While preparing for some examinations after graduating from secondary school, I faced a huge difficulty while searching for some genuine, complete guide to several topics over any online platform. I was very disappointed with
what people were putting over the web and I went back to books, I purchased some books as I had to qualify for exams. The major problem I faced was scattered information over the internet, so I thought why can’t someone just take
all of those over at one platform at one place? And so I thought of a solution. I found that there should be batches or clusters of different topics with easy navigation to all. But how can we create that without putting a
huge amount of money into it? Any solution?
Artificial Intelligence in education might sound a little strange to some people about how AI can teach us ‘how to learn’ or ‘what to learn’. Basically, we have been in a bubble
of limited thoughts where haven’t we used that big data that is already cooked up for use. If we think about it deeper, we can find that using the right models over the data that we have from online test portals, universities and
platforms like google classroom can make a big change in understanding how to make the academic structure clearer to the students.
There are two basic things that we need to understand, the use of left out data and Guiding over good solution for better academic fields slash curriculum. Let’s understand what I mean with these two terms. Suppose a university
has data about the results of its students and the data about the curriculum in the past hundred years. Let’s assume 145 million students have been graduated from that university and all the data is there in their records. But
what is the use of that data? If we look closer, many institutions have a good amount of data which they can use to improve their teaching efficiency.
Let’s take an example with several factors in hand. There are two universities, say ‘university A’ and ‘University B’. A student who qualified in 2019 from university A, major in mechanical sciences, failed to pass the examination
of architectural drawing. The university A then took the data from several students who have also failed in the past 10 years and evaluated the results. The committee found two cases, one is that those students who did not opt
‘engineering drawing’ in their first semester, got very low GPA as compared to those who have opted for ED in their first semester. Second, the students who were part of the football club, also scored a low GPA. So the university
committee decided to make some rules over opting any course or giving fundamental courses before taking any newly introduced course for any newbie and to see if someone is a member of any of the sports clubs. On the other hand,
the University B with sort of the same case decided to look over to the earlier record of the student like the GPA in earlier semesters, high school records and credits scored by the student, to give suggestions to upcoming students
about the selection of courses in the new semester. Both the universities are using data, one is using the data from the university record, and another is using the student records but what if we mix both of them together to make
the result better?
This theory is very simple to understand with these kinds of a simple example. If we go deeper, we can find advanced ways to improve grading system, teaching patterns, new curriculum design on the basis of earlier results,
feedback over new teaching model as well as about the old ones and also how new models are better than the traditional model(s).
Okay, let’s come to the problem I have faced earlier. I interacted with people from different sectors like university professors, startup founders, businessmen, students, etc. I think about many solutions which I will list
in the next article, but one solution that I found can be useful nowadays is a platform which would encourage people to contribute through what they have knowledge about and using recommendation models to find the highly-rated
academic resources, and eliminate the bunch of ‘not very useful’ or low-rated resources.
If we think a little deeper, we may find some better solutions.
Kirtivardhan Singh 21st July 2020, 20:00 IST
(Plagiarism in the article is subject to only coincidence. If you find any issue with the article, please email on 00129802019@vips.edu or ikirtivardhansingh@gmail.com.)