Can AI help solve education’s big data problems?

Insights offered by data can help create solutions to business problems. But managing big data can overwhelm rather than help.

“There’s so much data that’s lying around in everything that we are doing, and now the amount of data that is getting generated is so much. But how do you get meaningful insights out of this data?” Kavitha Prasad, Intel’s vice president and general manager of Datacenter, AI, and cloud execution and strategy, told ZDNet. s

“It’s beyond humanly possible right now to sit down with the data and figure out not just what the data is saying, but the interrelationship between the different sets of data that is getting collected, and to figure out what business insights are hidden behind this data.”

This issue doesn’t just impact tech companies. The education sector also faces challenges in managing, protecting, and extracting value from big data.

How can schools deploy AI to improve student outcomes?

Artificial intelligence can make it possible to solve big data problems. 

For example, Intel collaborated with Aible to use AI to help Nova Southeastern University. Aible can sense, explore, and optimize data through AI. 

Prasad said the university wanted data-based insight on two primary issues: Improving undergraduate student retention and optimizing student welfare. Intel and Aible, in collaboration with Dell, deployed AI on the school’s data sets. This project identified ways to potentially lower student attrition by 17% in 15 days. 

 Don Rudawsky, Nova Southeastern’s vice president of institutional effectiveness, wrote in an Aible case study highlighting the project: 

During a one-hour meeting, we went from a raw dataset to exploring insights in the data automatically highlighted by Aible, to creating and even deploying a predictive model. The collaboration with academic and financial aid advisors helped us further optimize the models and made them more useful — but we went end-to-end from raw data to deployed model in such a short amount of time.

Achieving a business outcome in 15 days is “huge,” Prasad added. 

The companies and the university achieved these results by running about 500 AI models across six projects. Each project took about 25 minutes to process. Using Xeon Scalable processors in tandem with Aible’s technology, developers can run applications without managing servers, Intel said.

Intel and Aible first began working together through Intel’s Disruptor Initiative. The program lets startups and established companies access Intel solutions and technologies. 

The partnership gives “that Intel muscle to these startup companies to make sure that their technologies can be leveraged broadly across the industries,” Prasad said.

Aible has also worked with customers in K-12 who wanted to get insights on student retention.

“Aible helped us proactively identify students who were leaving for reasons where we could address to retain them,” said Hywel Benbow, senior vice president of global data and analytics at GEMS Education. 

“When it comes to K-12, minimizing churn by itself is interesting, but Aible helped us distinguish where and how we could act to reduce churn and directly map the economic impact of taking those recommended actions.”

SEE: What is an AI engineer?

Executive: AI will be in ‘all realms of human life’

As AI expands, so do questions around using the technology ethically.

“AI is as good as the data is, and AI gives predictions,” Prasad said. “It’s still predictive.” 

And it’s up to humans to conduct the risk-benefit analysis of inaccurate predictions. 

“We acknowledge that there are a lot of implications associated with these AI technologies, and we are focused to make sure that it is used for social goodness,” she said. 

Prasad added that Intel follows “a continuous, rigorous process of review” for projects using AI. 

Prasad also predicts that AI soon will “impact every facet of human life.”

“The proliferation of AI is only expanding, and it’s very interesting to see where the businesses are trying to use techniques like AI to help achieve better business outcomes — be it in the context of education for student retention or figuring out what programs make sense from a student perspective.

“We are only at the tip of the iceberg,” she continued. “It’s only getting started, but eventually, it is going to expand a lot more into all realms of human life.”