International Journal of
Geo-Information Article
Detecting Themed Streets Using a Location Based Service Application Byeongsuk Ji 1 , Youngmin Lee 2 , Kiyun Yu 2,3 and Pil Kwon 2, * 1 2 3
*
KT R & D Center Convergence Lab., 151, Taebong-ro, Seocho-gu, Seoul 06763, Korea;
[email protected] Department of Civil and Environmental Engineering, Seoul National University, 1, Gwanak-ro, Gwanak-gu, Seoul 08826, Korea;
[email protected] (Y.L.);
[email protected] (K.Y.) Institute of Construction and Environmental Engineering, Seoul National University, 1, Gwanak-ro, Gwanak-gu, Seoul 08826, Korea Correspondence:
[email protected]; Tel.: +82-2-880-6286
Academic Editors: Georg Gartner, Haosheng Huang and Wolfgang Kainz Received: 26 April 2016; Accepted: 8 July 2016; Published: 12 July 2016
Abstract: Various themed streets have recently been developed by local governments in order to stimulate local economies and to establish the identity of the corresponding places. However, the motivations behind the development of some of these themed street projects has been based on profit, without full considerations of people’s perceptions of their local areas, resulting in marginal effects on the local economies concerned. In response to this issue, this study proposed a themed street clustering method to detect the themed streets of a specific region, focusing on the commercial themed street, which is more prevalent than other types of themed streets using location based service data. This study especially uses “the street segment” as a basic unit for analysis. The Sillim and Gangnam areas of Seoul, South Korea were chosen for the evaluation of the adequacy of the proposed method. By comparing trade areas that were sourced from a market analysis report by a reliable agent with the themed streets detected in this study, the experiment results showed high proficiency of the proposed method. Keywords: GIS; themed street; Isovist; mobile sensor
1. Introduction In order to encourage the local economy and to establish the identity and placeness (sense of place) of an area, various streets have been created in USA. For example, Broadway and Wall Street in New York City and Hollywood Boulevard in Los Angeles are well-known themed streets and can easily be found online or in a web map service. These types of themed streets not only offer areas of special characteristics for a city, but also provide a place where the community can spend their leisure time. It is also known that the development of themed streets increases as a city matures [1]. While a themed street is recognized by the public, it is difficult to illustrate its exact boundary because the themed street is usually expressed as lines on a map. Since density based clustering or aggregation of polygons is normally used to draw the boundaries as areal shapes on maps, illustrating a line-based area such as a themed street is limited. In reality, people travel via roads and their activity areas are based on roads. The road is the first impression of a city; the features of interest along a road are, therefore, strongly related to the features of interest of the city [2]. In other words, the image of a street area as seen from the road can form an impression of the city within which it is located. From this perspective, a boundary of a special space should be expressed based on the road in order to ensure the public understands the city intuitively. To express a place based on a road, the characteristics of the road need to be determined. In most cases, the characteristics are formed according to the points of interest (POIs) on the roads. ISPRS Int. J. Geo-Inf. 2016, 5, 111; doi:10.3390/ijgi5070111
www.mdpi.com/journal/ijgi
ISPRS Int. J. Geo-Inf. 2016, 5, 111
2 of 15
By categorizing the POIs and assigning them to the road, various types of themes on the roads can be identified. However, few research papers have focused on detecting themes on roads. Even the related researches only show one measured phenomenon, such as the level of crime on the roads. With this method, the face of a city where various events occur cannot be illustrated. This study, therefore, used data on peoples’ behavior and POIs obtained from mobile GPS and Wi-Fi sensors to detect various themed streets. The Themed Street Clustering Method (the TSCM) is suggested for this purpose. For this study, two subtle words are introduced; “the hot street” and “the themed street”. Although the two words can be used interchangeably, the meaning of each is still defined for this study. As a hot spot is a clustered area of relatively high values, it represents a road that has a high value of a unique index. For example, the popularity index of a street is high if the street is popular. A themed street, however, is a street with a special placeness that is due to the close co-location of popular places; therefore, in addition to the popularity of a themed street, it is also known for a specialty. It was confirmed through this study that various themed streets have been detected via the use of a mobile sensor and collected data according to the TSCM. 2. Related Works To analyze the life pattern and characteristics of people according to space, the “Livehoods” project was conducted [3]. Check-in data (a user manually tells the application when he/she is at a certain location by selecting from a list of venues on a smart device) was used, as acquired from the Foursquare service and the spectral clustering method, which aggregates similar data and creates clusters. Using the method, “Livehoods” clustered sections were generated by calculating similar values of locations and attributes. However, it was difficult to distinguish the exact themes of the sectioned area since the data was not categorized in a manner that allowed for this. That is, “Livehoods” places more significance on determining geographic boundaries than on detecting themes associated with such boundaries. Meanwhile, a study was conducted to analyze hot spots using check-in data from Jiepang, a Chinese location based social media service [4]. The testing area was divided into fishnet grids and the number of check-in data was counted. Each grid was colored based on the significance level that expresses hot spots of the check-in. With the result, the researchers insisted that the check-in data indirectly reflect the population and economy of its area by showing the correlation between population census and the number of counted check-in data. While this study is a good reference in terms of research on socio-economic active areas, different results can arise when the fishnet grid size changes (also known as modifiable areal unit problem). In addition, the study only defined the area in which many check-ins occurred, only representing the check-in hot spots. Therefore, the themed regions cannot be understood through this type of analysis. The studies introduced above ([3]) have limitations, whereby the results can only be expressed as areal shapes (polygons). Other problems arose, such as the lack of any meaning for the division of areas, since the study measures only check-in frequency. Again, with the methods introduced by previous studies, identifying themed streets presents a challenge. In the meantime, studies on network-based clustering methods have been conducted in order to supplement the limitations of areal based clustering methods. One of the most popular and widely used methods for analyzing point distribution is known as kernel density estimation (KDE) [5]. However, conventional KDE has many faults in finding hot spots on road networks. One of the main flaws is that the density area of not only in the networks but also of the other areas are detected and this leads to skewed results. For this reason, network kernel density estimation (NKDE) was suggested [6] and it has been used for various applications such as detection of the likelihood of a hot spot in vehicle incidents [7]. Not only NKDE, but also network spatial and temporal analysis of crime (NT-STAC) and network spatial scan statistics (NT-SaTScan) were introduced to detect crime occurrences (robbery, burglary, drug deals/use, etc.) on road networks [8]. In the study, the researchers insisted that using STAC and SaTScan to detect hot spots results is producing round shaped clusters
ISPRS Int. J. Geo-Inf. 2016, 5, 111
3 of 15
on a 2-D space, and argued that the methods are limited when linear spaces are analyzed. Therefore, they demonstrated that NT-STAC and NT-SaTScan can successfully detect the crime areas on road networks throughout their research. In hot spot detection studies based on road networks, hot spots were found in linear spaces; however, these studies mostly focused on the clustering data overlaid on road networks and the road networks themselves were not considered. That is, the results only connect point data to linear shapes and do not provide significant information about the road networks. To overcome the above limitations, Lu (2005) explained the phenomenon of public socio-economic activities constrained by roads in a city by expanding the point data clustering method, which shows hot spots in 2-D spaces, to road networks [9]. Particularly, the term “hot street”, a road network acting as a hot spot, was introduced by conducting clustering testing of vehicle burglary point data on street segments, where the basic analysis involved dividing the units by each junction. Finally, the statistically significant hot streets were detected using Poisson distribution on each street segment. Nevertheless, the limitations of the research can be summarized as follows. First, the data used in the study only show the locations of vehicle burglaries on road networks. Second, the Poisson distribution can only analyze discrete data, such as the number of counted point data. Lastly, since the lengths of road segments differ, the incident cases according to the length of road segments could not be normalized. Previously published studies do not include methods for detecting various themed streets containing abundant attributes. Furthermore, it has been recognized that a themed street cluster analysis and visualization based on road segments needed. Therefore, this study suggested the TSCM to address the limits of the previous studies. 3. Themed Street Clustering Method 3.1. LBSM Data The raw data used in this study were created using mobile GPS or Wi-Fi signal, which can be represented as (x Pnt1 , y Pnt1 ). Additional information such as users’ comments and the history of leisure activities are added using data in location-based social media (LBSM) such as Foursquare, a mobile application which allows a user to check-in where they have been and leave a comment. Particularly, the venue of Foursquare was utilized to detect themed streets. Essentially, the venues are equivalent to the POI, which includes a great deal of information such as traces (check-in counts), addresses, grades, and so forth. The venues are used because the data directly records peoples’ perceptions of POIs and their behavior at the POIs. For example, if a significant amount of data on a venue is accumulated, it can be assumed that the venue is popular and well-known [10]. In other words, by analyzing the characteristics of the venues, the public interest in the locations and historical data can be obtained, which is not provided with normal GIS data. Therefore, the venues used for this study are the most suitable way to find various themed streets. Meanwhile, Goodchild et al. (2012) argued that volunteered geographic information (VGI) [11] can be used to build detailed and real time spatial data economically. However, they also pointed out the poor quality and accuracy of the VGI data [12]. This present study also verifies the quality of the venues. The venues employed in this study from Foursquare using Venues API are illustrated in Figure 1. As shown below, the venues are clearly identified; however, some venues are located in empty spaces and in the middle of streets. It is confirmed that the identified venues have low location accuracy, similar to that described by Goodchild et al. (2012). Sillim, located in Seoul, South Korea, was selected as a test area to check the position errors of the venues (SW: 126.928, 37.482; NE: 126.931, 37.486). Among the 412 venues, the coordinates of 128 venue samples were manually recorded to measure the Euclidean distance error between the manually input coordinates of the sample data (x Pnt2 , y Pnt2 ) and the samples’ raw coordinates (x Pnt1 , y Pnt1 ). Also, the azimuth of error vector was measured to verify whether the position errors have any particular
ISPRS Int. J. Geo-Inf. 2016, 5, 111
4 of 15
direction. The calculation of the Euclidean distance error between Pnt1 and Pnt2 is described in Equation (1) and the2016, azimuth between Pnt1 and Pnt2 is measured using Equation (2): ISPRS 5, 44 of ISPRS Int. Int. J.J. Geo-Inf. Geo-Inf. 2016, 5, xx FOR FOR PEER PEER of 15 15 b Pnt1 Pnt2 pmq “
= =
`
x Pnt1 ´ x Pnt2
˘2
− −
˘2 ` ` y Pnt1 ´ y Pnt2
+ +
− −
˙ ˆ y Pnt1 − ´ y Pnt2 180 180 − pdegreeq ˆ =Pnt Pnt “ arctan 180 2 1 ∠ = ´ x Pnt2 × ∠ = arctan arctan x Pnt1 − × ππ π
−
(1) (1) (2) (2)
(1) (2)
Figure 1. Venue distribution in Sillim test area.
Figure1.1.Venue Venuedistribution distribution in Figure in Sillim Sillimtest testarea. area.
Again, represents represents Again, represents the the raw raw coordinates coordinates of of the the samples samples and and represents the the manually manually Again, Pntcoordinates thethe raw coordinates of thethe samples and to Pntthe the manually input 2 represents 1 representsof input exact samples. Applying equations 128 samples, input exact coordinates of the samples. Applying the equations to the 128 samples, the the average average exact coordinates of the samples. Applying the equations to the 128 samples, the average Euclidean Euclidean Euclidean distance distance error error was was about about 50 50 meters meters (Figure (Figure 2a) 2a) and and the the azimuth azimuth of of error error vector vector did did not not distance error was about 50 meters (Figure 2a) and the azimuth of error vector did not show any show any particular patterns (Figure 2b). show any particular patterns (Figure 2b).
particular patterns (Figure 2b).
(a) (a)
(b) (b)
Figure 2. (a) Euclidean distance error histogram; (b) azimuth error histogram.
Figure 2. (a) Euclidean distance error histogram; (b) azimuth error histogram. Figure 2. (a) Euclidean distance error histogram; (b) azimuth error histogram.
The The position position error error is is caused caused by by the the mobile mobile sensors. sensors. In In usual usual cases, cases, when when users users check check in in to to aa venue venue The position error is caused by the mobile sensors. Inwhich usualbelongs cases, when users check inthey to aare venue or or create create aa venue, venue, they they tend tend to to do do so so inside inside the the building building which belongs to to the the venue. venue. While While they are or create venue, tend do so that inside building whichby belongs to the venue. While they doing they might not the location measured the sensor is the doingathis, this, theythey might nottorealize realize that thethe location measured by the mobile mobile sensor is actually actually the are doing this, they might not realize that the location measured by the mobile sensor is actually the venue venue location, which is not always true. It is known that the average position errors of a positioning venue location, which is not always true. It is known that the average position errors of a positioning system using mobile GPS and Wi-Fi are 10 meters and 40 meters, respectively [13]. Thus, the position location, which is not always true. It is known that the average position errors of a positioning system system using mobile GPS and Wi-Fi are 10 meters and 40 meters, respectively [13]. Thus, the position error of concluded GPS and environments or patterns. using mobile GPScan andbe Wi-Fi are 10from meters and meters, respectively [13]. Thus, usage the position error of error of venue venue can be concluded from GPS and40Wi-Fi Wi-Fi environments or Foursquare Foursquare usage patterns. For this study, the coordinates of venues were manually edited to correct the position error. For study, thefrom coordinates ofWi-Fi venues were manually to correct the position error. As As venue can bethis concluded GPS and environments oredited Foursquare usage patterns. aaFor result, of venues extracted from raw were geocoded. Then, venues were spatially result, 87% of the the venues extractedof from raw data data were geocoded. Then,tothe the venuesthe were spatially this87% study, the coordinates venues were manually edited correct position error. joined joined with with aa corresponding corresponding building building polygon polygon layer. layer.
As a result, 87% of the venues extracted from raw data were geocoded. Then, the venues were spatially joined with a corresponding building polygon layer. 3.2. 3.2. Road Road Segments Segments and and Spatially Spatially Joined Joined Venue Venue Data Data Matching Matching
the network layer without dividing itit into 3.2. RoadUsing Segments and Data Matching Using the road roadSpatially networkJoined layer Venue without dividing into smaller smaller areas, areas, the the length length of of the the road road segment segment can can be be longer longer than than the the length length of of the the study study area. area. The The longer longer segment segment can can be be detected detected as as aa
Using the road network layer without dividing it into smaller areas, the length of the road segment can be longer than the length of the study area. The longer segment can be detected as a themed street,
ISPRS Int. J. Geo-Inf. 2016, 5, 111
5 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
5 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
5 of 15
because it can be affected more by the buildings to which the venues belong. Thus, road networks are themed street, because it can be affected more by the buildings to which the venues belong. Thus, splitroad at each junction 3), since the junction interrupts cognition of venues a continuous networks are(Figure split itatcan each junction (Figure since thethe junction interrupts the cognition of a and themed street, because be affected more by3), the buildings to which the belong.space Thus, people usually stopare at thepeople junction when walking along awhen road [9]. continuous space and usually stop at the junction walking along a road road networks split at each junction (Figure 3), since the junction interrupts the[9]. cognition of a continuous space and people usually stop at the junction when walking along a road [9].
Figure 3. Dividing a road into several road segments.
Figure 3. Dividing a road into several road segments. Figure 3. Dividing a road into several road segments.
The building polygon layer was matched to divided road segments to assign venue information The building polygon layer was matched to divided road segments to assign venue information to theThe segment. The matching condition between road segments and the to building layer isinformation defined as building polygon layer was matched to divided road segments assign venue to the segment. The matching condition between road segments and the building layer is defined considering the frontage space of acondition building, the building should touch the road and the building front to the segment. The matching between road segments and thesegment, building layer is defined as as considering frontage space ofUsing a building, the building should touch theroad roadsegment, segment, and building should bethe seen the road. Isovist, building are the matched to roadand segments that only considering thefrom frontage space of a building, the buildingpolygons should touch thethe building frontfront should be seen from thethe road. Using arematched matched road segments influence the target [14]. should be seen from road. UsingIsovist, Isovist,building building polygons polygons are toto road segments thatthat onlyonly influence the target [14]. Isovist is a visible area that can be seen from a given location (a view point), excluding any area influence the target [14]. beyond surrounding obstacles. Isovist can the process of people recognizing signboards andarea Isovist a visible area that canbe beseen seenshow from a a given given location (a(aview point), excluding anyany area Isovist is aisvisible area that can from location view point), excluding walking towards aobstacles. building the Isovist area, created from a building beyond surrounding obstacles.intuitively. Isovistcan canFurthermore, show the process of recognizing signboards andand beyond surrounding Isovist show process ofpeople people recognizing signboards centroid as a view point, does not increase due to neighboring buildings; this feature is very similar walking towards a building intuitively. Furthermore, the Isovist area, created from a building walking towards a building intuitively. Furthermore, the Isovist area, created from a building centroid people’s view in thenot realincrease world, sincetopeople’s sight buildings; are interrupted The centroid as point adoes viewofnot point, does due neighboring thisisfeature isbarriers. veryto similar as a to view point, increase due to neighboring buildings; thisalso feature veryby similar people’s Isovist area point therefore matches road segment, if theinterrupted building isby located within to people’s of view in thethe realappropriate world, since people’s sighteven are also barriers. The point of view in the real world, since people’s sight are also interrupted by barriers. The Isovist area complicated road structures, described in Figure 4a. Additionally, the building road layer was used as a Isovist area therefore matchesasthe appropriate road segment, even if the is located within therefore matches the appropriate road segment, even if the building is located within complicated barrier to interrupt the growing Isovist area. Due 4a. to this effect, incorrect all road complicated road structures, as described in Figure Additionally, the roadmatching layer wastoused as a roadsegments structures, every as described inwas Figure 4a. Additionally, theFigure road layer was used as a barrier to interrupt junction resolved illustrated in this 4b. incorrect matching to all road barrier toatinterrupt the growing Isovistasarea. Due to effect, the growing Duewas to this effect,asincorrect matching segmentsIsovist at everyarea. junction resolved illustrated in Figureto 4b.all road segments at every junction
was resolved as illustrated in Figure 4b.
(a) (a)
(b) (b)
Figure 4. Building layer and road segment matching examples; (a) matching to appropriate road segment; by using roadand layer as asegment barrier, Isovist areaexamples; from the building at corner matched only to Figure 4. (b) Building layer road matching to to appropriate road Figure 4. Building layer and road segment matching examples;(a) (a)matching matching appropriate road road segments satisfy match Otherwise, dashed line could be matched. segment; (b) bythat using roadthe layer as a condition. barrier, Isovist area from the building at corner matched only to
segment; (b) by using road layer as a barrier, Isovist area from the building at corner matched only to road segments that satisfy the match condition. Otherwise, dashed line could be matched. road segments that satisfy the match condition. Otherwise, dashed line could be matched.
ISPRS Int. J. Geo-Inf. 2016, 5, 111
6 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
6 of 15
To use Isovist, a certain range should be set based on peoples’ viewpoints. As mentioned above, To use Isovist, a certain range should set based peoples’ viewpoints. As is mentioned the centroids of each building were set tobe match theseonviewpoints. If the range set as theabove, same the centroids each building were area set tocould match viewpoints. If the range is set as the same distance for allofbuildings, an Isovist be these created inside buildings. To prevent this type of distance all buildings, Isovist area could be in created inside issue, thefor Isovist range wasan calculated as described Equation (3):buildings. To prevent this type of issue, the Isovist range was calculated as described in Equation ˜ ? ¸ (3): 2 (3) sight_dist pmq “ l ˆ √2 ` 5 (3) ℎ _ = × 2 +5 2 where where l isisthe theside sidelength length of of aa building building and and all all buildings buildings in in the the study study area area are are regarded regarded as as square square shapes. Then, the the width width of of the the frontage frontage space space of of aa building building and and sidewalk, sidewalk, five five meters, meters, was was added. added. shapes. Then, After After the the Isovist Isovist area area was was trimmed trimmed to to aa road road segment, segment, the the Isovist Isovist area area would would be be restricted restricted by by roads roads and neighboring buildings (Figure 5). and neighboring buildings (Figure 5).
Figure 5. Isovist from aa viewpoint. viewpoint. Figure 5. Isovist area area created created from
Because the the venues venues were were spatially spatially joined joined to to building building layers layers in in the the previous previous process, process, and and Isovist Isovist Because areas were created from the building, the Isovist area also contains the venue attributes. Thus, Isovist areas were created from the building, the Isovist area also contains the venue attributes. Thus, Isovist areas are are spatially spatially joined joined to to road road segments segments to to assign assign venue venue attributes attributes to to roads. roads. areas 3.3. Themed Cluster Detection Detection 3.3. Themed Street Street Cluster 3.3.1. Hot Value Each road segment should have unique values to establish criteria, whether or not the road segments For this study, thethe unique value is called the “hot valuevalue (Hi )” segments are areregarded regardedasasthemed themedstreets. streets. For this study, unique value is called the “hot and is derived by Equation (4), as(4), follows: ( )”itand it is derived by Equation as follows:
=Pˆ ×R ˆ ×D Hi “ i i i
(4) (4)
where is defined as the sum of the average popularity of venues ( ), the ratio of the venue where Hi is defined as the sum of the average popularity of venues (Pi ), the ratio of the venue buildings buildings ( ), and the density of the venue buildings ( ). is subsequently used as the input (Ri ), and the density of ∗the venue buildings (Di ). Hi is subsequently used as the input dataset of dataset of Getis-Ord’s in this paper. The calculation details of each factor are explained next. Getis-Ord’s Gi˚ in this paper. The calculation details of each factor are explained next. is the sum of opinions divided by the number of venues on the road and is given by Equation Pi is the sum of opinions divided by the number of venues on the road and is given by Equation (5), (5), as follows: as follows: ∑k ř , (5) = j“1 Ni,j Pi “ (5) k where is the total number of venues on the th road segment; , is the quantified numerical where k is the total number of venues on the ith road segment; Ni,j is the quantified numerical value value such as the number of check-ins, tips, likes, and others, which consider people’s positive actions such as the number of check-ins, tips, likes, and others, which consider people’s positive actions to the to the specific venues [10]. The , is then measured, as in Equation (6): specific venues [10]. The Ni,j is then measured, as in Equation (6): (6) , = ℎ , + , + , Ni,j “ check i,j ` tipi,j ` likei,j (6) refers to the popularity of the th venue amongst the venues on the th road segments, , when the number of road segments exist. That is, refers to the average popularity of the th road segment, when there are number of road segments.
ISPRS Int. J. Geo-Inf. 2016, 5, 111
7 of 15
Ni,j refers to the popularity of the jth venue amongst the venues on the ith road segments, when the n number of road segments exist. That is, Pi refers to the average popularity of the ith road segment, when there are n number of road segments. Second, if a similar theme of venues exists on a road with a sufficient number of venues, the road tends to take on the characteristics of the venues. This will help people recognize the road as a themed street in that venue. In Equation (7), Ri is a mathematical expression of the people’s recognition and the ratio of the venue buildings, as follows: Ri “
BVi BNi
(7)
where BNi is the number of buildings that touch the ith road segment; BVi is the number of buildings that have venue data belonging to the matched buildings on the ith road segment. That is, Ri is the ratio of the venue buildings on the ith road segment. Therefore, as Ri increases, the venue buildings on the target road segment become denser. Lastly, because the usage of only Pi and Ri can cause a statistical error (i.e., one popular building among two buildings and five popular buildings among 10 buildings could lead to the same result), the density of the venue buildings on the target road (Di ) is finally added to Hi . Di is the total number of the venue buildings on a length of road segment, as in Equation (8). Di “
BVi lengthi
(8)
where BVi is the number of venue buildings on the ith road segment; lengthi is the length of the ith road segment in meters; and Di is the intuitive index that indicates the density of venue buildings on a road segment. This equation is also used to distinguish whether or not a road has the same Pi and Ri . Consequently, Hi is the hot value of the ith road segment. A higher Hi implies the greater popularity of the venues, the higher ratio of the venue buildings, and the higher density of the venue buildings. It can therefore be concluded that it is more likely that road segments with a higher Hi will be detected as themed streets. 3.3.2. LISA Among the spatial cluster detection methods, this study considered using local indicator of spatial association (LISA) [15]. The spatial cluster detection method using LISA has been standardized previously in many studies. The specific LISA can be calculated, significantly important aggregation can then be extracted, and the aggregation can be named as a spatial cluster (hot or cold spot) after the significance of the value test [16]. Getis-Ord’s Gi˚ , among LISA, was chosen for this study. The most significant advantage of Gi˚ is that hot and cold spots can be intuitively identified through statistical results. Getis-Ord’s Gi˚ is measured as shown in Equation (9). ř Gi˚
“ d s
wij x j ´ x
ř
j“1 wij ´ ¯2 ř ř n j“1 wij2 ´ j“1 wij n ´1
j “1
(9)
where s is standard deviation; wij is the spatial weight between spatial unit i and j; and n is the total number of data. If the units are defined as adjacent, wij “ 1 and 0 otherwise. Since Gi˚ is regarded as a neighborhood, wij is 1. This also means that the spatial weight matrix diagonal values are not 0. The expectation value of Gi˚ is 0 and the variance is 1 [17]. Therefore, the significance test of Gi˚ is processed almost the same as the normal distribution test [16]. To measure the Gi˚ index, the spatial weight matrix should be modified for the spatial unit of this study, the road segments. If the road segments are connected, wij is 1, otherwise it is 0. Figure 6a shows
ISPRS Int. J. Geo-Inf. 2016, 5, 111
8 of 15
the nine road segments connected at junctions and the spatial weight matrix of the road segments is described Figure2016, 6b.5, x FOR PEER ISPRS Int.in J. Geo-Inf. 8 of 15 ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
(a) (a)
8 of 15
(b) (b)
Figure Spatial weight weight matrix (a) (a) sample roadroad segments; (b) spatial weight matrix Figure of Figure 6. 6.Spatial matrixexample: example: sample segments; (b) spatial weightofmatrix Figure 6. Spatial weight matrix example: (a) sample road segments; (b) spatial weight matrix of Figure 6a. Figure 6a. 6a.
To achieve the normality (expectation value of 0 and variance of 1) of Getis-Ord’s ∗ ,˚a normal ToTo achieve the normality (expectation value of and varianceof of 1)ofofGetis-Ord’s Getis-Ord’s G ∗ i , a normal achieve the normality of 00 that and the variance , a normal distribution of the spatial data (expectation is preferable.value Provided number 1) of spatial data is sufficient and distribution of the spatial data is preferable. preferable. Provided that number is sufficient ∗ of spatial distribution spatialspatial data is thethe number databeisdata sufficient and the number of the adjacent units is moreProvided than 30, that the normality ofof spatial can still assumed even ∗ of G ˚ can still be assumed and the number of have adjacent spatial is more than 30, the normality the number adjacent spatial unitsunits is more than the normality canofstill be assumed even i spatial if the spatialof data skewed distribution. On the30, other hand, if theofnumber data is small, even if the spatial data have skewed distribution. On the other hand, if the number of spatial data is ∗ spatial if the spatial data have skewed distribution. On the other hand, if the number of data is small, and the number of adjacent spatial units is less than 30, the normality of cannot be assumed, ∗ G ˚ cannot be assumed, small, and the number of adjacent spatial units is less than 30, the normality of and the number of adjacent spatial units is less than 30, the normality of cannot be assumed, i provided the skewness is moderate [18]. provided the skewness isisthis moderate [18]. provided the skewness moderate [18]. The spatial unit for study is split at the junctions. Hence, the number of adjacent spatial units spatial unit this study split junctions. Hence, the number spatial The spatial unitfor forabove, this study atat thethe junctions. the number of adjacent spatialcan units is The low. As mentioned if theisissplit distribution has aHence, moderate skewness, the of ∗adjacent index be ∗ is low. As As mentioned above, if normality. the has moderate skewness, index can be units is low. mentioned if the distribution Hi However, distribution a moderate skewness, Gi˚ positive index can measured assuming that itabove, has as ahas shown in Figure 7a, the has the a very assuming that it has normality. However, as shown in Figure 7a, has a very positive be measured measured assuming that it has normality. However, as shown in Figure 7a, H has a very positive skewness. Therefore, the skewed values were normalized using a natural logarithm to calculate the i ∗ skewness. Therefore, skewed values were normalized using naturallogarithm logarithmtotocalculate calculate theG˚ skewness. thethe skewed values were using a anatural the indexTherefore, (Figure 7b). Nevertheless, it is notnormalized critical to follow the normal distribution strictly, since i ∗ ∗ index 7b).does Nevertheless, it is notequations critical to regarding follow thethe normal distribution since Getis and(Figure Ord’s not provide clear calculation of strictly, the strictly, variance and index (Figure 7b). Nevertheless, it is not critical to follow the normal distribution since Getis ˚ value∗[17]. Getis andGOrd’s does not provide clear equations regarding the calculation of the variance and theOrd’s expected and i does not provide clear equations regarding the calculation of the variance and the the expected value [17]. expected value [17].
(a) (a)
(b) (b)
Figure 7. Test area histograms: (a) distribution of ; (b) distribution of natural logarithm value of . Figure 7. Test area histograms:(a) (a)distribution distribution of of H ;; (b) distribution of natural logarithm value of . Figure 7. Test area histograms: i (b) distribution of natural logarithm value of Hi .
Finally, the ∗ index was measured using ln . For this study, only the positive ∗ index ∗ ∗ ∗ Finally, theregarded index measured using ln . For thisofstudy, the of positive index (hot spot) was as was a measured target. With the significance 0.05,only aonly score of˚over 1.96 Finally, the Gi˚ index was using lnpHi q. For level this study, the positive∗ G index (hot i (hot spot) was as a target. With the significance level of 0.05, a score of˚ of over 1.96 was selected asregarded themed street clusters. spot) was regarded as a target. With the significance level of 0.05, a z score of Gi of over 1.96 was was selected as themed street clusters. selected as themed street clusters. 4. Results and Analysis 4. Results and Analysis 4. Results and Analysis Sillim and Gangnam, two districts in Seoul, were chosen for the examination of the results and Sillim andofGangnam, two districts the versatility the suggested method. in Seoul, were chosen for the examination of the results and Sillim and Gangnam, two districts in Seoul, were chosen for the examination of the results and the versatility the suggested The total of number of venuemethod. data is 312 in the Sillim test area. It was confirmed that the number the versatility ofnumber the suggested method. The total of venue data 312 in the test area.isIt154 was confirmed that the number of venue buildings and matched roadissegments toSillim the buildings and 152, respectively. of venue buildings and matched road segments to the buildings is 154 and 152, respectively.
ISPRS Int. J. Geo-Inf. 2016, 5, 111
9 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
9 of 15
The total number of venue data is 312 in the Sillim test area. It was confirmed that the number of venue buildings matched road to thethemed buildings is 154 and 152, The results and of the analysis of segments the restaurant street clusters in respectively. the Sillim test area are The results of the analysis of the restaurant themed ∗ street clusters in the Sillim test area are illustrated in Figure 8b. The red colored streets have a value that satisfies the significance level of ˚ illustrated inregarded Figure 8b. red colored have a Gcolored that satisfies level of i valuebuildings 0.05 and are asThe themed streets, streets while the blue referthe to significance the scale of popular 0.05 and are regarded as themed streets, while the blue colored buildings refer to the scale of popular venues ( ). To verify the normalization of the ln distribution of the restaurant theme, a normal venues (Pwas verify(Figure the normalization of the lnpHi q distribution of the restaurant theme, a normal i ). Todrawn Q-Q plot 8a). Q-Q plot was drawn (Figure 8a).
(a)
(b)
Figure 8. Sillim results: (a) (b) Restaurant Restaurant themed themed streets. streets. Figure 8. Sillim test test area area results: (a) Q-Q Q-Q plot plot for for normal; normal; (b)
The test test area area has has three three large large restaurant restaurant themed themed street street clusters. clusters. District District A A is is relatively relatively smaller smaller The than the other districts and is located some distance from them. District A is named a “Fashion and than the other districts and is located some distance from them. District A is named a “Fashion and Cultural Street”, originally developed by the local government. However, the public has criticized Cultural Street”, originally developed by the local government. However, the public has criticized the development, development, stating tax money money was was wasted wasted and and the the original original aim aim of of the the development development was was the stating that that tax not met, since most of the fashion-related stores are closed and restaurants now line the street [19]. not met, since most of the fashion-related stores are closed and restaurants now line the street [19]. District A A reflects reflectsthe theactual actualcondition condition directly showing functionality of street the street as a alley. food District directly byby showing thethe functionality of the as a food alley.It was observed that district B has many restaurants, with 92 restaurant venues within the district. It was observed that district B has restaurants, with 92(named restaurant venuesin within the district. Among these, about half are related tomany selling Korean sausages “sundae” Korean) (44 out Among these,sell about half sausages). are relatedThis to selling Korean sausages (named “sundae” in Korean) (44 out of 92 venues Korean unusual restaurant distribution was reported in the market of 92 venues sell Korean sausages). This unusual restaurant distribution was reported in the market analysis report on Sillim published by the Small Business Development Center (SBDC). The area is analysis“Sundae report on Sillimthe published by the Business Development Center (SBDC). Thethrough area is named Town”, Korean term for Small “Sausage Town”. The report was confirmed again named “Sundae Town”, the Korean term for “Sausage Town”. The report was confirmed again the test result. through the test result. District C has various restaurants such as Pizza Hut, McDonald’s, Chicken barbecue, etc. along District C has various restaurants as Pizzaclose Hut,toMcDonald’s, barbecue, etc. with Sillim-ro (Sillim Avenue). District such C, is located district B, toChicken the south; however, thealong two with Sillim-ro (Sillim Avenue). District C, is located close to district B, to the south; however, the two districts have not merged as one large cluster. This shows that the restaurant venues between districts districts have not merged as oneattention, large cluster. This shows that theCrestaurant venues between districts B and C do not receive enough which separates B and as different districts. B andToCdeduce do not receive enough attention, which and C as different the various themed streets, notseparates only the Brestaurants, but alsodistricts. the themes of cafés To deduce the various themed streets, not only the restaurants, but 9c) alsowere the tested themesusing of cafés (Figure 9a), fashion stores (Figure 9b), and entertainment facilities (Figure the (Figure 9a), fashion stores (Figure 9b), and entertainment facilities (Figure 9c) were tested using the same method. The most noticeable aspect of Figure 9 is that a fashion themed street cluster was not same method. The most noticeable aspect of Figure 9 is that street cluster not ˚ value themed detected (Figure 9b). None of the road segments satisfies the aGfashion of the significant levelwas of 0.05, ∗i detected (Figure 9b). None of the road segments satisfies the value of the significant level of 0.05, although a few fashion venues appear in the data. However, a building located to the southeast side of although a few fashion venues in the data.popularity. However, aItbuilding located the southeast side the intersection in the test area appear shows the highest can be seen thatto the popular fashion of the intersection in the test area shows the highest popularity. It can be seen that the popular fashion stores are entered through a mall rather than via streets. stores are entered through a mall rather than via streets.
ISPRS Int. J. Geo-Inf. 2016, 5, 111 ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
(a)
10 of 15 10 of 15
(b)
(c)
Figure 9. 9. Themed results in in Sillim Sillim test test area: area: (a) Figure Themed street street cluster cluster detection detection results (a) Café Café themed themed street; street; (b) (b) Fashion Fashion themed street; (c) Pub themed street. themed street; (c) Pub themed street.
The G˚∗ z-score z-scoreand andp-value p-value of each theme are plotted sothe that the overall detections of the The of each theme are plotted so that overall detections of the themed i themed streets can be observed (Figure 10). The values are ordered according to the z-score in streets can be observed (Figure 10). The values are ordered according to the z-score in accordance ∗ accordance with its p-value. Notice that the data shows the calculations of the road segments that ˚ with its p-value. Notice that the data shows the Gi calculations of the road segments that contain containinformation. venue information. Furthermore, the frequency of eachis theme is different because of venue Furthermore, the frequency of each theme different because not all ofnot the all road the road segments contains venues. The roads that do not contain any data were disregarded. segments contains venues. The roads that do not contain any data were disregarded.
(a)
(b)
(c)
(d)
∗ Figure 10. segments of of each theme in the Sillim testtest area: (a) Figure 10. Gi˚z-scores z-scoresand andp-values p-valuesofofthe theroad road segments each theme in the Sillim area: restaurant; (b) café; (c) fashion; (d) pub. (a) restaurant; (b) café; (c) fashion; (d) pub.
Four different different venues venues were were used used to to detect detect the the various various themed themed street street clusters. clusters. While While themed themed Four streets represent the placeness of a region, they are also important in forming the business streets represent the placeness of a region, they are also important in forming the business district. district. Therefore, the the sum sum of of all all the the themed themed street street clusters clusters mentioned mentioned above above can can be be regarded regarded as as aa business business Therefore, district. The total area was visually compared to that in a market analysis report from the SBDC of Korea, presented in June 2008.
ISPRS Int. J. Geo-Inf. 2016, 5, 111
11 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
11 of 15
district. The total area was visually compared to that in a market analysis report from the SBDC of Korea, in trade June 2008. A presented map of the area from the market analysis report (Figure 11a) and the themed street A map of the area from the market report (Figure thedid themed street clusters along withtrade buildings (Figure 11b) is analysis illustrated below. The 11a) test and result not reveal clusters along with buildings (Figure 11b) is illustrated below. The test result did not reveal abnormally abnormally incorrect areas and most of the areas created by the test were inside of the area described incorrect areas and most thevisual areas created by theittest inside the area described in the SBDC in the SBDC report. Fromofthis comparison, can were be seen thatof the test method is reliable. It was report. From comparison, it can be method is reliable. It was observed that observed thatthis thevisual market report presented in seen 2008that did the not test provide updated information. The themed the market report presented in 2008 did not provide updated information. The themed street clusters street clusters had expanded to the south compared to the 2008 report. Also, the Fashion and Cultural had expanded the south 2008not report. Also,tothe Fashion and Cultural street was not street was not to surveyed bycompared the SBDC,tosothe it was possible make a comparison. surveyed by the SBDC, so it was not possible to make a comparison.
(a)
(b)
Figure 11. 11. Visual Visual assessment assessment test: test: (a) Figure (a) trade trade area area from from the the Small Small Business Business Development Development Center Center (SBDC) (SBDC) market analysis report; (b) trade areas from the SBDC (red line) and sum of all the themed street street market analysis report; (b) trade areas from the SBDC (red line) and sum of all the themed clusters (dashed black line). Blue polygons are buildings that have venues data. clusters (dashed black line). Blue polygons are buildings that have venues data.
In the the meantime, meantime,the thetotal totalnumber numberofofvenues venuesininthe theGangnam Gangnam test area 1570 and number In test area is is 1570 and thethe number of of confirmed venue buildings and matched road segments to the buildings is 425 and 242, confirmed venue buildings and matched road segments to the buildings is 425 and 242, respectively. respectively. Theplot normal Q-Qthat plotthe indicates that the distribution of ln of restaurant venues in The normal Q-Q indicates distribution of lnpH i q of restaurant venues in the Gangnam area theslightly Gangnam area isskewed slightly(skewness negativelyofskewed (skewness −0.61)12a. as As shown in Figure 12a. As is negatively ´0.61) as shown inof Figure previously mentioned previously mentioned though, a strict following of the normal distribution is not critical since Getis˚ though, a strict following of the normal distribution is not critical since Getis-Ord’s Gi does not ∗ Ord’s does not provide clear equations for the calculation of the variance and the expected value provide clear equations for the calculation of the variance and the expected value [17]. The detection [17]. detection of the themed clusters from the test area results the Gangnam area are of theThe themed street clusters from thestreet test results of the Gangnam are of illustrated in Figure 12b. illustrated in Figure 12b. Three restaurant themed street clusters are detected in the Gangnam area. The special aspect of the area is that districts A and B form the back alleys of Gangnam-daero (the widest road shown in Figure 12b). It was reported that many restaurants and pubs are densely located in the back alleys of the Gangname-daero [20]. Considering the study, it can be concluded that the test result has objective accuracy. Unlike districts A and B, district C reveals a limitation of the test method. In fact, district C is not a cluster, but it is detected as only a single hot street (a street acting as a hot spot) segment. Only one restaurant venue building was assigned to the target road segment, and the z score of Gi˚ was 1.98, with a significance level of 0.05, whose values are within the average range of the study areas. However, the popularity value pPi q was 1045. This value is extremely high considering that the average Pi in the Gangnam area is around 80. This shows that an extremely high value of Pi could render a road segment as a themed street.
ln of restaurant venues in the Gangnam area is slightly negatively skewed (skewness of −0.61) as shown in Figure 12a. As previously mentioned though, a strict following of the normal distribution is not critical since Getis-Ord’s ∗ d. The detection of the themed street clusters from the test of2016, the 5, Gangnam area are illustrated in Figure 12b. ISPRS Int.results J. Geo-Inf. 111 12 of 15
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
12 of 15
(a)
(b)
Figure 12. Gangnam area test results: (a) Q-Q plot for normal; (b) Restaurant themed streets.
Three restaurant themed street clusters are detected in the Gangnam area. The special aspect of the area is that districts A and B form the back alleys of Gangnam-daero (the widest road shown in Figure 12b). It was reported that many restaurants and pubs are densely located in the back alleys of the Gangname-daero [20]. Considering the study, it can be concluded that the test result has objective accuracy. Unlike districts A and B, district C reveals a limitation of the test method. In fact, district C is not a cluster, but it is detected as only a single hot street (a street acting as a hot spot) segment. Only one restaurant venue building was assigned to the target road segment, and the z score of ∗ was 1.98, with a significance level of 0.05, whose values are within the average range of the study areas. However, the popularity value was 1045. This value is extremely (a) (b) high considering that the average in the Gangnam area is around 80. This shows that an extremely high value of Figure Gangnam area results: (a) Figure 12. Gangnam area test results:street. (a) Q-Q Q-Q plot plot for for normal; normal; (b) (b) Restaurant Restaurant themed themed streets. streets. could render a 12. road segment as test a themed Similar to the Sillim test area, the café (Figure 13a), fashion (Figure 13b), and pub (Figure 13b) Similar to the Sillim test area, the café (Figure 13a), fashion (Figure 13b), and pub (Figure 13b) themes are tested. themes are tested.
(a)
(b)
(c)
Figure 13. 13. Themed Themed street street cluster cluster detection detection results results in in Gangnam: Gangnam: (a) Café themed themed street; street; (b) (b) Fashion Fashion Figure (a) Café themed street; (c) Pub themed street. themed street; (c) Pub themed street.
The G˚∗ z-score z-scoreand andp-value p-valueofofeach eachthemes themesare areplotted plottedininFigure Figure14. 14. The i The SBDC market analysis report of Gangnam was presented in May 2008, in which all of the themed street clusters in the Gangnam area were visually compared. Figure 15 shows a map of the trade area from the market analysis report (Figure 15a) and the themed street clusters along with buildings that have venue data (Figure 15b). Excluding district C in Figure 12b, which as explained above, reveals the limitation of the testing method, the SBDC market analysis report shows that most of the themed street clusters are located inside of the area. This demonstrates that the test method has reliability and versatility.
(a)
(b)
(a)
(b)
ISPRS Int. J. Geo-Inf. 2016, 5, 111
(c)
ISPRS Int. J. Geo-Inf. 2016, 5, x FOR PEER
(c) (a)
13 of 15
13 of 15
(d)
(b)
Figure 14. ∗ z-score and p-value of road segments of each themes in Gangnam test area: (a) restaurant; (b) café; (c) fashion; (d) pub.
The SBDC market analysis report of Gangnam was presented in May 2008, in which all of the themed street clusters in the Gangnam area were visually compared. Figure 15 shows a map of the trade area from the market analysis report (Figure 15a) and the themed street clusters along with buildings that have venue data (Figure 15b). Excluding district C in Figure 12b, which as explained (d) above, reveals the limitation(c) of the testing method, the SBDC market analysis report shows that most of the themed street clusters are located inside of the area. This demonstrates that method has Figure 13. ∗ z-score and p-value of road segments of each themes in Gangnam the testtest area: (a) Figure 14. Gi˚ z-score and p-value of road segments of each themes in Gangnam test area: (a) restaurant; reliability and versatility. restaurant; (b) café; (c) fashion; (d) pub. (b) café; (c) fashion; (d) pub.
(a)
(b)
Figure 15. 15. Visual Visual assessment assessment test: test: (a) (a) trade trade area area from from the the SBDC SBDC market market analysis analysis report; report; (b) (b) trade trade areas areas Figure from the SBDC (red line) and sum of all the themed street clusters (dashed black line). Blue polygons from the SBDC (red line) and sum of all the themed street clusters (dashed black line). Blue polygons are buildings buildings that that have havevenues venuesdata. data. are
5. Conclusions 5. Conclusions The TSCM TSCM has has been been suggested suggested throughout throughout this this study study to to detect detect themed themed streets. streets. Themed Themed streets streets The encourage local economy and provide social and cultural spaces. The TSCM is not limited to merely encourage local economy and provide social and cultural spaces. The TSCM is not limited to merely detecting areas areas where wheresimilar similar stores storesare are densely densely located, located, but but also also detects detects various various themed themed streets streets using using detecting LBSM data, data, providing providing rich richinformation. information. LBSM Comparingthe thetrade tradeareas areasfrom fromthe the market analysis report prepared from a field survey Comparing market analysis report prepared from a field survey withwith the the test results of this study, the reliability ofmethod the method was confirmed. test results of this study, the reliability of the was confirmed. The most most significant significant contribution contribution of of this this study study is is that that various various themes themes were were detected detected using using the the The suggested TSCM and that the method produced consistent results. Also, through the identification suggested TSCM and that the method produced consistent results. Also, through the identification of themed streets, local governments may be able to understand the socio-dynamics of target areas, or they may engage in the re-development of the existing town planning [21]. Using this method, planners can receive up-to-date data regarding the specific usage of urban areas without the need for field surveys or the risk of outdated literature investigations, and budgetary spending is consequently reduced. Besides, like the finding of this study whereby the “Fashion and Cultural Street” in Sillim test area is used for different purposes, planners can identify the way that official
ISPRS Int. J. Geo-Inf. 2016, 5, 111
14 of 15
of themed streets, local governments may be able to understand the socio-dynamics of target areas, or they may engage in the re-development of the existing town planning [21]. Using this method, planners can receive up-to-date data regarding the specific usage of urban areas without the need for field surveys or the risk of outdated literature investigations, and budgetary spending is consequently reduced. Besides, like the finding of this study whereby the “Fashion and Cultural Street” in Sillim test area is used for different purposes, planners can identify the way that official planning outcomes have been transformed. With these perspectives, an enhanced decision making process can be used for re-development of urban areas. Furthermore, the TSCM produced objective results irrespective of the location of the test area, since the method converts LBSM data attributes to mathematical values. Also, using road segments as basic spatial units for analysis, the test results were intuitive and distinguishable by avoiding formation of ambiguous spatial areas such as those created by KDE or conventional hot spot areas. Finally, not only the number of POIs and check-in counts, but also other attributes were mathematically measured to find meaningful themed streets. However, this study also has some limitations. First, a quantitative measurement to evaluate the themed street clusters was not carried out. Second, the age range of people using the Foursquare service is limited. Although this is a general problem of the analyses for which LBSM data are used, additional data from different sources can be used to represent the whole population. Third, human error and participation leads to mistakes like typos and subjective opinions of venues when LBSMs are used. Also, internet access and other types of physical limitations could lead to the ununiformed distribution of LBSM data. Such limitations should be known by those researchers who utilize LBSM data. Lastly, a number stores or buildings (venues), such as large malls or well-known restaurants, can skew the detection of themed streets if their popularity levels are extremely high. To overcome the abovementioned limits, an evaluation method needs to be studied. Also, errors caused by a few venues should be avoided and continuously identified and investigated by modifying the hot value (Hi ), which reflects the characteristics of road segments. Acknowledgments: This research was supported by a grant (15CHUD-C061156-05) from the National Spatial Information Research Program funded by the Ministry of Land, Infrastructure and Transport of the Korean government. Author Contributions: Byeongsuk Ji, Youngmin Lee and Pil Kwon provided the core idea for this study. Byeongsuk Ji implemented the methodology and carried out the experimental validation. Kiyun Yu and Pil Kwon wrote the main manuscript. Conflicts of Interest: The authors declare no conflicts of interest.
Abbreviations The following abbreviations are used in this manuscript: LBSM VGI POI NT-STAC NT-SaTScan NKDE KDE
Location Based Social Media Volunteered Geographic Information Point of Interest Network Spatial and Temporal Analysis of Crime Network Spatial Scan Statistic (Network Kernel Density Estimation) (Kernel Density Estimation)
References 1. 2. 3.
Park, C.-B. A study on the revitalization of specialized street in a viewpoint of downtown regeneration in physical environment. J. Archit. Inst. Korea Plan. Des. 2009, 25, 285–292. Jacobs, J. The Death and Life of Great American Cities; Vintage: New York, NY, USA, 1961. Cranshaw, J.; Schwartz, R.; Hong, J.sI.; Sadeh, N. The livehoods project: Utilizing social media to understand the dynamics of a city. In Proceedings of the International AAAI Conference on Weblogs and Social Media, Dublin, Ireland, 5 June 2012; p. 58.
ISPRS Int. J. Geo-Inf. 2016, 5, 111
4.
5. 6. 7. 8. 9. 10.
11. 12. 13.
14. 15. 16. 17. 18. 19. 20. 21.
15 of 15
Wang, M.; Qin, L.; Hu, Q. Data mining and visualization research of check-in data. In Proceedings of the 2012 20th International Conference on Geoinformatics (GEOINFORMATICS), Hong Kong, China, 15–17 June 2012; pp. 1–4. Xie, Z.; Yan, J. Kernel density estimation of traffic accidents in a network space. Comput. Environ. Urban Syst. 2008, 32, 396–406. [CrossRef] Okabe, A.; Satoh, T.; Sugihara, K. A kernel density estimation method for networks, its computational method and a GIS-based tool. Int. J. Geogr. Inf. Sci. 2009, 23, 7–32. [CrossRef] Mohaymany, A.S.; Shahri, M.; Mirbagheri, B. GIS-based method for detecting high-crash-risk road segments using network kernel density estimation. Geo-Spat. Inf. Sci. 2013, 16, 113–119. [CrossRef] Shiode, S. Street-level spatial scan statistic and stac for analysing street crime concentrations. Trans. GIS 2011, 15, 365–383. [CrossRef] Lu, Y. Approaches for Cluster Analysis of Activity Locations Along Streets: From Euclidean Plane to Street Network Space; Texas State University: San Marcos, TX, USA, 2005. Li, Y.; Steiner, M.; Wang, L.; Zhang, Z.-L.; Bao, J. Exploring venue popularity in foursquare. In Proceedings of the 2013 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Turin, Italy, 14–19 April 2013; pp. 3357–3362. Goodchild, M.F. Citizens as sensors: The world of volunteered geography. GeoJournal 2007, 69, 211–221. [CrossRef] Goodchild, M.F.; Li, L. Assuring the quality of volunteered geographic information. Spat. Stat. 2012, 1, 110–120. [CrossRef] Paek, J.; Kim, J.; Govindan, R. Energy-efficient rate-adaptive GPS-based positioning for smartphones. In Proceedings of the 8th International Conference on Mobile Systems, Applications, and Services, San Francisco, CA, USA, 15–18 June 2010; pp. 299–314. Benedikt, M.L. To take hold of space: Isovists and isovist fields. Environ. Plan. B Plan. Des. 1979, 6, 47–65. [CrossRef] Anselin, L. Local indicators of spatial association—Lisa. Geogr. Anal. 1995, 27, 93–115. [CrossRef] Lee, S.-I.; Cho, D.; Sohn, H.; Chae, M. A GIS-based method for delineating spatial clusters: A modified amoeba technique. Korean Geogr. Soc. 2010, 45, 502–520. Aldstadt, J.; Getis, A. Using amoeba to create a spatial weights matrix and identify spatial clusters. Geogr. Anal. 2006, 38, 327–343. [CrossRef] Getis, A.; Ord, J.K. Local spatial statistics: An overview. Spat. Anal. Model. GIS Environ. 1996, 374, 269–285. Jeong, H. Wasting the Tax-Payer’s Money, "Fashion and Cultural Street“. Available online: http://www.ytn.co.kr/ _ln/0103_200610020814127112 (accessed on 11 July 2016). Kwak, K.S. A study on trade city area analysis: Focused on the trade city areas of Kangnam, Seocho; The Graduate School of Business Administration, Kwangwoon University: Seoul, Korea, 2010. Rösler, R.; Liebig, T. Using data from location based social networks for urban activity clustering. In Geographic Information Science at the Heart of Europe; Springer: Cham, Switzerland, 2013; pp. 55–72. © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
