FRONTEO succeeds in improving an AI algorithm that captures the subtleties of human behavior

2022.08.24 Press release

--To the press -

FRONTEO succeeds in improving an AI algorithm that captures the subtleties of human behavior

Up to 80% reduction in the number of review documents required to find 44% of relevant documents in digital forensics

Masahiro Morimoto, President and CEO
2-12-23 Konan, Minato-ku, Tokyo
(Code number: 2158 TSE Growth)

 FRONTEO Co., Ltd. (Headquarters: Minato-ku, Tokyo, President: Masahiro Morimoto, hereinafter FRONTEO) is one of the core technologies that make up the AI ​​engine "KIBIT" developed in-house.1, classification performance that reduces the number of evidence-related documents and unrelated review documents in digital forensics (information security and analysis investigations targeting information recorded on digital devices) by up to 44% compared to the past by improving the algorithm. has been successfully improved.


 The amount of data managed by companies is increasing year by year, and digital forensics collects several TB of data per evidence holder (custodian) and extracts limited data from a huge amount of documents. Documents related to evidence must be found within the specified period.For attorneys involved in investigations, time, cost and investigation quality are major challenges.Among them, document review in the process of discovering relevant information is said to account for about 1% of the time and cost, making the use of AI essential. KIBIT is already being used in the legal tech field in the United States and Japan, and has helped solve problems by significantly reducing the amount of documents related to reviews and the associated time and costs.


 In Landscapescaping, which is one of the core technologies that make up KIBIT, this improvement is for words that are highly likely to indicate the subtleties of human behavior, among the highly rare words that do not appear frequently in the data. We have developed a unique technology that formulates the degree of relevance to documentary evidence in a more precise manner based on statistics.For example, inverse document frequency (IDF)* is commonly used for rare words2In addition to this, we used a method that considers statistical errors caused by the number of occurrences of words and a method that calculates the degree of relevance of evidence from the obtained rarity.The result is up to a 80% reduction in the number of human-reviewed documents required to find 44% of the evidence-relevant documents, and an improved recall rate*3 improved. (See Figure XNUMX, using FRONTEO test data) 


 As a pioneer of digital forensics and discovery in Japan (discovery procedures in the US civil litigation system), FRONTEO will continue to develop and improve AI technology that will help improve the efficiency of fraud investigations and litigation support.


Figure XNUMX. Comparison before (orange) and after (green) improvement.
Compared to the review without AI (random sampling, black dotted line),
It showed high classification performance even before the improvement, but after the improvement, further improvement in classification performance has been achieved.


*1 Landscaping: Based on the amount of mutual information, FROTNEO's unique core technology with excellent intuition and explainability that can quantify the degree of relevance of each word in evidence documents.It is designed for analysis of data with a low proportion of relevant documents as evidence, and we expect to apply it to more digital forensic cases in the future.
*2 Inverse document frequency (IDF): A measure of how rare a word appears in a few documents in the data.
* 3 Recall Rate: Recall rate.Percentage of all data that is relevant as evidence that is correctly predicted to be relevant.



■ About KIBIT URL:
"KIBIT" is an artificial intelligence that analyzes text without relying on keywords, using a unique machine learning algorithm that reproduces the "tacit knowledge" possessed by specialists and business experts.Highly accurate analysis in a short time is possible with a small amount of teacher data.


FRONTEO uses the in-house developed AI engines "KIBIT", "Concept Encoder (trademark: conceptencoder, reading: concept encoder)", and "Looca Cross", which are specialized in natural language processing. It is a data analysis company that supports the business of companies by extracting meaningful and important information from a huge amount of text data. Since its establishment in August 2003, it has been expanding globally to Japan, the United States, South Korea, and Taiwan, focusing on legal tech businesses such as "e-discovery (electronic discovery)" and "digital forensic investigation" that support corporate international litigation. Has been deployed.Based on the AI ​​technology cultivated in this business, we will expand the business field to the life science field, business intelligence field, and economic security from 8, and by using AI to "turn text data into knowledge", We contribute to solving various corporate issues such as drug discovery support, dementia diagnosis support, financial, personnel, and sales support. Listed on TSE Mothers (currently TSE Growth) on June 2014, 2007. Obtained a first-class medical device manufacturing and sales business license in January 6 (permit number: 26B2021X1), and notified the managed medical device sales business in September of the same year (notification number: 13 Minato Misei Equipment No. 1).The capital is 10350 thousand yen (as of March 9, 3).


* FRONTEO, KIBIT, conceptencoder, and Looca Cross are registered trademarks of FRONTEO in Japan.



<Contact for inquiries from the press>

Public Relations Officer, FRONTEO Inc.



ja 日本語
Machine Translation by Google. : close x