Keita Takeuchi, Masashi Iwasaki, Masato Shinjo

Similarity-Based Network in the Industrial Community of Joyo City

Introduction

Similarity-based network in the industrial community of joyo city. Explore Joyo City's industrial structure via similarity-based network analysis of company interviews. Identify similar companies & new classifications across industrial categories.

Abstract

Data utilization is becoming increasingly widespread in a variety of fields around the world, and has become especially important in the industrial world. Data utilization techniques and approaches can contribute to the development of not only individual companies but also certain groups of companies. In this paper, we consider the industrial structure of Joyo City, Japan, by analyzing data collected through interviews with company presidents and managers. The main purpose of this paper is to grasp it in terms of similarity across industrial categories. We first express the features of each company as a vector with entries determined from the interview data. We then compute vector similarities in order to draw a graphical network, in which nodes corresponding to similar companies are linked by an edge. From the resulting network, we derive the most similar companies in the same and different industrial categories for each company. Moreover, we then classify Joyo City's companies into new groups across the standard categories.

Review

The paper, "Similarity-Based Network in the Industrial Community of Joyo City," addresses a pertinent and timely topic concerning the application of data utilization techniques to understand and foster development within industrial communities. Recognizing the increasing importance of data-driven insights in the industrial sector, the authors propose an innovative approach to analyze the industrial structure of Joyo City, Japan. The core concept revolves around moving beyond conventional industrial categories to identify latent similarities between companies, aiming to contribute to the development of both individual firms and the broader industrial group. Methodologically, the paper outlines a clear and intriguing process. It involves collecting qualitative data through interviews with company presidents and managers, which is then systematically transformed into quantitative vector representations for each company. These vectors subsequently enable the computation of similarities, forming the basis for constructing a graphical network where nodes (companies) are linked by edges indicating similarity. A key contribution highlighted is the ability to derive the most similar companies, both within and across existing industrial categories, ultimately leading to the classification of Joyo City's companies into novel groups. This data-driven approach promises a granular understanding of inter-company relationships and potential synergies that might otherwise remain unobserved. While the abstract presents a compelling framework, the full paper would benefit from detailed elaboration on several key aspects. The specifics of how interview data is quantified into vector entries, the choice of similarity metric, and the methodology for constructing the network will be critical for evaluating the robustness and validity of the findings. Additionally, more information regarding the scope and scale of "Joyo City's industrial community" and the representativeness of the interview sample would enhance the paper's impact. Nevertheless, the proposed methodology offers significant potential for regional economic planners and policymakers, providing a valuable tool for visualizing industrial ecosystems and identifying new avenues for collaboration and strategic development.

Full Text

Comments

You need to be logged in to post a comment.