Tomoaki Akitomi

Profiting from IoT: the key is very-large-scale happiness integration

Big data without link to value is merely a cost. We have studied how to profit from data with Internet-of-Things technologies for over 10 years to reach the answer: the Wearable Happiness Meter. It allows us to integrate the measure of both wellbeing and productivity of 7-billion people worldwide, which was the dream of the 18th-century philosopher Jeremy Bentham, numeration of the greatest happiness of the greatest number to measure the right and wrong. Knowing right and wrong with the 10x speed over conventional financial feedback accelerates the growth of the enterprise, the economy, and the individual to maximize the worldwide happiness. Here the integration is not only on a chip, but in the distributed massive chips embedded in the society.

A Method for Analyzing Influence of Emotions of Posts in SNS Conversations

We devised a method for analyzing communication, which focuses on influences between two sequential posts in a conversation on a social-network service (SNS). Posts expressed on an SNS are classified into three emotions: positive, negative, or neutral. Two influences were analyzed: the emotion of a persons preceding post on the emotion of a post from another person, and the emotion of a persons post on the emotion of a subsequent post from the same person. To analyze these influences, the basis of the frequency of emotional transitions of two sequential posts in actual conversational data is evaluated by probability and information entropy analysis. Three tendencies were found: first, a person who posted positively before is likely to post positively again, second, a neutral emotion of a persons post is more likely to be induced when both that person and another person expressed a neutral emotion in the past, third, negative posts do not have a strong influence on the emotion of subsequent posts of that person and other persons. These findings may be useful, for example, in promoting positive conversations.

Automatic Label Data Abstraction Based on Information Entropy (Application Paper)

There is currently a big demand for automating big data analysis. In the data analysis field, data abstraction or summarization playes an important role in the extraction of generalized information from large scale data. We developped an artificial intelligence computer system with the aim of automating big data analysis and came up with a method that can abstract numerical type data (age, height, time, etc.). However, it could not abstract or summarize label type data (customer ID, product code, name, etc.). In the present work, we have developed a label abstraction method based on information entropy. Experiments using open real data showed that the proposed method achieved an extraction accuracy of 80% evaluated by f measure. We intended to apply the proposed method to our artificial intelligence and perform further evaluations.