The client is one of the most renowned publishing companies globally.
The client wanted to integrate a reliable and efficient assessment system to assess the oral fluency of non-native English speakers. Their existing system was time-consuming and relied on manual evaluation, making it prone to errors and inconsistencies.
Critical Success Factors
- Find an AI system with ability to accurately evaluate oral fluency, including factors like pronunciation, grammar, vocabulary, and intonation.
- The flexibility to accommodate a diverse range of accents and dialects.
- User-friendliness and easy navigation.
- Researched and identified multiple solutions available in the market and reviewed how each solution utilized Natural Language Processing (NLP) and Machine Learning (ML) techniques.
- Adopted an AI-based assessment tool consisting of two components: a speech recognition system and a scoring engine.
- The speech recognition system utilized NLP techniques to transcribe spoken words into text.
- The scoring engine employed ML algorithms to evaluate the text according to predefined criteria and generate a score.