Smart Mentor

Empowering Learning for a Brighter Future

With SmartMentor, every student can have a personalised AI tutor, mentor, and partner dedicated to helping them achieve their learning goals.

This innovative system powers individualised learning through structured learning dialogues, tailored resources, and data-driven insights that continuously improve the experience. 

Designed for international school students and parents seeking an internationalised-education for their children, the system guides learners to reach their potential in English and develops the skills to thrive in a global future.

By understanding each learner’s strengths, needs, and goals, SmartMentor opens possibilities once thought impossible.

• Automatic speech recognition, natural language processing, and machine learning technologies tailor interactions to each student’s unique needs, as well as continual learning and improvement over time.
• In 15-minute learning sessions, SmartMentor engages students in stimulating learning dialogues using speech and text. Through these conversations, the system opens students to a world of inspiring multimedia and engaging texts organised around their curriculum. Content is personalised to their skills, goals, and interests.
• As students read, listen, speak, and write, SmartMentor tracks their progress. It serves as a partner that gets to know their strengths and opportunities for growth, then pushes them to achieve more than they thought possible.
• By capturing all their work and data, the system builds an in-depth learning portfolio for each student. Students and teachers gain a holistic view of achievements and challenges, fuelling data-driven insights to enhance the curriculum and help all students thrive.

Supercharged Learning

SmartMentor offers unmatched value through:

Structured Learning Dialogues

Through dynamic conversations, the mentor sharpens skills, opens new worlds of knowledge, and helps learners achieve more than they thought possible. 

Each learning dialogue features:

Overview of Performance Metrics

SmartMentor uses a range of performance metrics to identify the learners’ level, track progress and provide personalised feedback.

By gathering and analysing these performance metrics, SmartMentor can provide recommendations that are tailored to the learner’s specific needs and goals. 

In addition, SmartMentor uses the information to generate reports and visualisations that summarise the learner’s progress and achievements over time, highlighting areas of strength and weakness.

Communicative competence

Measuring communicative competence can be a challenging task for an AI system, as it requires an understanding of language use in a range of contexts, including social, cultural, and situational factors.

However, SmartMentor will use a variety of techniques to measure a learner’s growing communicative competence, including:

Scoring rubrics: SmartMentor can use scoring rubrics to assess learners’ communicative competence in specific language skills: speaking, writing, reading and listening. These rubrics can be developed based on a set of predefined criteria for each language skill, which can be used to assess learners’ performance on specific tasks. We suggest adapting relevant CASE level descriptors and CEFR descriptors to operationalise scoring rubrics in the system.

Natural language processing (NLP): SmartMentor can use NLP techniques to analyse learners’ language use in real-time during interactive dialogues, such as detecting errors in grammar, syntax, and vocabulary use. NLP can also be used to measure learners’ proficiency in specific language skills, such as reading and listening comprehension.

Adaptive testing: SmartMentor can use adaptive testing techniques to measure learners’ communicative competence by tailoring the difficulty of questions and tasks to their individual abilities. This approach can provide a more accurate and detailed assessment of learners’ language proficiency by adjusting the level of difficulty based on their previous responses.

Analysis of learner data: SmartMentor can analyse data on learners’ performance over time to identify patterns and trends in their communicative competence. This can include data on learners’ accuracy, fluency, complexity, and appropriateness in language use across different language skills and contexts.

Learning Process

SmartMentor does more than provide instruction and content. The system actively supports learners to enhance their ability to learn effectively and efficiently.

In each learning dialogue, the AI gathers and analyses data to create a detailed profile of each learner’s strengths and identify areas for further development.

Here is an overview of data the system will use:

SmartMentor combines these data points to create a grade, using an algorithm that weighs each metric according to its importance and relevance to the learner’s overall performance. 

For example, time spent on the platform may be given less weight than the frequency and type of mistakes made, as these metrics are more directly related to the learner’s language competence. 

The algorithm can then generate a score or grade that reflects the learner’s strengths, weaknesses, and overall level, and provide recommendations for further improvement.

Training Data

The effectiveness of an AI-powered learning assistant is highly dependent on the quality and quantity of data used to train the system. For this study, the AI system will be trained on a large dataset of levelled authentic content in English – text, audio and video – and language learning materials and supplementary resources.

The dataset will be selected from reliable sources that are in alignment with international benchmarks for English proficiency as well as local curriculum content areas, values and attitudes, key concepts and generic skills.

Safety is a priority. The dataset will be filtered and pre-processed to remove any inappropriate or sensitive content, such as offensive language or cultural stereotypes, to ensure that the system is safe and appropriate for young learners to use.

A Dynamic Learning Portfolio

The AI-powered learning assistant will collate a variety of artefacts created by the learner to generate a comprehensive learning portfolio.

Some of the specific types of artefacts include: