Mobigram (CEO Jeon Sang-hoon), which is a blockchain-based AI meter development and big data company, has been working in collaboration with Professor Song Joo-young (University of Pennsylvania) and Professor Song Tae-min (Sahmyook University) to develop Artificial Intelligence (AI) using social big data to predict diet successes and failures, and plans to launch a smartphone application in the second half of this year that can provide Diet prediction AI services.
The diet prediction AI developed by Mobigramuses text mining and opinion mining on the contents of online channels (Tweeter, Blog, café, news bulleting, etc), and predicts the success rate of diet by modeling the learning data after the machine learning about the diet method (exercise therapy, diet) and the diet success/failure.
“The Diet Prediction AI service presents diet related questions such as lifestyle, exercise, etc. through the smartphone app to users who are currently on a diet or who are planning to do a diet, and when users enter an answer, the app shows the user’s diet success/failure probability. In addition to the probability of success of the diet, the user will also be able to access an advisory function, an obesity calculator, and an exercise method function to increase the success rate of the diet plan,” said an official at Mobigram.
Mobigram is also planning to develop ‘Machine Learning-based youth crime prediction AI’ along with Song Joo-young and Song Tae-min. The youth crime prediction AI is a predictive program designed to prevent juvenile offenders from committing crimes, and it is already a model that can predict the re-offending of juvenile offenders based on machine learning using personal environment survey data, including past experiences, social and individual factors.
“In the future, we will continue to deveop Ai to provide customized AI-based services to all sectors of society, such as finance, crime, and health care,” said Jeon Sang-hoon, CEO of Mobigram.