AI Use Cases in Education

Application of Artificial Intelligence especially ML-DL (Machine Learning – Deep Learning) in Education is still primitive compare to other industry and sectors. One of the prime reason is absence of big behavioral data. Students and parents are attached to the institutes. Institutes don’t capture their behavioral data. They don’t consider to deploy a platform for such data capture being unaware of possible return. Large education institutes with thousands of students can adopt such platform to benefit from it. There are many private pre and k-12 school chains who has hundreds of institutes and more than hundred thousand students. States/governments can adopt such platform at any level. There is also a mindset of no considering students/parents as commercial users.

Here are my ideas where big behavioral data can be useful by reducing manual waste or allowing early intervention thus improving outcome.

Dyslexia detection

Children struggling with reading and writing disability who needs early attention can be beneficial by using this advanced technology. These children lacks concentration and their eye movement is random compare to general trend. Normally anybody reads in a particular pattern like left to right, top to bottom. It is possible to capture eye ball movement by attaching sensors in display unit. Data can be stored and compared with thousands of such data stored in the big data database to get the degree of attention deficiency. Manual observation takes long time, few months to few years. Such comparison can be done quickly allowing medical and intervention help to apply early.

Assessment

MCQ, EMQ based or voice based examination are already widely adopted, be it for quick assessment or entrance examination. There are many merits of objective tests. However, it has many de-merits as well, like it fails to assess creativity, writing skill, subjective questions, etc. Assessment of these type of skills requires huge manual effort and depends on assessor capability and bias. Machine learning technology powered by natural language processing (NLP) especially with latest development in deep learning can fill up this void. It is possible to evaluate writing skill, subjective questions by adoption of technology which has human brain like functionality. This will be one of the complex solution to achieve though.

Child Future Stream/Career Prediction

Children are natural learners. Each one of them is special and they have special likes and dis-likes. In group learning or class room scenario, their interests are ignored often. Their future journey and counselling depends on teachers’ observation capability and annual summative assessment. Observations are also not always recorded properly and not available on time. It is possible to capture student behavioral data over a period based on their activities (and discipline) in lab, library, online content search, class test, games, SUPW, cultural and other events, and attendance. A supervised learning model can be helpful to predict a stream/career where the student could be more successful as professional.

Personalized Learning Centre

Every child has their own grasping capability and takes their own time to learn. Pushing a child through standard curriculum and pace mayn’t give the desired outcome. A child learns best when curriculum is personalized and teacher works as facilitator. Personalized learning requires significant attention and investment where technology can play its role bigtime. Machine learning can suggest new topic based on the children pace, progress and performance. Many new initiatives has made progress in this area however more can be achieved.

Digital Marketing

Knowing students/parents behavioral data (choices, browsing history, etc.) can help to sale new education service/product/inputs. Deep learning can use these data to predict next intervention.

It is possible to improve education outcome and leapfrog rote learning by adopting strategy powered by technology. Technology adoption in the above areas are becoming popular. Institutes, who are pioneer in adopting such strategy will remain ahead in terms of efficiency, better learning outcome and revenue generation. While we are working on these use cases, you may leave your suggestions or ideas of other use cases.