The Science

AI, ML, DS are big words in the software industry now. In the technology sector, you do get big words regularly, and a good part of them eventually die out; more like a fad. But not these three. They have earned their places and are here to stay.

Artificial Intelligence, Machine Learning, and Data Science are not new really. Depending on how you look at it, they have been around for decades, and centuries in the case of data science. Old civilizations and empires collected data about their operations, economies, and residents, and looked into it for information, trends, and convergences. It was always a laborious manual undertaking until it got automated recently. Catalyzed by progressive hardware and software technologies, AI, ML, and DS are nowadays driving a sharp exponential rise in the output of the many fields that implement them. The months of drilling manually, sorting through data, finding some patterns in some behavior, have now become minutes. Moreover, machines started to learn from historical information, or from training data sets, to improve their “knowledge” and forecast.

This said, and for all their worth, these three fields are in their promising beginning, with lots of telltale signs.

Our Work

At White Mountain Technologies, we moved data science to the center. Data has always been a focal point in our work. Our UniversiTools and Skoolee information management software for the educational institutions are being used by tens of thousands of users, with repeating cycles of communication, processes, and information exchange. The data that these two systems preserve is clean, categorical, and theme-oriented. It is classifiable because of the very nature of the business of technology in education. Consequently, with the software’s underlying data structures, it is then just common sense that we dig deeper, and intelligently, into this data and see what we can find. And we found a lot.

In our data science work, we go beyond vanilla reports, statistics, and graphs. We dig out pedagogical, social, and behavioral trends. We work on understanding the natural orientation of students, where they find themselves, and how they can deliver better results. We also study how teachers perform, where some do better than others, and how to raise the performance of all of them. And we identify actions that the school or college might take to reach potentials it did not tackle before.

Of course, educational institutions lend a great helping hand in this matter. Responding to targeted data collection effort is one. Running some specific software algorithms that we put in effect is another. Data on grades, academic achievements, attendance, participation, motivation, and performance over the years is readily available. The data that some schools might not have, and which we need to supplement the existing data set, is that pertaining to the psychological, social, and creative domains. We need it to dig deeper into behavioral and thought processes. To collect such data, if it is not already available, we work with the educational institutions to go beyond the regular academic undertakings. It is imperative that the more stakeholders involved in the process the better. When the educational institution and its instructors, teachers, parents, and students team up, the results can be more informative, indicative, and effective.

To Do

As a future extension to our work, we also look to evaluate what the educational sector produces versus what the job market actually needs. There is a gap, and companies recruiting fresh graduates put effort and time in coaching them. By the time they become productive, a sizable investment has been spent. Identifying the specifics of the gap is a prerequisite to narrowing it down. Our concern here is not the material and courses that are being taught. Rather, we want to sort through the data to extract the thought, creativity, and motivation processes that are in effect, and to see how they can be improved.

The work is challenging and rewarding. The education data is there, bustling with a wealth of information. The science that looks through this data and generates results is in our hand and growing remarkably. When have put things together and worked on it diligently. The results are impressive.