Page 395 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

Additionally, AI-driven retention strategies will be implemented:

• Engagement Prediction: Identifying volunteers at risk of disengagement.
• Leadership Identification: Recommending leadership roles for highly engaged volunteers.
• Personalized Communication: Sending tailored reminders and recognition messages. 4.4-Productivity

Example: A volunteer who frequently participates in food distribution may receive a message like:
"You’ve supported 500 families this year—would you like to join our leadership program?"

3 Use Case Requirements

• REQ-01: Machine Learning Models – predictive analytics, analyzing historical data to
promote NPO initiatives and volunteers matching / retention
• REQ-02: User friendly interface – To interact with the AI driven name validator
AI Models of the solution

The endpoint relies on Natural Language Processing (NLP) models, specifically:

• Semantic Similarity Models (Sentence-BERT [1]) to compare the meaning of names in
Arabic and English.
• Fuzzy Matching Algorithms ( Levenshtein distance, token sort ratio) for partial or character-
level similarity.
• These models act like a recommendation engine, suggesting close or semantically similar
names based on a given input.

359