arXiv subjects

Susumu Hashimoto

Publications and source records attributed to Susumu Hashimoto.

4 recordsLinked to original sources

arXiv2026

Comparative visualization of knowledge structures using edge-difference graphs and network flow analysis: A case study of Wikipedia philosopher networks

Visualizing knowledge structures as graphs is common, but making them intuitively understandable remains challenging. Existing methods, such as macroscopic statistical metrics and whole-graph visualizations, often fail to capture local differences in conceptual relationships and suffer from severe visual clutter as networks grow large. To address these limitations, we propose a comparative visualization method that combines Edge-Difference Graphs with network flow analysis. The method first constructs Edge-Difference Graphs by extracting edges unique to each graph from graphs sharing a common node set, reducing redundancy while preserving the overall graph structure. It then identifies diverse paths between specific nodes by solving a minimum-cost maximum-flow problem. By incorporating a cost based on Adamic-Adar similarity, it penalizes routes that pass through generic hub concepts, enabling the extraction of contextually specific paths. We applied the method to networks of 20th-century French philosophers constructed from the French, German, English and Japanese editions of Wikipedia. The results reveal distinctive relational paths that reflect how each linguistic community receives and contextualizes these philosophers. This study provides a framework for the comparative analysis of large-scale knowledge structures and deepens our understanding of cultural and structural differences.

cs.SI↗

arXiv2025

Sample-Cluster-Select: A new framework to obtain diverse approximate solutions of combinatorial optimization problems

When solving real-world problems, practitioners often hesitate to implement solutions obtained from mathematical models, especially for important decisions. This hesitation stems from practitioners' lack of trust in optimization models and computational results. To address these challenges, we propose Sample-Cluster-Select (SCS) for solving practical combinatorial optimization problems under the assumption of potentially acceptable solution set. SCS first samples the potential solutions, performs clustering on these solutions, and selects a representative solution for each cluster. SCS aims to build trust by helping users understand the solution space through multiple representative solutions, while simultaneously identifying promising alternatives around these solutions. We conducted experiments on randomly generated instances, comparing SCS against multi-start local search and $k$-best algorithms where efficient methods exist, or evolutionary algorithms otherwise. The results show that SCS outperforms multi-start local search and $k$-best algorithms in most cases, while showing competitive performance against evolutionary algorithms, though not surpassing some of their variants. Most importantly, we found that the clustering approach provides insights into solutions that are difficult to obtain with existing methods, such as local structures of similar potential solutions and neighboring solutions of representative solutions. Thus, our approach helps practitioners understand the solution space better, thereby increasing their confidence in implementing the proposed solutions.

math.OC↗

arXiv2024

Spin orbit torque-driven motion of quasi-Bloch domain wall in perpendicularly magnetized W/CoFeB/MgO structure

The motion of chiral magnetic domain walls (DWs) driven by spin-orbit torque (SOT) has been extensively studied in heavy metal/ferromagnet heterostructures with perpendicular magnetic anisotropy. This study specifically focuses on SOT-driven DWs in near Bloch-states, which we refer to as ``quasi-Bloch DWs". These quasi-Bloch DWs exhibit slower motion compared to Neel-type DWs, offering potential for achieving highly controllable DW positions. Here, we investigate the characteristics of SOT-driven motion of quasi-Bloch DWs in perpendicularly magnetized ultra-thin films consisting of W/CoFeB/MgO. For analyzing the DW motion, we employ a one-dimensional model incorporating parameters derived from experimental data obtained from our samples. Our model successfully reproduces the experimental results, which reveal variations in the direction and threshold current density of DW motion among different samples. Through theoretical analysis, we unveil that the DW remains in quasi-Bloch states during motion, with SOT serving as the primary driving force rather than spin transfer torque (STT). The direction of motion is determined not only by the sign combination of Dzyaloshinskii-Moriya interaction (DMI) and spin Hall angle but also by the strength of DMI, STT, and extrinsic DW pinning. Furthermore, we provide analytical expressions for the threshold current density required for SOT-driven quasi-Bloch DW motion. These findings provide valuable insights for the design of future DW devices with specific film structures.

cond-mat.mes-hall↗

arXiv2021

A 3-approximation list scheduling algorithm for a single-machine scheduling problem with a non-renewable resource and total weighted completion time criterion

This paper studies a single-machine scheduling problem with a non-renewable resource (NR-SSP) and total weighted completion time criterion. The non-renewable resource is consumed when the machine starts processing a job. We consider the case where each job's weight in the objective function is proportional to its resource consumption amount. The problem is known to be NP-hard in this case. We propose a 3-approximation list scheduling algorithm for this problem. Besides, we show that the approximation ratio 3 is tight for the algorithm.

cs.DS↗