Researchers Database


FacultyFaculty Division of Humanities and Social Sciences Research Group of History,Sociology and Geography
Last Updated :2022/10/29


Profile and Settings

  • Name (Japanese)

  • Name (Kana)



  • Doctor of Science, Tokyo Metropolitan University

Research Interests

  • Urban Geography, Geographic Information Science, Quantitative Geography

Research Areas

  • Humanities & social sciences, Human geography
  • Humanities & social sciences, Geography

Research Experience

  • Apr. 2014, 奈良女子大学研究院人文科学系・教授
  • Apr. 2007, Mar. 2014, 奈良女子大学文学部・准教授
  • Apr. 2001, Mar. 2007, 奈良女子大学文学部・助教授
  • Oct. 1995, Mar. 2001, 東京都立大学理学部・助手


  • 1995, Tokyo Metropolitan University, Graduate School of Science, 地理学
  • 1989, Tokyo Metropolitan University, Faculty of Science, 地理学

Association Memberships

  • 経済地理学会
  • 人文地理学会
  • 日本地理学会
  • 地理情報システム学会
  • 日本都市地理学会
  • 日本社会学会


Published Papers

  • Refereed, Geographical Review Of Japan, 日本地理学会 ; 2009-, Generalization of Central Place Theory in Hierarchical Structures Using Mathematical Programming, ISHIZAKI Kenji, 2015, 88, 4, 305-326, 326
  • Refereed, 地理学評論, 日本地理学会 ; 2009-, 数理計画法による中心地理論の体系化―単一財の立地について―, ISHIZAKI Kenji, 2014, 87, 2, 87-107, 107
  • Refereed, 建築雑誌, コンビニエンス・ストアの品揃えと立地, ISHIZAKI Kenji, 2001, 1470, 22-25
  • Refereed, 多摩ニュータウン研究, 多摩ニュ-タウン学会, 地理情報システムで描く多摩ニュータウンの居住環境, ISHIZAKI Kenji, 1998, 1, 1-12, 12
  • Refereed, Geographical Review of Japan, Series B, Spatial Competition and Marketing Strategy of Fast Food Chains in Tokyo, Kenji Ishizaki, 1995, 68, 1, 86, 93, Scientific journal
  • Refereed, Geographical Review of Japan, The Association of Japanese Geographers, A reinterpretation of Christaller's central place theory : from marketing principle to generalized maximal covering location problem, ISHIZAKI Kenji, Previous research efforts (Ishizaki, 1992) demonstrate that the marketing principle of Christaller's central place theory can be formulated as the set-covering problem. Results from this model applied to a hypothetical lattice network bring up the question of how to generate the hexagonal system of cen-tral places. Christaller required a set of six K-places, located at the median points between their neigh-boring B-places, to cover unsupplied areas about k-d units from B-places (Fig. 1). However, only three K'-places will entirely serve unsupplied areas (Fig. 2). Why are K-places located at the median points between B-places ?
    The primary purpose of this paper is to propose a reinterpretation for solving the above problem and reformulate the marketing principle using a location-allocation model. Furthermore, the relaxed models reflecting real assumptions are developed to expand the notions of the marketing principle. Last, relationships between the marketing principle with the relaxed models in this paper and well-known location-allocation models are discussed, in order to investigate the implications of the market-ing principle.
    Christaller's central place theory insures a continuous type of good and successively inclusive hier-archy. Satisfying these constraints, the supply of goods from a central place is represented as a bundle of outer limits of the range of goods at some central place (Fig. 3). Consider adding the loca-tion of the central place with a bundle of the outer limits of the range of goods. The most efficient location of central place K is at the center point between two B-places on a line (Fig. 4), which where
    aim=demand of node i for good m
    Zim=1 if node i is covered with good m
    Yjm=1 if node j serves good m
    pm=the number of supply points for good m
    wm=weight for good m
    M=the set of good m
    dy=distance from node i to j
    Rm=the outer limit of the range of good m
    Objective function (1) is to maximize weighted demand for good m. Constraint (2) allows Zim to equal 1 only when one or more supply points of good m are established at sites in the set Ni(Rm). The number of supply points of good m is restricted to equal pm in constraint (3). Constraints (4) and (5) insure successively inclusive relaxation: the former defines which supply points for some good q possess good r; the latter defines which demand points covered by some good s are likely to be suppl- ied good t.
    Consequently, the marketing principle and the relaxed models are summarized in Table 2 and in the linked well-known location-allocation models (Fig. 7). The relaxed models are based upon the max- imal covering location problem (MCLP), which is useful for public facility location planning (especial- ly emergency medical services). GMMP has all the characteristics of maximal covering extensions. Therefore, GMMP expanded from the marketing principle is defined as the generalized maximal covering location problem integrating MCLP and its extensions. This definition may make the marketing principle interpretable as the basis of various contemporary public facility location models., 1995, 68A, 9, 579-602, 602
  • Refereed, The Human Geography, 人文地理学会 ; 京都, Location-allocation models and their development, ISHIZAKI Kenji, 1994, 46, 6, 604-627, 627
  • Refereed, Comprehensive urban studies, 東京都立大学都市研究センター, MDPREFによる都市公園の選好分析, ISHIZAKI Kenji, 本稿は、東京区部とその周辺に位置する親水性の8公園(井の頭恩賜公園、水元公園、有栖川宮記念公園、葛西臨海公園、上野恩賜公園、日比谷公園、浮間公園、和田堀公園)を対象とし、都市公園の選好構造を解明しようとした。175人の大学生・院生に対し、1公園につき1セット4枚の写真からなる合計8セットの写真を呈示し、8公園の選好順位データをえた。この175人 x 8公園の選好順位行列に対しMDPREF(選好データを分析するためのベクトル・モデル型ノンメトリックMDS)を適用し、2次元のジョイント空間を求めた。葛西臨海公園と上野恩賜公園が正の大きな値をとり、有栖川宮記念公園と和田堀公園が負の大きな値をとるジョイント空間の第1軸は、「調和がとれた自然環境」を表わす選好次元、同じく日比谷公園と上野恩賜公園が正の大きな値をとり、水元公園が負の大きな値をとるジョイント空間の第2軸は、「開放性」を表わす選好次元と解釈された。選好の個人差は、この二つの軸に対する各人の重みづけの違いによって生じているのである。This paper is concerned with a preference analysis of eight city parks with waterside landscape in and around Tokyo City (Fig.2). Eight sets of photographs,each of which consists of four landscape scenes for one park(see Appendix 2),were presented to 175 undergraduate and graduate students as the sample. Then they were asked to see the photographs to rank the parks in term of their preference. A preference ranking data-matrix thus obtained serves as the input for MDPREF (Chang and Carroll, 1968),a non-metric MDS algorithm applied to preference data,to represent the preference structure in a two一dimensional joint space. Interpreting the recovered joint space (Fig.3),based on external information or individual respondents' reasons for their evaluations,leads to a conclusion that the two dimensions of "harmonious natural environment" and "openness" underlie the respondents' preference judgements of city parks: for the former dimension,Arisugawanomiya Memorial Park and Wadabori Park are highly evaluated and Kasai Seaside Park and Ueno Park low evaluated; for the latter dimension,Mizumoto Park is highly evaluated and Hibiya Park and Ueno Park low evaluated. Inokashira Park and Ukima Park,in a sense,appear to lack distinguishing features since they are ambiguously evaluated for the both dimensions. These results suggest that the parks covered with green and/or those full of an open atmosphere are preferred while the parks impressing respondents artificially are less preferred., Sep. 1993, 49, 47-66, 66
  • Refereed, Geographical review of Japan, Series B., The Association of Japanese Geographers, 立地-配分モデルによるクリスタラー中心地理論の定式化の試み, ISHIZAKI Kenji, This paper reconstructs the marketing principle of Christaller's central place theory using the location-allocation model. According to the marketing principle, the system of central places is constructed in terms of the upper limit of the range by satisfying the following constraints :
    (a) All consumers must be supplied with all goods from central places.
    (b) The central places at a higher-level offer not only same order goods but also all the other goods of a lower-level (i. e., successively inclusive).
    In this way, Christaller's requirement organizes the system of central places which is based upon the condition of a minimum number of central places (Saey, 1973). Therefore, the marketing principle is formulated as the set-covering problem which minimizes the number of central places required to satisfy the demand for coverage subject to the maximum distance constraint. The model can be represented as a hierarchical programming (Daskin and Stern, 1981),
    miniZm=WΣxjmj+ΣjSim (1)
    subject to
    Σjaijmxjm-Sim=1 (2)
    xjm-xj(m-1)_??_0 (m=2, ……, L) (3)
    Sim_??_0 (4)
    where dij=the distance from demand node i to potential center site j; Rm=the range of good m, W=a very large positive weight;aijm=1 if the distance from i to j is less than or equal to Rm, 0 otherise; xjm=1 if good m is offered from node j, 0 otherwise; Sim=number of additional central places capable of serving node i with good m.
    The objective function (1) has two hierarchical objectives : the primary objective is to minimize the number of central places; the secondary objective is to minimize the extent of market overlap. Constraints (2) and (4) define the above constraint (a) in Christaller's theory. Constraint (3) insures that the hierarchy is successively inclusive (i. e., the above constraint (b)).
    The model is applied to a hypothetical lattice network (Fig. 5) which is composed of two parts : an inner network where the entry of central places is possible and all nodes must be served from central places; and an outer network where nodes are not potential center sites but demand nodes that do not have, to covered with all goods. To build the hierarchy, both the top-down method and the bottom-up method (Fig. 3) are applied. Results are summarized as follows :
    (1) As can be seen in Fig. 8 and Table 3, the system of central places generated by the bottom-up method is that of the K =3 system developed by Christaller. The solution result using the top-down method (Fig. 7 and Table 2), however, clearly does not fit Christaller's K =3 system. The system generated by the latter method reveals a complicated configuration of central places and has a number of hierarchical levels.
    (2) As compared with the result by the top-down method, we notice that the system built using the bottom-up method is characterized by a concentration of functions (Fig. 9) and, therefore, shows efficient hierarchical organization.
    Consequently, the above model is valid only if it is constructed by the bottom-up method. In order to generate the K=3 system by the top-down method, it is necessary to reinterpret Christaller's objective function., Oct. 1992, 65A, 10, 747-768, 768
  • Refereed, Annals of the Japan Association of Economic Geographers, The Japan Association of Economic Geography, The geographical development of fast food stores in Tokyo city area from the viewpoint of locational policy, ISHIZAKI Kenji, 1990, 36, 2, 129-140, 140


  • Not Refereed, 理論地理学ノート, 都市間結合を考慮した中心地モデルの定式化, ISHIZAKI Kenji, Jul. 2017, 19, 99-104
  • Not Refereed, 理論地理学ノート, A multiobjective location-allocation model for Christaller’s administrative principle, ISHIZAKI Kenji, 2016, 18, 21-26
  • Not Refereed, Geographical Reports of Tokyo Metropolitan University, Department of Geography, Tokyo Metropolitan University, A multiobjective maximal covering location problem incorporating inter-city traffic network, ISHIZAKI Kenji, Location-allocation models provide a valuable tool for operationalizing the central place theory. This paper attempts to integrate the marketing principle and the traffic principle in Christaller's central place theory using multiobjective model. The model is formulated with two objectives: (a) minimize the total population uncovered by each service, and (b) minimize the total cost of constructing arcs between centers. Some computational results indicate that the model is possible to derive the flexible system of central place hierarchy., 2015, 50, 50, 37-44, 44
  • Not Refereed, 奈良女子大学地理学・地域環境学研究報告, 地図変換技法の代替的アプローチに関する一考察, ISHIZAKI Kenji, 2010, 7, 49-64
  • Not Refereed, 古代学, 奈良女子大学, 明治前期の奈良盆地における中心地の階層構造-GIS・数理モデルによる立地分析-, ISHIZAKI Kenji, 2009, 1, 1, 69-75, 75
  • Not Refereed, 情報処理学会研究報告, 明治期の奈良盆地における集落の中心性について, ISHIZAKI Kenji, 2009, 2009-CH-83, 189-203
  • Not Refereed, 理論地理学ノート, 2種の点分布間における空間的適合に関する一考察, ISHIZAKI Kenji, 2006, 15, 1-12
  • Not Refereed, 地理月報, 商業施設の立地, ISHIZAKI Kenji, 2005, 485, 1-4
  • Not Refereed, 地理, 古今書院, 福祉政策-公共施設の立地・配分モデル, ISHIZAKI Kenji, 2001, 46, 3, 42-45, 45
  • Not Refereed, 地理, 古今書院, よりよい主題図を作成するために, ISHIZAKI Kenji, 1999, 44, 12, 36-46, 45
  • Not Refereed, Comprehensive urban studies, 東京都立大学都市研究センター, 多摩ニュータウン内における住宅地移動-多摩市の事例-, ISHIZAKI Kenji, 本稿では、多摩ニュータウン内での住宅地移動の実態について、多摩市でのアンケート調査に基づき報告した。明らかになった点は以下のとおりである。(1)転居の背景には、前住地における住宅状況に対する不満がある。(2)多摩ニュータウンが現在も建設途上にあり、前住地付近でも新しい住宅物件が供給され続けていることもあって、新居は前住地からあまり遠くない範囲で探されている。(3)対象世帯の世帯主が持ち家を指向する40-50歳代に集中しているため、持ち家間移動が借家・賃貸住宅から持ち家への移動を上回っている。(4)新居の選定に当っては、妻の意見が反映される場合が多く、主な選択基準は住宅状況であり、利便性、快適性が副次的に考慮される場合もある。This paper reports on intra-urban migration in Tama New Town, Tokyo Prefecture, based on questionnairing in Tama City. The questionnairing was conducted on 109 households who had changed their residences within Tama City from 1990 through 1995. The results are summarized as follows: the decision to move is related with dissatisfaction with housing conditions of their former residences; movers could search for the new residences in the area not so distant from their former residences because newly constructed houses continue to be supplied near their former residences in Tama New Town being still in the building; moves between the owner - occupied houses exceed those between the owneroccupied house and the rented house since many householders are in their forties and fifties who incline to own their houses; wife's will is likely to be more esteemed in choosing the new residence, and many movers regard housing conditions as the principal choosing criterion while some take into account accessibility and amenity, too., 1999, 70, 5-15, 15
  • Not Refereed, Notes on Theoretical Geography, GIS approach to evaluation of livign environment in Tama new town, ISHIZAKI Kenji, 1998, 11, 31-52
  • Not Refereed, Comprehensive urban studies, 東京都立大学都市研究センター, 店舗特性・立地特性からみた世田谷区におけるコンビニエンス・ストアの立地分析, ISHIZAKI Kenji, 本稿では、世田谷区におけるコンビニエンス・ストアの立地を対象として、現地調査に基づく品揃えのデータから抽出した店舗特性と、地理情報システムを用いて把握した店舗の立地特性について分析し、両者の関係を考察した。その結果、1)店舗数がほぼ飽和状態にある対象地域では、とくに上位チェーン企業による店舗の空間的競合が生じていること、2)各企業によって種々の店舗特性の構成に差異がみられ、他企業とは明確な差別化を図っていること、3)それらの店舗を多様な立地特性に対して集中的あるいは分散的に立地させることによって、コンビニエンス・ストア企業は需要および企業開の競合を見据えた戦略・立地適応を図っていること、などが明らかになった。The purpose of this paper is to verify the relationships between store merchandising and locational characteristics in convenience stores. It appears that the distribution of stores in Setagaya ward of this study area is saturated as regards the number of stores and the spatial competitions are intensive through regression analysis between night-time/ day-time populations and the number of convenience stores in Tokyo 23 wards. The stores are classified into seven types of store merchandising and into five types of locational characteristics, applying factor analysis and cluster analysis to the data of individual store's goods from field survey and the measured demand data from trade area analysis using geographic information systems. Consequently, each convenience store firm possesses a specific type of store merchandising for particular market segmentation and adopts a various target market selections according to its marketing ability., 1998, 65, 45-67, 67
  • Not Refereed, Comprehensive urban studies, 東京都立大学都市研究センター, 都市公園に対する知覚と選好の関係について-PREFMAPによる分析-, ISHIZAKI Kenji, 本稿は、東京区部とその周辺に位置する8公園(井の頭恩賜公園、水元公園、有栖川宮記念公園、葛西臨海公園、上野恩賜公園、日比谷公園、浮間公園、和田堀公園)を対象にして、都市公園に対する知覚と選好の関係を明らかにしようとした。そのために、8公園の風景写真を121名の大学生被験者に呈示し、知覚データと選好データを入手した。まず知覚データをKYST-2Aで分析し、それぞれ、「自然の環境が生み出す安らぎ感」、「都市公園の知覚された広さ」と命名される軸からなる2次元の知覚空間を求めた。そして、それを刺激空間座標として、選好との関係をPREFMAP(ベクトル・モデル型)で分析したところ、都市公園に対する知覚(公園の類似性判断)と選好(公園の噌好判断)の間には強い対応関係があることが判明した。The purpose of this paper is to elucidate the relationship between perception and perference for urban parks, taking as an example eight parks in and around Tokyo City. The perceptual and preference data on personal judgements of park-landscape scenes were collected through a questionnaire survey of a total of 121 undergraduate students. In order to get a stimulus configuration for an input to PREFMAP, a two-dimensional solution was sought by a non-metric MDS analysis of the perceptual data in terms of park similarity; the two dimensions were interpreted as the relaxed air produced by nature-environment and the perceived extent of urban parks, respectively (Fig.3) . A PREFMAP analysis -phase IV analysis or vector model- using this stimulus configuration revealed that perception of urban parks. that is to distinguish between them, was closely related to their preference judgement., 1994, 53, 57-70, 70
  • Not Refereed, 地理, 古今書院, 開発の波に揺れる秋留台, ISHIZAKI Kenji, 1993, 38, 9, 110-117, 117

Books etc

  • 978-4780312133
  • 人文地理学事典, 丸善出版, ISHIZAKI Kenji, 分担, Sep. 2013, 206-207, Not Refereed
  • 歴史GISの地平-景観・環境・地域構造の復原に向けて, 勉誠出版, ISHIZAKI Kenji, 分担, Mar. 2012, 159-170, Not Refereed
  • 大学的奈良ガイド-こだわりの歩き方, 昭和堂, ISHIZAKI Kenji, 分担, Apr. 2009, 23-37, Not Refereed
  • 地理学の声, 古今書院, ISHIZAKI Kenji, 分担, Nov. 2008, 28-52, Not Refereed
  • 18・19世紀の人口変動と地域・村・家族, 古今書院, ISHIZAKI Kenji, 分担, Mar. 2008, 66-88, Not Refereed
  • 地理空間分析, 朝倉書店, ISHIZAKI Kenji, 分担, Sep. 2003, 61-83, Not Refereed
  • 立地論入門, 古今書院, ISHIZAKI Kenji, 分担, Apr. 2002, 130-138, Not Refereed
  • GIS-地理学への貢献, 古今書院, ISHIZAKI Kenji, 分担, Feb. 2001, 216-228, Not Refereed
  • 都市をとらえる-地理情報システム(GIS)の現在と未来-, 東京都立大学都市研究所, ISHIZAKI Kenji, 分担, Mar. 1996, 173-211, Not Refereed
  • 都市の環境創造-環境と対話する都市, 東京都立大学都市研究所, ISHIZAKI Kenji, 分担, Mar. 1995, 221-255, Not Refereed


  • 日本地理学会研究奨励賞, 1996

Research Projects

  • Apr. 2017, Mar. 2022, 17K03247, Principal investigator
  • 地理情報システムや数理モデルを用いた地理空間分析, 0, 0, 0, Competitive research funding
  • Spatial Analysis using GIS and Mathematical Modelling, 0, 0, 0, Competitive research funding

Copyright © MEDIA FUSION Co.,Ltd. All rights reserved.