Sanpou Co., Ltd. (Headquarters: Shibuya-ku, Tokyo; Representative Director and President: O-yuh Watanabe), which operates career information services for high school students, announced the implementation and release of "AI Matching System MichiCal" on its university advancement portal site "Sanpou Shingaku Net" (https://www.sanpou-s.net/). This new feature uses AI to personalize support for school selection.

Background of Development: The Era of University Entrance Exams Shifts from "Deviation Scores" to "Matching"

Currently, about 60% of private university entrants advance through school recommendation or comprehensive selection (year-end selection). With personal statements and interviews being emphasized, examinees are required to have a clear sense of purpose: "What do I want to learn at that school?" School choices based on elimination, such as "because it's a famous school" or "because my deviation score matches," can lead to a decline in motivation after enrollment or early withdrawal. Therefore, an essential matching that deeply connects students' subjective "interests and concerns" with the school's "educational content" has become indispensable.

Furthermore, vocational schools, professional colleges, and vocational training centers also offer diverse vocational education in response to changes in societal needs. While career options have expanded dramatically, finding a school that truly matches one's aptitude and future vision from among the subdivided departments and curricula has become a difficult challenge that requires even more advanced information gathering and judgment.

However, many high school students in the field say, "I don't even know what I'm interested in," and it is extremely difficult for busy teachers to secure interview time to delve into each student's interests. Finding the optimal school from a vast amount of school and department information has become an extremely heavy burden for high school students, parents, and busy high school teachers.

"MichiCal" was born to solve these problems and provide an opportunity to "Meet Career" (encounter an unknown career).

Five Strengths of "MichiCal"

1. Facilitative Dialogue: Providing "Awareness" and "Serendipity" to Students Undecided About Their Future Paths

Even if a student is "undecided about what they want to do," the AI asks multi-faceted questions based on their "likes" and "strengths." For example, from a vague interest like "I like animals," it presents specialized fields that students might not have discovered on their own, such as not only veterinarians and pet groomers but also "animal welfare support." It can reveal professions, academic fields, and the attractions of universities, junior colleges, and vocational schools nationwide that students themselves might not have known in a dialogue. It provides serendipity (unexpected discoveries) that help students realize their own potential, guiding ambiguous hopes towards concrete career path candidates.

2. Overwhelming Data Volume: Learning 12 Million Characters of Proprietary Research Information

It covers information on approximately 3,500 schools and 13,600 departments, from national universities, junior colleges, and vocational schools to public vocational training facilities. Based on our company's unique surveys and proprietary research, the AI has learned the "real appeal of schools," totaling approximately 12 million characters, including class content, research themes, internal and external practical training, qualifications obtained, and even post-graduation career paths.

3. Safety Design Compliant with MEXT Guidelines

We have designed a safe system compliant with the "Guidelines for the Utilization of Generative AI in Elementary and Secondary Education" formulated by the Ministry of Education, Culture, Sports, Science and Technology (MEXT). By using RAG (Retrieval-Augmented Generation) technology, which limits the basis of responses to specific databases, we minimize inappropriate answers and AI-specific "plausible lies (hallucinations: misinformation generation)." We will ensure reliability that can be safely used even in official guidance settings such as triple interviews.

4. Adoption of "Zero Data Learning" and Thorough Personal Information Protection

Dialogue content entered by students is processed in a "Zero Data Retention" environment, meaning it is not used for AI (LLM) re-learning. This service adopts an API infrastructure through a development vendor, with a contract concluded with OpenAI to not use usage data for learning. This provides a secure AI consultation environment where students' personal consultation content is protected with high confidentiality.

5. Supporting Career Guidance Sites: Reducing Teacher Burden and Alleviating Parental Anxiety

By utilizing "MichiCal" as an introductory step for selecting open campus participants, career aspirations surveys, and interview guidance, students' self-understanding is deepened, and the interview workload of high school teachers is reduced. Furthermore, it functions as a compass for modern career choices for parents' generations who are confused by the drastic changes in the university entrance examination environment.

Service Overview

Name: AI Matching System MichiCal for School Selection

Implementation on: Sanpou Shingaku Net (https://www.sanpou-s.net/)

Target: High school students, parents, high school teachers, and those wishing to advance to universities, vocational schools, etc.

Main functions: What you want to learn, your future aspirations, etc.

FACT BOX

  • Source: PR TIMES
  • Category: New Product