SBIR-STTR Award

This SBIR Phase I project will develop an Artificial Intelligence (AI) /Machine Learning (ML) driven solution to automate the process of giving qualitative feedback on student reflections. The automated reading of student reflections for quality and content is a new and innovative technology that is currently not available in the market, as all applications in this realm focus on grammar and writing structures. The innovation will understand the specific actions and strategies students are using (many of which are unique to an individual learner), and help students get the support they need to improve in those techniques or try new ones. This process would build learning skills and growth-oriented belief systems in students, both of which lead to significantly improved academic and career outcomes. Fundamentally, the proposed project will help students learn how to learn.This SBIR Phase I project will build the first phase of personalized insights for impactful student goal-setting and reflection. The company will leverage natural language processing (NLP) and ML algorithms to 1) parse, process, and analyze reflections written by students, 2) return a score for each reflection based on a research-based rubric, and 3) notify teachers when a student has a low score and needs support. At the end of Phase I, the company will examine whether the prototype functions as intended with 90%+ accuracy, if teachers can integrate the prototype into their classroom practice, and if it shows early promise of improving the quality of student reflections and learning. With a proprietary dataset of 400K+ written student reflections and data points, the company is uniquely positioned to develop and scale this product vision. Phase II research will develop personalized feedback, strategies and suggestions to help students and teachers find the best strategies faster.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

The broader impact/commercial potential of the Small Business Innovation Research (SBIR) Phase II project may be a fundamental shift in academic and life outcomes for students, especially those from low-income and vulnerable populations, by building metacognition and Social Emotional Learning (SEL) skills. This project will advance the company's technology that builds critical SEL and metacognition skills in a way that is student-led and integrated with academic content and routines. The effort expands teacher capacity in a measurable way. The project will apply Natural Language Processing (NLP) techniques to the company?s proprietary data of student reflections and teacher feedback, will train Machine Learning (ML) driven predictive algorithms to measure reflection quality and will build a learning strategies recommendation engine. This innovation will allow teachers and students to expand their toolkits of effective pedagogically-sound learning strategies. The deeper application of the innovation unlocks commercialization potential coupled with transformational outcomes for student learning.The Small Business Innovation Research Phase II project will build on the company's NSF-funded research and Phase I results that support the feasibility and commercialization of applying NLP techniques and ML algorithms based on the company?s proprietary data to ?read? student reflections, rate them on a quality rubric, and help teachers provide feedback. Stage I of this project will be a lower effort, fast value objective to apply data science and analytics to roll-up real-time quality assessments to measure growth at the student. The project will provide reports on the reflection quality to school administrators and feedback tips to teachers based on student reflection quality and classroom context. Stage II would be a higher effort. One objective will be to build a Strategy Recommendation Engine driven by NLP and ML techniques and based on the latest research in pedagogy and learning sciences. This technology will analyze past reflections to identify which learning strategies a student has tried, then provide teachers with suggestions on pedagogy-driven strategies they can be recommended to improve the students' learning.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.