DeBERTa-v3-base-mnli-fever-anli from MoritzLaurer respectively. ###### Steps 1. first use the nllb model to translate from amahric to english 2. then give a zero shot classification on that english text Also there is an implemtation using gradio if you want to see it on a ui.