Kyoung Jun Lee., Baek Jeong., Yujeong Hwangbo., Youngchan Kim., Sungwon Bae., & Taehoon Baek. (2022). AMPER(Aim-Measure-Predict-Evaluate-Recommend): The Paradigm of Digital Me. 23rd International Center for Electronic Commerce.
Abstract
AI services can help people live healthier and richer lives, and such services can be
collectively referred to as Digital Me. The purpose of this paper is to research Digital
Me service and suggests AMPER (Aim-Measure-Predict-Evaluate-Recommend)
approach to implement a general Digital Me algorithm for providing actual services.
Digital Me is defined as an AI-based product service system (PSS) that makes it
possible to manage the individual’s state (health, beauty, memory, knowledge, finance,
and happiness, etc.) in real- time. We studied cases of Digital Me service of edutech
and healthcare. In order to enable a Digital Me service, it is necessary to measure,
predict and evaluate the user’s future states, and recommend actions to improve the
states. We developed a structure of a purely data-based algorithm that set a usercentric
aim (A), measures the user’s states (M), predicts the user’s future states (P),
evaluates and compares the user’s performance (E), and recommends desirable action
(R). As a result of experimenting with the proposed algorithm structure with EdNet
dataset, we verified the AMPER structure for the user-centric aim of improving the
user’s English score that measures the correct answer to the question solved by the
user, predicts the correct answer to the next question using Transformer model,
evaluates the user’s English ability, and recommends the question that will improve
English scores the fastest.