In recent years, hot-desking that allow workers to choose their own workspace have been attracting attention to accommodate the diversified work styles of workers and to improve productivity.
However, when workers are free to choose their own seats, problems arise such as fixed seating due to biased preferences. There is also a preference externality where workers prefer to be seated closer to their coworkers. However, there is no model that considers the positional relationship between workers and their coworkers when workers choose their seats.
In contrast, companies expect an increase in incidental informal communication. However, if workers who are always closely related to each other are assigned to seats close to each other, accidental informal communication is unlikely to occur. Models exist that aim to increase informal communication, but excessive decentralization has occurred, and communication has become more difficult. Therefore, what is needed, is an algorithm that considers not only workers' preferences for seating, but also their relationships with their coworkers, so that seating assignments can be made in a way that encourages incidental informal communication.
In this study, worker externality is represented by a utility function that expresses the strength of the relationship between workers in terms of a network. By doing so, we propose an algorithm to search for two-sided many-to-one stable matching that takes externality into account. The strength of the relationship in terms of externality is calculated from the community structure in the network.