Financial Systems

1. Applying AI to Identifying Suspicious Transactions in Securities Firms and Exchanges

The securities industry is making growing use of artificial intelligence (AI) for tasks such as the automated trading of stocks and bonds, decisions about opening accounts, or giving credit for loans.

As part of this trend, Hitachi AI Technology/H (Hereafter referred to as AT/H) is being deployed in the reviewing of stock trades, a job that in the past required a specialist, where it is being used in place of such specialists to screen trades for possible suspicious activity. Past trading data, the results of reviewing trades, and market data, all of which were pre-processed based on the knowledge of the specialists and were supplied to AT/H to generate a model that quantifies risk assessments that are close to what the specialists had done. The aim is to improve efficiency by feeding transaction data into this model so that it can produce risk assessments similar to those determined by the specialists and reduce the number of trades that ultimately need to be checked by a person(specialist). As of September 2018, this solution had been adopted by a number of companies.

Along with enhancing the AT/H model by expanding its scope of application to cover a larger number of types of suspicious transactions, future plans also include expanding its scope of application to cover other financial products than stocks as well as all other types of commercial transaction activity.

1. Flowchart of using AT/H to check suspicious activity

2. Interactive Digital Services for Financial Institutions

3. Deposit Slips-adapted ATM Using Recognition Technology that Works Against Undefined Forms

A deposit slips-adapted automated teller machine (ATM) enables the processing of various different types of deposit slips, such as those for tax payments, by using a recognition technology that works against undefined forms. This means it is able to handle tens of thousands of different tax payment types and other deposit slips, which could only be dealt with by a teller.

With this ATM, users now can process their own payment transactions, which can reduce congestion at the teller counter, and users no longer need to fill out payment request forms, which results in bringing greater convenience and better service to users while helping tellers make work more efficient. A key feature is the use of a form recognition technology, which Hitachi developed, to identify and acquire necessary information from an image of the form captured by a scanner. Due to this technology, it is able to acquire the necessary information from forms without having to either predefine regions to scan or specify their formats in advance, so the ATM is able to process a much wider range of payment slips than in the past.

Hitachi will continue to respond flexibly to the increasing and diverse requirements of financial institutions to support their objectives, including service quality improvement and efficiency gains.

3. System diagram