Visualizing the gay community in Beijing with location-based social media

Full text

Featured graphic
Visualizing the gay
community in Beijing with
location-based social media
Bo Zhao
Oregon State University, USA
Daniel Z Sui
The Ohio State University, USA
Zhaohui Li
The George Washington University, USA
Compared with the growing worldwide social acceptance of the lesbian, gay, bisexual, and
transgender (LGBT) community (Brown and Knopp, 2008; Ferreira and Salvador, 2015;
Gates and Ost, 2004), discussions of LGBT-related matters are still a taboo in China. To
raise public awareness of this community, we aim to estimate the distribution of gay people
in Beijing using location-based social media (LBSM) feeds. To do that, we collected the gay
population data from a popular LBSM app – Jack’d. As an app dedicated to gay social
networking, Jack’d primarily attracts the use of gay people. A user who launches the
location-based function will leave a geolocation (in the format of a pair of coordinates) in
the Jack’d database. On 28 September (Wednesday) 2016, we retrieved the geolocations of
active users every 6 hours from the Jack’d database via an API (Burrell et al., 2012). As a
result, we had 5209 users at 00:00, 1006 users at 06:00, 4972 users at 12:00, 5543
1
users at
18:00, and 5214 users at 00:00 of the next day. For each sample, we created a fishnet-grid
map layer through aggregating the geolocations into equal area hexagons (each edge of the
hexagon is 300 meters in length) by QGIS. By ordering these five map layers into a time
sequence, we transformed the layers into several space–time volume elements (voxels) by
Voxler. The voxels were rendered in a rainbow-like color ramp using a ray-casting algorithm
(Roth, 1982). This algorithm simulates the mechanism of a ray travelling from the eye of an
observer to the observing object. Here, an observing object is modeled by a system of
numerous tiny points. The ray, cast from the eye of the observer, travels through the
points. Only the points along the rays are visualized. This algorithm can represent a much
clearer texture of a voxel (Hoang et al., 2016). Moreover, in order to illustrate the core
portion of the gay community, we visualized a set of net surfaces made by points of an equal
value (seven users per sampling unit in space–time). And, a base layer of Beijing city (created
in QGIS) was overlaid on the bottom of the voxels; the axes and labels were post-processed
in Adobe illustrator.
Corresponding author:
Bo Zhao, College of Earth, Ocean, Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA.
Email: bzhao@coas.oregonstate.edu
Environment and Planning A
2017, Vol. 49(5) 977–979
!The Author(s) 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0308518X16685885
journals.sagepub.com/home/epn

As shown in the graphic (Figure 1), the space spreads out along the x and y axis, and the
time sequence is indicated by the z axis. The area most populated by gay people is the
Chaoyang district. With numerous gay bars, Chaoyang is acknowledged as the recreation
center of LGBT people. Communication University of China (CUC) is located to the East,
and many gay students live there, many of whom are relatively active at night. This contrasts
with the Haidian district, which also lies with several universities (e.g., Tsinghua Univ.,
Peking Univ., Beihang, etc.), where the gay people are more active during school hours.
Several small clusters are found around Xidan (a major shopping area) during business
hours and in the Beijing Airport during the flight rush hour. This map also reveals a hot
spot of gay activities in the Tongzhou district at night – this may be partially resulted from
relatively low-rent housing in Tongzhou and its convenient accessibility to Chaoyang
for hangouts.
By geovisualizing a LBSM feed with ray-casting rendered volume model and net surfaces,
this graphic, for the first time, depicts the spatial-temporal dynamics of gay community in a
metropolitan area of China. Though ray-casting rendering is widely adopted for medical
(Moltz et al., 2009), engineering (Ray et al., 1999), and environmental uses (Dick et al., 2009;
Zhang et al., 2016), it is seldom utilized for visualizing human activities across space and
time. The ray-casting rendered volume model effectively illustrates the texture of the spatial-
temporal distribution, and the net surface enables us to recognize the core concentration. In
addition to its graphical innovation, this figure is also a timely and geographically accurate
representation of gay activities, although LBSM feeds expose an unavoidable bias in regard
to representing the whole population. For instance, there are only a few records from
middle-aged and elderly Jack’d users. Even so, this graphic is still of great value in
discovering those gay-friendly places which are not easily identified in physical space.
Perhaps more importantly, this graphic can facilitate urban governance by optimizing the
allocation of public health resources (i.e. the access to gay condoms) and providing
accessible counseling services about sexual identity.
Figure 1. The distribution of gay community in Beijing over 28 September 2016 (authors, 2016).
978 Environment and Planning A 49(5)

Acknowledgements
For their invaluable assistance with this article, I want to thank Qingxu Huang, Jane Darbyshire, Nan
Zhao, Hui Kong, and Xining Yang.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this
article.
Note
1. A total of 5543 – the sample population at 18:00 – accounts for approximate 1.3 to 2.6% of the total
gay population in Beijing. The ratio is calculated based on (1) the statistics from the Chinese
government in 2004, 2 to 4% of the adult male population is composed of gay men. And (2) the
entire population of Beijing was 21.7 million in 2015.
References
Brown M and Knopp L (2008) Queering the map: The productive tensions of colliding epistemologies.
Annals of the Association of American Geographers 98(1): 40–58.
Burrell ER, Pines HA, Robbie E, et al. (2012) Use of the location-based social networking application
GRINDR as a recruitment tool in rectal microbicide development research. AIDS and Behavior
16(7): 1816–1820.
Dick C, Kru
¨ger J and Westermann R (2009) GPU ray-casting for scalable terrain rendering.In: The
proceedings of EUROGRAPHICS, Munich, Germany, March 30–April 3 2009.
Ferreira E and Salvador R (2015) Lesbian collaborative web mapping: Disrupting heteronormativity in
Portugal. Gender, Place & Culture 22(7): 954–970.
Gates GJ and Ost J (2004) The Gay & Lesbian Atlas. Washington, DC: The Urban Institute.
Hoang NP, Asano Y and Yoshikawa M (2016) Your neighbors are my spies: Location and other
privacy concerns in dating apps. In: The 2016 18th international conference on advanced
communication technology (ICACT), Pyeong Chang, South Korea, 31 January–3 February 2016.
Moltz JH, Bornemann L, Kuhnigk J-M, et al. (2009) Advanced segmentation techniques for lung
nodules, liver metastases, and enlarged lymph nodes in CT scans. IEEE Journal of Selected Topics in
Signal Processing 3(1): 122–134.
Ray H, Pfister H, Silver D, et al. (1999) Ray casting architectures for volume visualization. IEEE
Transactions on Visualization and Computer Graphics 5(3): 210–223.
Roth SD (1982) Ray casting for modeling solids. Computer Graphics and Image Processing 18(2):
109–144.
Zhang T, Li J, Liu Q, et al. (2016) A cloud-enabled remote visualization tool for time-varying climate
data analytics. Environmental Modelling & Software 75: 513–518.
Zhao et al. 979