SC publications has been a pioneer in education publication since half a century in India. With more than XYZ published books that cater to the whole range of K-12 education system in India, the publication house has been critical in student’s life across the nation. They have ventured into digital transformation of the content for almost a decade and have a wide range of content available for digital consumption for the school students. With an explosion in EdTech applications in Indian diaspora, SC publications were lagging behind these personalized content delivery systems.
SC Publications approached Palpx to develop a personalized content delivery algorithms for better user retention and gaining a competitive edge in the highly competitive market.
With the new product feature on horizon, the publication house managed to win various bids which could have resulted in school contracts going to its competitors.
Challenges we solved
Preparing the base data : Enabling age old questions/content with deep tagging for smarter insights.
Schools still assess a students performance with marks they score in examinations. It is a very limited understanding of students actual abilities. To generate a much more personalized assessment and content delivery system, all the educational content needed to be understood and tagged for various insights that can be gained. For example : A question with multiple tagging system.
The videos were converted from older flash format to new generation format and with added tags which enabled advanced searching capabilities within video content as well.
This tagging was done to almost 1000+ videos spanning across k-12 syllabus.
But the most daunting task was to tag the assessment system, which included a massive 60000+ database of questions. This included setting up a team of subject matter experts, education experts & psychologists across the range of K-12 education system and assigning a system of ability, skillset, difficulty, blooms taxonomy based tags to each of the questions. Palpx is developing a machine learning system which can now understand a new question thrown at it and can define automatic tags with 90% accuracy. This has massive implications in future which allows any new content and questions to be input into the system and system can automatically assign right tags with 90% accuracy.
Advanced Content suggestion & Deep Assessment
Based on the in-depth tagged content, we built a lot of algorithms at every stage of content consumption. Measuring how much time and how many times a student spends on a particular video to what pace the student reads a particular article, the system is built to enable very deep insights into student learning behavior and to learn a students pitfalls and strengths with much more accuracy. Every test question a student goes through helps the system to assess his abilties, his skillset, his thought process in line with blooms taxonomy and much more. This creates a very deep profile of students learning DNA. This is then used to suggest right content and directions at every step of learning. Such deep assessment is unique in the industry and is shaping up to be a very powerful platfrom for students, teachers and parents alike.
Co-creating the future together.
This is a project which could not have been possible with equal contribution from both the sides. We approached the project on a partnership basis where we had equal stakes in finding business gaps and developing the product together. From having brainstroming sessions together to arrive at the best approach towards solving problems that came in between to forming teams that included both SC publications and Palpx team working together. (as in case of forming a team of subject matter experts and developing the application codes together). This helped in fast forwarding the project by leaps and bounds.
Solutions & Technologies Involved
User Experience, Research, Natural Language Processing, Predictive modelling