• No results found

Travelers’ Route Choice: Comparing Relative Importance of Metric, Topological and Geometric Distance

N/A
N/A
Protected

Academic year: 2021

Share "Travelers’ Route Choice: Comparing Relative Importance of Metric, Topological and Geometric Distance"

Copied!
8
0
0

Loading.... (view fulltext now)

Full text

(1)

Procedia Engineering 142 ( 2016 ) 18 – 25

1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the organizing committee of CUTE 2016 doi: 10.1016/j.proeng.2016.02.008

ScienceDirect

Sustainable Development of Civil, Urban and Transportation Engineering Conference

Travelers’

Route Choice: Comparing Relative Importance of

Metric, Topological and Geometric Distance

Amila Jayasinghe

a,

*, Kazushi Sano

b

, Rattanaporn Kasemsri

c

and Hiroaki Nishiuchi

d a,c

Doctoral Student, Graduate School of Civil & Environmental Engineering, Nagaoka University of Technology, Nagaoka, 940-2137, Japan b

Professor, Nagaoka University of Technology, 1603-1, Kamitomiokamachi, Nagaoka, Niigata Prefecture, 940-2137, Japan d

Assistant Professor, Nagaoka University of Technology, 1603-1, Kamitomiokamachi, Nagaoka, 940-2137, Japan

Abstract

This study investigates the relative importance of ‘metric distance’(MD),‘topological distance’(TD) and ‘geo-metrical distance’(GMD) in determining the route choice behavior of motorized travelers by mode. The study is based on data which has traced the travelers’ actual movements by using mobile GIS applications. MD, TD and GMD were calculated based on space syntax tool within a GIS environment. The findings indicated that GMD is more appropriate than MD in case of travelers’ who use car and motorcycle, whereas MD is more appropriate than GMD travelers’ who use PT and taxi for mode choice modeling, travel demand simulations and automobile navigation systems.

© 2016 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the organizing committee of CUTE 2016.

Keywords: Route choice; metric distance; geometric distance; gognitive behaviour; space syntax; travel demand.

1.Introduction

Modeling travel demand and development of navigation systems have been extensively focused on identifying and quantifying the most suitable or optimal path for travelers [1] in the fields of traffic and transport planning and engineering. ‘Rational behavior theory’ referring to the domain of traffic assignment has explained that individual travelers select the best route that maximizes their utility by comparing all possible alternatives and measuring their attributes. Stochastic choice models which explain the route choice behavior of travelers based on a utility function assesses the cost of various aspects of the journey with focus to minimize the cost while maximizing the utility. The

* Corresponding author. Tel.: +81-080-2675-1504; fax: +81-0258-47-1611. E-mail address: s147013@stn.nagaokaut.ac.jp and amilabjayasinghe@gmail.com

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

utility function is a complex measure comprised of many factors such as length, travel time and traffic congestion. The function differs by people to people depend on the different levels of their knowledge and the level of understanding of the network [2]. However, another cluster of studies including the works of Zhang [3]; Turner and Dalton [2]; Jan, Horowitz and Peng [4]; and Tversky [5] have highlighted that the utility function which have been developed based on length, congestion, travel time are far away from the actual situation and have been overlooked the traveler’s own perceptual and cognitive understanding of the road network. Further, Jiang et al [6] have highlighted that “drivers evaluate the alternative routes by individual experience, cognition, and attitudes which are not considered in the Expected Utility Theory (EUT) or Random Utility Theory (RUT) models” and Witlox argue that travellers’ reported distance estimates may cause a serious bias on maximize their overall utility [7]. When it comes to the navigation systems, many automobile navigation systems are developed based on Dijkstra algorithm which is known as the shortest-path algorithm. The algorithm identifies the shortest route between an origin and a destination in terms of the metric distance [8]. Metric distance is a simple measure that is used in most of the navigation systems to identify the best path. But recent route choice experiments have shown that minimizing turns is also an important factor determining the route choice of drivers [9] in contrast to the shortest-path algorithm only based on metric distance. “The notion of distance that people carry around in their heads, i.e., the so-called cognitive, estimated or subjective distance, is very different from the objective, real world distance [10]…. Human beings seem to incorporate a far more complex unity of (not always logical) criteria for path selection (i.e., least effort, shortest path, shortest time path, etc) which cannot be modelled in one simple algorithm” [7]. Ramming [10] has highlighted the need for more ‘human like’ route to be suggested to drivers’ rather confirming mainly to metrically the shortest path. Therefore, metric distance-based navigation systems may not be able to capture the driver’s intent fully.

As per the notion of ‘movement economies’, which has been explained in ‘cities as movement economies’, people move in lines and tend to approximate lines in more complex routes [11]. This has further suggested that the metric distance assumption is may not suitable, “not perhaps because we do not seek to minimize travel distance, but because our notions of distance are compromised by the visual, geometrical and topological properties of networks” [11]. Further, cognitive behavioral theories of human-way-finding, which is explained in neuroscience, have highlighted the role of ‘hippocampi’ in this regard. Hippocampus is a part of the brain that involved in body functions such as spatial orientation, navigation, and memorization. It conveys information about places based on landmark, unit distance and directional changes [12]. Accordingly, unit distance and directional changes can be considered as playing the key role in the route choice of humans. In addition, Dalton [13]; Duckham and Kulik [14]; and Hochmair and Frank [15] also have proposed that the directional change might be useful to direct travelers to their destination more simply in comparison to the metric distance. Further, Hillier and Iida [16] have argued that “topological and geometric complexities are critically involved in how people navigate [movements] urban grids [road network]. This has caused difficulties with orthodox urban modelling, since there it has always been assumed that…, it will be on the basis of metric distance”. Further to this, researchers’ argued that, past memories, current experiences [17]; structure and functional complexity of city [18]; mode and distance of travel [7] influence on individual’s estimate of distance.

In such background, this paper attempts to substantiate above mentioned statements. Accordingly, this study argues that, individual travelers’ route choice has greeter influence from the geo-topological distance (topological and geo-metric) than metric distance. This attempt constructively contributes to overcome three limitations which have been noted in emerging research in the domains of traffic and transport engineering, and urban planning. First, larger cities in the USA, Europe and Australia have been predominately referred in most of the previous studies in this domain. Very limited reference has been given to the medium-scale and small-scale cities in South Asia where human behaviors driven by diversified socio-cultural forces. Secondly, most of the available studies have been focused on either non-motorized movements or aggregated motorized movements. Investigating the impact of specific modes used by travelers has been paid a limited attention. Thirdly, most of the studies are based on data which people have reported as their route choice. Otherwise the observed traffic flow rates at gate locations. Very limited studies are based on the data which traced the actual movements of travelers. The main objective of this study was to explain the relative importance of ‘metric distance’, ‘topological distance’, and ‘geo-metrical distance’ in determining the route choice behavior of travelers by mode with reference to motorized movements. This paper used data on travelers’ actual movements which has traced by using mobile GIS application. A split of four modes

(3)

was considered discussing the route choice of travelers’. Colombo, which is an emerging South Asian city of Medium-scale, was referred as the case study. ‘Metric distance’ (MD), ‘Topological distance’ (TD), and ‘Geo -metrical distance’ (GMD) have been calculated employing space syntax technique. When comparing the relative importance of these three measures in determining the route choice, it was revealed that GMD as the most important factor among either car riders or motorcyclists whereas MD as the most important factor among taxi and PT users.

The paper has been structured into four sections. In section 2, we described the method of study which was employed in recording of the travelers’ movement tracks and identifying the relative importance of routes by distance (i.e. MD, TD & GMD respectively). We discussed the results in the section 3. Conclusions and recommendations were given in the final section.

2.Method of study

2.1.Study area

The study was conducted in Colombo Metropolitan Area (CMA), which can be considered as the main urban agglomeration in Sri Lanka. CMA is one of the medium-scale cities in South Asia with 5.8 million residential population and it accounts 30% of the country’s total population (DCS-SL, 2012). 40% of the trips in Colombo are made by public modes such as bus and railway. Around 38% of trips are made by private modes including motorcycles 14.1%, taxies 12.9% and cars 11.1%. The remaining trips are made by non-motorized transport modes [19]. The total number of trips in CMA is around 700,000 per day [19].

2.2.Data set - travelers’ route choice behavior

“The travel behavioral survey based on the questionnaire is not always sufficient for the measurement of travel behavior in space–time dimensions” [20]. This study employed open source mobile GIS application imbedded to cell-phone in tracing travelers’ movements. The sample included the movements of 250 travelers which have been traced within first 4 months of 2015. Travelers were asked to switch on the mobile tracking application which was installed in their cell phones, on all the journeys they took for their day to day activities. Each participant responded to a survey with questions about socio-economic characteristics (see Table 1.), the importance of various factors in choosing a given route and frequency of travelling in the route.

At the end of the period, the GPS based movements tracks were collected. The movement tracks were downloaded and geographically adjusted to road network by using GIS application. The tracks were underwent accuracy checking considering the continuity and geographical overlapping referring to the actual road network. 22% of tracks were removed due to lack of accuracy and 3,091 tracks were selected for further analysis. Then these tracks were assigned to road segments of the study area (see Fig. 1.) and categorized into 31 O-D pairs. O-D locations were selected based on the level of concentration of tracks at the points of start and end respectively. The considered O-D points were adjusted to the nearest well-known node (i.e. small town, popular intersection). Referring to these 31 O-D pairs, 410 routes were identified considering the routes which have been selected by at least one traveler. O-Distance of each route in terms of metric distance (MD), topological distance (TD), and geo-metrical distance (GMD) was calculated as per the method introduced by Hillier & Iida [16] (see table 2.) by a space syntax tool embedded to GIS application. In space syntax, geo-metrical distance is referred as angular distance.

Table 1. The characteristics of the sample of 250 travelers who participated in the survey

Sex % Mode % Income level (SLR) % Age %

Male 62 Car 27 <10,000 16 <20 08

Female 38 Motorcycle (MC) 23 10,000-25,000 46 20-30 26

Taxi 26 25,000-50,000 34 30-40 38

Public Transport (PT) 24 >50,000 04 40-50 18

(4)

Sex % Mode % Income level (SLR) % Age %

>60 03

Fig. 1. (a) road network of the study area; (b) travelers’ movement tracks Table 2. Methods of calculating the distance

Metric (MD) Topological (TD) Geo-metric (GMD) - Angular

MD is the cumulative ‘metric’ distance between two points.

Ex. DistanceAE = LengthAB+ LengthBC ..

= 5+7+10.6+5.7 = 28.3 km

TD is the cumulative number of ‘turns’ between two points.

Ex. DistanceAE

= Turn at B + Turn at C= 2 turns

GMD is the cumulative ‘angle change’ between two points.

Ex. DistanceAE

= 90/180×2 + 45/180×2= 1.5

2.3.Ranking of routes based on distance

All routes between each O-D pair were ranked based on MD, TD and GMD respectively. For instance, travelers

who travel from ‘Katubedda’ to ‘Nugegoda’ (refer to Fig.2, i.e., O-D pair ID 10) have used three alternative routes (i.e. route 1 in red, route 2 in purple and route 3 in blue in Fig.2.). Fig. 2 shows how these three routes are ranked by MD, TD, and GMD values computed by space syntax tool embedded to GIS application.

It shows that the route ranks are different when organized the data by MD, TD and GMD, respectively. For instance, the first rank was obtained by route 1 by MD, route 2 by GMD and route 3 by TD. In order to identify the measure which best represent the route choice, route ranks of three methods were compared with the actual number of travelers who have chosen the given route. Where there is the strongest inverse relationship between route rank and the number of travelers can be considered as the best measure (i.e. out of MD, TD and GMD) in explaining the

(5)

route choice of travelers. For this purpose, this study has compared the movement tracks obtained by 250 travelers with the MD, TD and GMD values computed by space syntax tool embedded to GIS application. First level of comparison was undertaken without considering the mode of travel, second level with considering the mode of travel and third level with considering both mode of travel and journey length. The findings of the study at each level

discuss how MD, TD and GMD are capable in representing the traveler’s route choice.

Fig. 2. an example of ranking routes between a selected O-D pair in terms of MD, TD and GMD

Histograms below (Fig. 3) indicate the number of travelers’ who selected each route by route rank which has been computed by MD, TD and GMD respectively. In all three graphs, the number of travelers shows an inverse relationship to route rank.

Fig. 3. The number of travelers’ (frequency) route choice based on route rank in terms of (a) MD, (b) TD and (c) GMD

The capability of MD, TD and GMD in explaining the route choice was assessed comparing the cumulative percentage distribution of a number of travelers who have chosen each route by route rank derived by MD, TD and GMD respectively (see table-3). When referring to rank-1, 23% of travelers’ has traveled on the route which was

selected as rank-1 according to the GMD whereas only 13% has traveled on the route which was selected as rank-1 according to the MD. Though this shows GMD better represent the traveler’s route choice compare to MD, this

requires further investigation. The next sections will investigate this further with reference to mode of travel and journey length. Then the number of passengers was categorized based on mode of travel and journey length. Figure 4 summarized the result of that. When refer the results pertaining to car or MC users, the results clearly indicated that high percentage of travelers (i.e. 59%, 53%, 56% and 52% of car riders who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively, and 48%, 44%, 48% and 48% of motorcyclists who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively) who travels along the routes which was recorded as rank-1 according to the GMD. In contrast, the percentage of travelers who travels along the route which was recorded as rank-1 according to the MD is low (i.e., In the case of cars; for 5-10km it is 46%, 10-15km it is 23%, 15-20km it is 7% and 20-25km it is 5% and in the case of motorcyclists; for 5-10km it is 39%, 10-15km it is 15%, 15-20km it is 6% and 20-25km it is 2%). In other words, the percentage gap between number of travelers’

(who used Car or MC) who select the top ranked routes (i.e., 1st, 2nd) by GMD and number of travelers’ who selected top ranked routes by MD is increased with the journey length. This finding can be substantiated by the claim of

O-D pair ID Route ID MD (meters) Rank-MD TD (no. of links) Rank-TD GMD (angular change) Rank-GMD 10 1 10,833 1 48 2 401.77 2 2 10,987 3 77 3 243.23 1 3 10,942 2 46 1 767.59 3 a b c

(6)

Hiller et al. [21] that is “urban space is locally [shorter journey length] metric and global [longer journey length]

topo-geometric”.

Table 3. The cumulative percentage distribution of travelers’ route choice based on route rank in terms of MD, TD and GMD

Rank-MD Cum. % Travelers Rank-MD Cum. % Travelers Rank-TD Cum. % Travelers Rank-TD Cum. % Travelers Rank-GMD Cum. % Travelers Rank-GMD Cum. % Travelers 1 13% 6 61% 1 20% 6 70% 1 23% 6 69% 2 29% 7 67% 2 36% 7 73% 2 38% 7 77% 3 43% 8 71% 3 50% 8 76% 3 48% 8 80% 4 49% 9 74% 4 58% 9 80% 4 59% 9 83% 5 57% 10 78% 5 64% 10 84% 5 65% 10 85%

Fig. 4. The cumulative frequency distribution of travelers’ route choice based on route rank and by mode and journey length

Results pertaining to PT or taxi users clearly indicated that high percentage of travelers (i.e. 73%, 41%, 34% and 48% of PT users who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively and 75%, 33%, 23% and 10% of taxi riders who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively) travels along the routes which was recorded as rank-1 and rank-2 according to the MD. In contrast, the percentage of travelers who travels along the route which was recorded as rank-1 and rank-2 according to the GMD is low (i.e. 29%, 29%, 0% and 0% of PT users who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively and 49%, 15%, 11% and 10% of taxi riders who travelled within the range of 5-10km, 10-15km, 15-20km, 20-25km respectively). Accordingly, the behavior of travelers’ who used PT or taxi cannot be explained

by the Hiller et al. [21] claim, though it explained the behavior of travelers’ who used Car or MC. While we explain the travelers route choice based on MD, TD and GMD, how travelers explain the reasons for their route choice? This

GMD MD TD GMD MD TD GMD MD TD GMD MD TD 1 59% 46% 45% 53% 23% 25% 56% 7% 24% 52% 5% 27% 2 61% 62% 46% 58% 27% 44% 59% 10% 45% 59% 8% 43% 3 69% 69% 56% 70% 41% 56% 66% 23% 47% 68% 10% 51% 4 75% 81% 59% 88% 43% 59% 77% 29% 56% 74% 21% 57% 5 93% 83% 64% 95% 50% 68% 83% 36% 67% 82% 22% 62% 6 100% 92% 81% 100% 51% 72% 92% 40% 74% 83% 28% 73% 1 48% 39% 26% 44% 15% 26% 48% 6% 37% 48% 2% 32% 2 59% 44% 44% 59% 21% 42% 57% 14% 46% 57% 17% 45% 3 70% 74% 85% 62% 37% 50% 62% 35% 57% 62% 24% 47% 4 94% 85% 89% 70% 41% 55% 69% 44% 59% 71% 32% 48% 5 100% 86% 91% 84% 52% 68% 75% 57% 62% 70% 37% 59% 6 100% 87% 100% 90% 54% 71% 77% 59% 64% 75% 41% 62% 1 32% 45% 32% 13% 19% 5% 3% 1% 7% 4% 4% 12% 2 49% 75% 57% 15% 33% 43% 11% 23% 23% 10% 10% 20% 3 59% 84% 89% 31% 55% 58% 26% 34% 31% 26% 26% 29% 4 66% 93% 94% 34% 67% 63% 35% 44% 41% 34% 34% 33% 5 69% 98% 96% 44% 75% 75% 43% 55% 54% 36% 36% 36% 6 72% 100% 100% 48% 81% 79% 49% 67% 65% 44% 54% 56% 1 13% 12% 29% 18% 16% 5% 0% 15% 0% 0% 21% 0% 2 29% 73% 45% 29% 41% 17% 0% 34% 0% 0% 48% 0% 3 35% 82% 91% 39% 64% 18% 3% 45% 1% 21% 59% 0% 4 38% 87% 93% 45% 69% 18% 37% 61% 5% 38% 65% 0% 5 40% 91% 95% 47% 71% 29% 39% 69% 8% 42% 69% 3% 6 42% 94% 100% 48% 75% 41% 42% 80% 9% 43% 83% 34% 30%> 10%30-50% 35%50-75% 47% 75%< 75% Ta x i PT Mode Ca r M o to rcycl e

Average journey length

5-10km 10-15km 15-20km 20-25km

Route Rank

(7)

was questioned during the survey of 250 respondents and the main reasons mentioned were travel time, cost, road condition, convenience and familiarity. The study attempt to interpret the route choices derived from computing MD, TD and GMD with the travelers responses (see Table-4).

Accordingly, travel cost and the travel time as mentioned as the key determinant of the route choice of taxi and PT users. Convenient and familiarity are mentioned as the key determinant of the route choice of car and MC users. It is explained in literature that MD has direct relationship with cost and travel time compare to GMD and TD. Convenience and familiarity related with cognitive behavior (i.e. which has direct relationship with GMD and TD than MD) of human way finding. In supporting the claim on applicability of TD and GMD, Hiller et el [21] further mention that “although it is perfectly plausible that people try to minimize distance, their concept of distance is, it seems, shaped more by the geometric and topological properties of the network more than by an ability to calculate metric distances. In general we might say that the structure of the graph governs network effects on movement and how distance is defined in the graph governs cognitive choices”. Though the route choice of car users and mortar cyclists substantiates this claim the validity becomes less prominent when reference to taxi and PT users who consider the travel time and cost the most.

Table 4. Percentage distribution of travelers’ responded as per key reasons to select a route by mode of travel.

Key reason to select route % of travelers’ by Car % of travelers’ by MC % of travelers’ by Taxi % of travelers’ by PT

Travel time 13.6% 13.9% 38.5% 25.5%

Travel cost 13.2% 11.8% 24.6% 49.3%

Road condition 10.9% 06.6% 18.9% 04.3%

Convenience 42.1% 44.6% 09.7% 07.6%

Familiar road 20.2% 23.1% 08.3% 13.3%

3.Conclusions and recommendations

The main objective of this study was to explain the relative importance of MD, TD and GMD in determining the route choice behavior of travelers by mode with reference to motorized movements. In order to accomplish this objective, the number of travelers (who used motorized modes) compared with the route ranks which obtained based on MD, TD and GMD. Study is based on data which has traced the travelers’ actual movements by using mobile GIS applications. A split of four modes (Car, MC, Taxi and PT) was considered discussing the route choice of travelers’. MD, TD, and GMD have been calculated employing space syntax tool within a GIS environment. In order to identify the measure which best represent the route choice, route ranks of three methods were compared with the actual number of travelers who have chosen the given route. For this purpose, this study has compared the movement tracks obtained by 250 travelers (3,091 tracks) with the MD, TD and GMD values. First level of comparison was undertaken without considering the modal split, second level with considering the mode of travel and third level with considering both mode of travel and journey length. Results revealed from this study can be summarized in three points as follows. First, the results indicated that, 23% of travelers’ has traveled on the route which was selected as rank-1 according to the GMD whereas only 13% has traveled on the route which was selected as rank-1 according to the MD. Second, the results pertaining to car or MC users, the results clearly indicated that high percentage of travelers who travels along the routes which was recorded as rank-1 according to the GMD. In contrast, the percentage of travelers who travels along the route which was recorded as rank-1 according to the MD is low. This phenomenon is further strengthened with the journey length. These two points are in line with works of Hiller et al. [21] as well as Dalton [13], Duckham & Kulik [14], Hochmair & Frank [15]. They have interpreted that human beings perceive the space mostly from geometrical and topological views rather than metric distance and the space is locally metric and global topo-geometric.

Hence, it can be concluded that, all three measures are important in explaining the route choice but it is more appropriate to consider geometric perspective in case of travelers’ who use Car and MC whereas metric perspective in case of travelers who use PT and taxi. Accordingly, this study suggests that GMD factor which is currently not represented in practice can be added as a key attribute in route choice modeling and travel demand in traffic

(8)

simulation and developing navigation systems. This offers the potentials of improving travel demand models’ and navigation systems accuracy and reliability while simulating more human like routes.

The current study examined only impact of shortest distance in terms of GD, TD and GMD. An essential extension of this study would necessitate extending the research in future to incorporate network centrality characteristics which are significantly influenced by topo-geometric aspect of distance.

References

[1] P. H. L. Bovy and E. Stern, Route Choice: Wayfinding in Transport Networks, Dorfrecht: Kluwer Acdamic Publishers, 1990, p. 9. [2] A. Turner and N. Dalton, "A simplified route choice model using the shortest angular path assumption," in Proceedings of the 8th

International Conference on GeoComputation, Michigan, USA, 2005.

[3] L. Zhang, "The Behavioral Foundation of Route Choice and Traffic Assignment," Transportation Research Record: Journal of the Transportation Research Board, vol. 2254.1, pp. 1-10, 2011.

[4] O. Jan, A. J. Horowitz & Z. Peng,"Using GPS data to understand variations in path," Transportation Research Record: Journal of the Transportation Research Board, vol.1725,pp.37-44, 2000.

[5] B. Tversky, "Distortions in cognitive maps," Geoforum, vol. 23, no. 2, pp. 131-138, 1992.

[6] X. Jiang, Y. Ji, M. Du and W. Deng, "A Study of Driver's Route Choice Behavior Based on Evolutionary Game Theory," Computational Intelligence and Neurosicence, 2014.

[7] F. Witlox, "Evaluating the reliability of reported distance data in urban travel behaviour analysis," Journal of Transport Geography, vol. 15, no. 3, pp. 172-183, 2007.

[8] S. Nakajima, D. Kitayama and Y. Sushita, "Route Recommendation Method for Car Navigation System based on Estimation of Driver’s

Intent," in Proceedings of IEEE International Conference on Vehicular Electronics and Safety (ICVES), Istanbul, 2012.

[9] L. Jeffrey and V. Blue, "Real-Time Multiple-Objective Path Search for In-Vehicle Route Guidance Systems," Transportation Research Part C: Emerging Technologies, vol. 6, no. 3, p. 157–172, 1998.

[10] M. Ramming, "Network Knowledge and Route Choice," PhD thesis, Massachusetts Institute of Technology, 2002. [11] B. Hillier, Space is the Machine: A Configurational Theory of Architecture, Cambridge: Cambridge University Press, 1999. [12] R. G. Gooledge, Processess, Wayfinging Behavior: Cognitive Mapping and other Spatial, Gooledge, R G;, Ed., Maryland: The Johans

Hopkins University Press, 1999.

[13] C.R.Dalton,"The secret is to follow your nose:Route path selection & angularity,"Environment & Behavior,vol.35, no.1, pp.107-131, 2003. [14] M. Duckham and L. Kulik, ""Simplest” Paths: Automated Route Selection for Navigation," In COSIT 2003, Lecture Notes in Computer

Science, vol. 2825, p. 169–185, 2003.

[15] H. Hochmair and . A. U. Frank, "“Influence of estimation errors on wayfinding decisions in unknown street networks analyzing the least-angle strategy," Spatial Cognition and Computation, vol. 2, p. 283–313, 2002.

[16] B. Hillie and S. Iida, "Network and psychological effects in urban movement," in Proceedings of Spatial Information Theory: International Conference, Berlin, 2005.

[17] R. D. Golledge and R. J. Stimson, Analytical Behavioural Geography, Croom Helm: New York, 1987. [18] D. J. Walmsley, Urban Living. The Individual in the City, Essex: Longman Scientific & Technical, 1988.

[19] JICA, "Final Report - CoMTrans Urban Transport Master Plan," Japan International Copperation Agency (JICA), 2014.

[20] Y. Asakura and E. Hato, "Tracking survey for individual travel behaviour using mobile communication instruments," Transportation Research Part C, vol. 12, p. 273–291, 2004.

[21] B. Hillier, A. Turner, T. Yang and H. Park, "Metric And Topo-Geometric Properties Of Urban Street Networks, Some convergences, divergences and new results," The Journal of Space Syntax, vol. 1, no. 2, pp. 258-279, 2010.

[22] S. L. Department of Census & Statistics, "Census of Population & Housing-2011," Department of Census & Statistics, Sri Lanka., 2012. [23] T. Kishimoto, S. Kawasaki, N. Nagata and R. Tanaka, "Optimal Location of Route and Stops of Public Transportation," in 6th International

Space Syntax Symposium, İstanbul, 2007.

[24] N. Raford, A. Chiaradi and J. Gil, "Space Syntax: The Role of Urban Form in Cyclist Route Choice in Central London," 1 4 2007. [Online]. Available: https://www.academia.edu/454861/Space_Syntax_The_Role_of_Urban_Form_In_Cyclist_Route_Choice_In_Central_London. [25] B. Jiang, "Flow Dimension and Capacity for Structuring Urban Street Networks," Physica A: Statistical Mechanics and its Applications,

vol. 387, pp. 4440-4452, 2008.

[26] T. Lotan, "Effects of familiarity on route choice behavior in the presence of information," Transportation Research, Part C, vol. 5, no. 3-4, p. 225–234, 1997.

[27] A. Turner, "From axial to road-centre lines: a new representation for space syntax and a new model of route choice for transport network analysis," Environment and Planning B: Planning and Design, vol. 34, no. 3, pp. 539-555, 2007.

[28] A. Hassan and M. D. Hoque, "Simplified Travel Demand Modeling Framework: in the Context of a Developing Country City," in the CODATU XIII, Ho Chi Minh City, Vietnam, 2008.

References

Related documents

Art is the Head of Mobext for Asia-Pacific, the mobile marketing agency brand of Havas Media, one of the world’s fastest-growing media agencies.. He is also the concurrent

Based on many interactions and discussions, the following important topics for future research (not covered in ongoing projects, but considered very important for future

features you guys are so lucky to have GIGABYTES of data to use Why is my Bluetooth on my Lumia 530 not connecting to my car Printable View net is a software job search resource

In these systems, users explicitly manipulate their brain activity instead of using motor movements to produce signals that can be used to control computers or communication

Please discuss location and open hours of hall concession stands, food service areas, and other Reliant Park food facilities with your Event Manager and Catering

Writing 3 Distribution Requirements 6 15 Fourth Semester CS 226 Computer Science IV 3 MTH 150 Elementary Statistics 3 BA 153 Management Foundations 3 Distribution Requirements 6

[r]

“the capability to provide service bundles which support the development of USOs’ entrepreneurial competencies in a sustainable way within the university structure.” Service