Lelapa AI Buzuzu-Mavi Challenge
Helping Africa
$1 300 USD
Completed (11 months ago)
Natural Language Processing
Sentiment Analysis
Machine Translation
492 joined
118 active
Start
Jan 09, 25
Close
Apr 06, 25
Reveal
Apr 07, 25
Please what are the labels to use for sentiment analysis?
Help Ā· 19 Feb 2025, 12:01 Ā· 4
I need help for those with successful submissions. I don't know which labels to use for the sentiment analysis.
I tried == > 0 : Negative, 1 : Neutral , 2 : Positive
I tried also the original labels : 'Kyakkyawa', 'Tsaka-tsaki', 'Wastani', 'Korau', 'Hasi', 'Chanya'
But my score didn't change. It's like other values are expected for the scoring script.
Hey @nostml :)
For sentiment prediction, the scoring script uses 0, 1, and 2 as the expected labels:
There’s no need to modify or manually adjust the labels during submission, the script will handle the predictions automatically. Just ensure your outputs align with these values.
Are we allowed to use the anyother inkuba dataset for training apart from the one provided in the competition?
I found this from the github: https://github.com/Lelapa-AI/zindi-inkuba-notebook/blob/main/Notebooks/Starter-Notebook.ipynb
Maybe {0: positive, 1:negative, 2:neutral}?
"sent_instruction = "Please identify the sentiment reflected in this text based on the following guidelines: Positive: if a text implies positive sentiment, attitude, and emotional state. Negative: if a text implies negative sentiment or emotion. Neutral: if a text does not imply positive or negative language directly or indirectly. Provide sentiment labels only""@nostml, I found it out!
- 0: positive
- 1: neutral
- 2: negative
https://github.com/Lelapa-AI/zindi-inkuba-notebook/blob/main/utils/eval.py#L18-L22