Two hundred and forty Texas Computer Science students recently participated in a data mining and analytics competition hosted by SparkCognition, an Austin-based enterprise artificial intelligence company with software solutions that help clients analyze complex data, reveal actionable insights, and identify and automate optimal responses. The company awarded scholarship prizes to the first, second, and third place winners of $10,000, $3,000, and $2,000 respectively.
This competition was included in the curriculum for “Introduction to Data Mining,” taught by Angela Beasley, Assistant Professor of Instruction at Texas Computer Science. In groups of four, students enrolled in the course teamed up to build data mining applications designed to tackle issues faced in everyday life. Students participating in the competitions used SparkCognition’s model building platform Darwin, a machine learning (ML) product that accelerates data science by automating the building and deployment of models.
“It is a truly unique opportunity for students and the company alike,” said Beasley.
By participating in this competition, students will gain real-world experience in applying these techniques and learn the power of using machine learning to make better-supported decisions.
“The best ways to learn data mining is to use your own data set in an experimental way. I can tell you what to do, but the best way is not coached through,” said Beasley.
Each team presented their project to a panel of SparkCognition executives, who graded on the impactfulness and creativity of the team’s application. The first-place winner of the competition built an application around Darwin with the intent to solve the increasing issue of media bias. The application was able to take any website or article and assign it a sensationalism and media bias score to determine what was written with the intention to excite readers versus to objectively inform the reader and share factual accounts.The first runner-up developed an application that was able to take data points from facial scans and images to identify and predict the mood of an individual. This was followed by the third place winners, who used data to determine what drives the first impression in a relationship and use the data to predict whether or not people would match on a dating service.
In addition to introducing the competition to other universities, Bruce Porter, who serves as both a Texas Computer Science professor and Chief Science Officer for SparkCognition, plans on continuing the program with future UT students. Porter has also stated that he hopes this event is the start of a lasting, fruitful relationship between SparkCognition and Texas Computer Science.
“I hope that (the competition) will be an incentive for students to think and challenge themselves on really serious, hard problems and then to be rewarded for their hard work. These kind of projects where the students are able to find their own project, something they personally create. And to pursue that with vigor — that's what these kind of competitions can provide,” said Porter.