ExaWizards Inc., a company that develops AI-enabled services to solve social issues (Headquarters: Minato-ku, Tokyo; Representative Director & President: Ko Ishiyama; hereafter, “ExaWizards”) introduced a system to quicky recommend the information that meets the user’s conditions to the job information services of dip Corporation, a leading job information service provider (Headquarters: Minato-ku, Tokyo; President & CEO: Hideki Tomita; hereafter, “dip”).
We confirmed that this system could quickly recommend highly accurate information based on the conditions set by the user signing up with dip’s job information services. ExaWizards is planning to offer various recommendation services to businesses in different industries on this occasion.
☑Background to the development of this system—Low-load machine learning and job matching
A challenge job information services present is to find the best way to pick the information which matches the user signing up with the services out of the huge database where data is seamlessly accumulated on a daily basis.
AI-enabled recommendation systems are becoming common; however, there are such challenges as updating the system with new data added day after day, and quickly showing results.
The recommendation system ExaWizards developed this time can quickly complete the learning phase necessary for using AI and the inference phase for job matching with a small amount of computer resources.
Specifically speaking, first, it is capable of learning and job matching in a short time regardless of the size of data on jobseekers and job offers; second, it allows the user to set conditions for job matching.
On top of that, with user profile information being limited, it can show highly-accurate matching results. The user can use the service just by entering their minimum profile information, which makes them feel less burdened when registering with the system.
☑ Features of AI algorithm—Development of a new collaborative filtering algorithm
This system is a “collaborative filtering” recommendation system in which the data on a number of users selected based on their similarity to the active user is used to make an inference (recommendation) for the active user. ExaWizards developed a new AI-based collaborative filtering algorithm this time.
High-speed learning and response, which were impossible with existing algorithms, were realized with this new algorithm. The updated information can also be reflected in matching results within a short period.
☑ Effects brought about by the introduction of this algorithm—High-speed processing and effective outcomes expected
This algorithm is used for companies to send job offer emails to job seekers, and indispensable for the system to show job and candidate data close to the requirements of both sides.
The Proof of Concept (PoC) conducted when the system was introduced to some customers on a trial basis showed that the system could process large datasets at a high speed with this new algorithm although there were no systems that could do it with other collaborative filtering algorithms. According to the PoC, the number of job offer emails read by jobseekers and the number of applicants both increased more than twice as much as before the introduction of the new algorithm.
[ExaWizards Corporate Profile]
Company name: ExaWizards Inc.
Location: 21F, Shiodome Sumitomo Bldg., 1-9-2, Higashi Shimbashi, Minato-ku, Tokyo
Representative: Ko Ishiyama, Representative Director & President
Description of business: Industrial innovation and resolution of social issues through the development of AI-enabled services
<Inquiries about news releases>
Public Relations Department, ExaWizards Inc.