Quadratic Unconstrained Bin Matching System Reduces Parental Stress, Fosters Intergenerational Connections in Japan

The increasing prevalence of isolated parenting in Japan presents a significant social challenge, contributing to parental stress and potentially hindering child development, and researchers are now exploring innovative solutions to foster stronger community support networks. Yuuma Matsumoto from the Institute of Science Tokyo, alongside Taisei Takabayashi and Rima Sato from Tohoku University, and their colleagues, demonstrate a novel childcare support service that connects parents with experienced senior community members, aiming to provide crucial psychological support and intergenerational connections. This team tackles the complex problem of optimal matching, considering both individual compatibility and practical constraints like supporter availability, by employing the power of quantum annealing. Their work shows that this quantum approach generates significantly better and more diverse matching options compared to traditional methods, and a successful pilot program in Sendai City confirms the framework’s potential for real-world implementation and flexible scheduling, offering a promising avenue for addressing the growing needs of modern families.

Quantum Matching Optimised with Annealing

This research explores using quantum annealing, a technique employing D-Wave systems, to solve a matching problem within a community support network. The goal is to optimally connect supporters (volunteers) with visitors (those needing assistance) based on their preferences and needs, comparing the performance of quantum annealing against traditional simulated annealing. The study utilized real-world data collected from a community support system in Sendai City, Japan. The problem involves formulating the assignment of supporters to visitors as a mathematical problem, maximizing overall compatibility.

Quantum annealing demonstrated promising results in finding good solutions, potentially outperforming simulated annealing by generating a wider range of potential matches and identifying better pairings. This research demonstrates the potential of quantum annealing for optimizing resource allocation in community support systems. The problem was successfully translated into a mathematical format suitable for quantum annealing, facilitated by the OpenJij framework and D-Wave quantum annealers. A greedy algorithm generated initial solutions, providing a starting point for the optimization process. This research provides a practical application of quantum annealing to a real-world optimization problem, highlighting its potential for improving resource allocation and matching in community support systems.

Matching Parents and Seniors via Quantum Optimisation

Researchers developed a service connecting parents with experienced senior community members, aiming to provide psychological support and intergenerational exchange. A key challenge was efficiently matching individuals, considering compatibility, supporter workload, and scheduling constraints, which the team formulated as a complex mathematical optimization problem using Quadratic Unconstrained Binary Optimization (QUBO). They rigorously evaluated its performance against simulated annealing. The QUBO formulation allows the team to represent the matching problem in a way suitable for Quantum Annealing (QA), an algorithm that leverages quantum mechanics to explore a vast solution space.

QA searches for the lowest energy state of a system, effectively identifying optimal matches between parents and supporters, potentially overcoming limitations of classical solvers by utilizing quantum tunneling. To test this, they designed experiments employing both QA and simulated annealing to solve randomly generated mathematical problems, demonstrating QA’s ability to sample a more diverse set of high-quality matches. The methodology involved a detailed questionnaire with 10 items, grouped into categories encompassing cognitive tendencies, daily life values, communication styles, and learning interests, forming the basis for quantifying compatibility between users and supporters. Each questionnaire item yielded a score quantifying compatibility, ranging from 0 to 3 based on the degree of agreement in responses.

These individual item scores were summed to create a comprehensive compatibility score for each potential pairing. To further refine the matching process, practical constraints, such as time slot availability and supporter capacity, were incorporated, addressed through penalty terms and pre-filtering. This combined approach enabled the team to generate multiple high-quality matching candidates, validated through a proof-of-concept field experiment conducted in Sendai City, Japan.

Parent-Supporter Matching via Optimisation Algorithms

Researchers developed a matching framework to connect parents with experienced senior community members, aiming to provide psychological support and address isolation in parenting. The core of this system relies on quantifying compatibility between individuals using data gathered from questionnaires, ultimately formulated as a mathematical problem. Questionnaires comprised 10 items, categorized into cognitive tendencies, daily life values, communication style, and learning interests, providing a multifaceted profile for each participant. Compatibility scores were calculated for each user-supporter pair based on responses to these questionnaires, with item scores ranging from 0 to 3, reflecting the degree of agreement between individuals.

A perfect match on an item received a score of 3, while a difference of three points resulted in a score of 0. These individual item scores were summed to generate an overall compatibility score, representing the total affinity between a user and a supporter. To ensure realistic matching, the framework incorporated practical constraints, including limits on the number of users assigned to each supporter and the number of supporters assigned to each user. These constraints were addressed through penalty terms within the mathematical formulation, effectively discouraging solutions that violated the specified limits.

Quantum Annealing Improves Intergenerational Matching Diversity

This research demonstrates a compatibility-based matching framework designed to connect isolated parents with experienced senior community members, offering psychological support and fostering intergenerational exchange. The framework formulates the matching process as a mathematical problem and assesses the performance of Quantum Annealing (QA) against Simulated Annealing (SA). Results indicate that QA can improve the diversity of matches found.