Mapping Protest in 3D with Twitter Data

As one part of my docotoral thesis, I have made the video that shows the relationship between ‘London End Austerity Now’ Protest on 20^th June 2015 and the Twitter acitivity on that day.

The video gives you some details about the protest, the data and 3D visualisation.

If the following YouTube video is not displayed on your device, please use this link. 

Mapping London’s Twitter Activity in 3d

Image 1. The tweet density from 8am to 4pm on 20th June 2015, Central London

Twitter Mapping is increasingly useful method to link virtual activities and geographical space. Geo-tagged data attached to tweets containing the users’ location where they tweeted and it can visualise the locations of users on the map. Although the number of the geo-tagged tweets is a relatively small portion of all tweets, we can figure out the density, spatial patterns and other invisible relationships between online and offline.

Recently, studies with geo-tagged tweets have been developed to analyse the public response to specific urban events, natural disasters and regional characteristics (Li et al., 2013) [1].  Furthermore, it is extending to traditional urban research topics, for example, revealing spatial segregation and inequality in cities (Shelton et al., 2015) [2].

 

Twitter mapping in 3D can augment 2d visualisation by providing built environment contexts and improved information. There are many examples of Twitter mapping in 3d such as A) #interactive/Andes [3] , B) London’s Twitter Island [4], C) Mapping London in real time, using Tweets [5]. A) and B) build up 3d mountains of the geo-tagged tweet on the map.  In the case of C), when the geo-tagged tweets are sent in the city, the heights of nearest buildings increase in the 3d model. These examples are creative and show different ways to view the integrated environments.

From a Networking City’s view, if we make a Twitter visualisation more tangible in a 3d urban model, it would help us to have a better understanding how urban environments are interconnected with the invisible media flow.

 

To make the visualisation, the Twitter data has been collected by using Big Data Toolkit developed by Steven Gray at CASA, UCL. All 53,750 geo-tagged tweets are collected on 20^th June, 2015 across the UK. As we can see from Table 1, the number of tweets was at the lowest point at 5am and reached to the highest point at 10pm with 3495 tweets. Moreover, Video 1 shows the location of the data in the UK and London on that day in real time.

 

Table 1. The Number of Geo-Coded Tweets in the UK on 20th June, 2015

 



Video 1. The location of Geo-Coded Tweets in the UK on 20th June, 2015

When we calculate the density of the data, London, particularly Central London, contains the largest number of the tweets. (Image 2)

 

 

 

Image 2. The density of Geo-Coded Tweets in the UK on 20th June, 2015

In order to focus on the high density data, 6 km x 3.5 km area of Central London is chosen for the 3d model. Buildings, bridges, roads and other natural environments of the part of London have been set in the model based on OS Building Heights data[6]. Some Google 3d warehouse buildings are added to represent important landmark buildings like St.Pauls, London Eye and Tower Bridge as you can see from Image 3, Image 4 and Image 5.

 

 

Image 3. The plan view of Central London model

Image 4. The perspective view of Central London model

Image 5. The perspective view of Central London model (view from BT Tower)

The geo-tagged data set is divided into one hour periods and distributed on the map to identify the tweet density in the area. Through this process, we can see how the density is changing depending on the time period. For example, the tweets are mainly concentrated around Piccadilly Circus and Trafalgar Square between 10am and 11am, but  there are two high-density areas between 12pm and 1pm (See Image 6, Image 7, Image 8 and Image 9)

Image 6. The tweet density between 10am and 11am on 20^th June 2015

Image 7. The tweet density between 12pm and 1pm on 20^th June 2015

Image 8. The tweet density from 12am to 12pm

Image 9. The tweet density from 12pm to Midnight

 

 

As we’ve seen above, the 2d mapping is useful to understand the relative density in one period such as which area is high and which area is low between 12pm and 1pm. However, we cannot understand the degree of intensity in the highest peak areas. It is believed that 3d mapping is needed at this stage. We can clearly see the density of the tweet data in each period and the intensity of the tweet density across the time periods from Image 10 to Image 14.

West End area shows high density throughout the whole day but City area shows the peak only during lunch time. This pattern likely relates to the activities of office workers in City and leisure/tourist in West End.

Image 10. The tweet density in 3d between 10am and 11am on 20^th June 2015

Image 11. The tweet density in 3d between 12pm and 1pm on 20^th June 2015

 

Image 12. The tweet density in 3d from 12am to 8pm

Image 13. The tweet density in 3d from 8am to 4pm

Image 14. The tweet density from 4pm to Midnight

 

 

 ________________________________________

[1] Linna Li , Michael F. Goodchild & Bo Xu (2013) Spatial, temporal, and socioeconomic patterns in the use of Twitter and Flickr, Cartography and Geographic Information Science, 40:2, 61-77

 

[2] Taylor Shelton, Ate Poorthuis & Matthew Zook (2015) Social Media and the City: Rethinking Urban Socio-Spatial Inequality Using User-Generated Geographic Information, Landscape and Urban Planning (Forthcoming), http://papers.ssrn.com/abstract=2571757

 

[3] Nicolas Belmonte, #interactive/Andes,   http://twitter.github.io/interactive/andes/  (Strived on 15^th August 2015)

 

[4] Andy Hudson-Smith, London’s Twitter Island – From ArcGIS to Max to Lumion, http://www.digitalurban.org/2012/01/londons-twitter-island-from-arcgis-to.html#comment-7314
(Strived on 15^th August 2015)

 

[5] Stephan Hugel and Flora Roumpani, Mapping London in real time, using Tweets, https://www.youtube.com/watch?feature=player_embedded&v=3fk_qxGZWFQ  (Strived on 15^th August 2015)

[6] OS Building Heights-Digimap Home Page  http://digimap.edina.ac.uk/webhelp/os/data_information/os_products/os_building_heights.htm  (Strived on 15^th August 2015)

 

A sense of new academic term

 

New academic term of UCL officially started on 22nd September 2014. Thousands of new students walk around the campus, and they fill Bloomsbury area with vibrant energy. Dr. Adam Dennett begun his GI Systems and Science class for new post-graduate students of CASA with welcoming slide on 1st October. In this year, CASA opened two new post-graduate courses: MSc SmartCities and Urban Analytics and MRes Smart Cities. Therefore, he has developed a lot of the course materials and structure for the students during last summer days.

 

On the same day, when CASA held SHOW AND TELL, which is a traditional CASA event to introduce each other, I apparently realised that the new term is just started! Most members of the lab came up and introduced themselves at this inaugural meeting.  

 

Emer Coleman’s seminar was followed on 7^th October under the title of “Open Data and the City: Looking back and Looking Forward”. She explained open data as a way of engagement and empowerment and how citizen can be benefited and can participate in making better urban environment. Several good cases, such as Hello Bristol, were mentioned.

After the presentation, many questions were emerging from the audience. Transparency, security, effectiveness and so on. However, I was uncomfortable when she criticised, with some sentences from Adam Greenfield’s “Against the smart city”, big corporations that IBM and Cisco have been pushing smart city idea for money rather than people or better society. I could not catch the difference between the big brands, which get profit by providing new city systems and solutions, and her company, which get profit as well by providing efficient transport solution and application. There might be a matter of size.

These adventures would be enough to feel a sense of the new term. However, UCL email was unusually hacked on 9^th October. All UCL students got 3000 emails (including me) with bello. It was a big issues not only in the campus but also in the UK as The Independentannounced. Steven Gray, a specialist of large datasets at CASA, analysed what has happened with his Big Data Toolkit and posted it on his blog.
 

Luminous Cities: offering an alternative way of geotag

Image1. The webpage of Luminous Cities_Manhattan

Studying human behaviours and communication in time and space has been regarded as the important factor of modern urban planning. In this digital era, collecting online data and analysing the data provide an opportunity to understand the intention and the process of the behaviours and the communication which had not been revealed.

Geotag, which is attached on Social Network Service (SNS), is concerned as one of connecting link between the internet and urban. Mainly, there are two types of geotag. One is user-generated geotag that SNS users identify the places on their contents. The other is automatically generated with spatial coordination by the services. It represents the political, social and economic characteristics of the places as well as the physical location of the user or the data produced.

There are many good examples of mapping the geotag data of SNS. Eric Fischer’s well known mapping images reveal not only the density of the geotag data but also social aspects in cities such as the invisible dimensions of tourism in New York (Image 2). Twitter Languages in London by James Cheshire and Ed Manley shows the popularity of languages depends on different locations in London ((Image 3).

Image2. The mapping geotag data of locals and tourists by Eric Fischer 

Image3. Twitter Languages in London, James Cheshire and Ed Manley

Luminous Cities is the project to demonstrate the interactive map of Flickr geotag data supported by CASA at UCL and CSAP at the University of Leeds. It has developed by Gavin Baily and Sarah Bagshaw. The project does not remain the displaying density and distribution of the geotag, but offers in-detail contents of the geotag such as user, tag, time of the day and timeline over 50 cities in the world. With the multiple contents, Luminous Cities could be a platform to check out the geotag data of Flickr based on personal interest, and to view their cities from a different side. When it comes to Networking City, who is interested in protest and demonstration in the city, it would be a helpful tool to examine the relationship between protests or occupy tags of Flickr in London and actual events of them. Also, some interesting results may be emerging when we compare two data sets: Flickr and Twitter.

Image4. Berlin user geotag map from the webpage of Luminous Cities

Image5. London occupy geotag map from the webpage of Luminous Cities

Image6. Tokyo geotag map, Zoom out, from the webpage of Luminous Cities

Image7. Tokyo geotag map, Zoom in, from the webpage of Luminous Cities

You can find more things from following links.

Flickr was shown as the highest growing application in 2013 by Mashable

http://mashable.com/2013/10/21/fastest-growing-apps/?utm_cid=mash-com-fb-main-link

Mapping the world with Flickr and Twitter by Guardian

http://www.theguardian.com/technology/pda/2011/jul/13/flickr-twitter-maps

Infographic Of The Day: Using Twitter And Flickr Geotags To Map The World

http://www.fastcodesign.com/1664462/infographic-of-the-day-using-twitter-and-flickr-geotags-to-map-the-world

GIS Course Note 02: Research Application, Software and Data Sources

The second lecture of GIS comprised mainly three parts, the examples of practical research by using GIS, GIS software and the way to gain relevant data for the research.  

In the beginning, Dr. Adam Dennett, the lecturer of CASA, informed the aim of the lecture to understand the basic elements of social science research using GIS and the diverse analytical approaches with it. He showed several example maps, which are related to population, crime, deprivation, health care, flooding, and education, and the way how to read economic, social and physical characteristics from the maps and its meaning in the projects. (Image 1)



Image 1


And then, he moved to GIS software industry which has been significantly growing. As interest and the utilisation of GIS are increasing, GIS software market is expanding almost 10% every year and now it is used in all industries and public sectors such as business, public safety, military and education. The popular GIS tools: Arc GIS, MAP Info, Quantum GIS, Python and R, and specific points of each tool were introduced. Also, small description of GIS cloud and online GIS tools was following. (Image 2)

Image 2

In the last part, he said of various kinds of the data and the way of gathering the data which is the key element to proceed the research. Easily, we can classify the data according to the way of gathering. On the one hand, we can use the open data, which are provided by public sectors and other organisations. On the other hand, we need to collect the data through participation and measuring by ourselves. Some websites of the UK, which contain the open data or shapefiles, and the characteristics of each website were mentioned. As we can see Image 3, some other methods like WebScarping and Volunteered Geographic Information were shown as alternative ways to collect the data by ourselves, when the given data are unclear, and the goal of the research needs the specific data.

Image 3

 

The lecture was finished with the emphasis on caution when using the open data and the mapping with it. Much of the data are made with inadequate formats like pdf, or do not include any spatial reference, so we need to be careful to collect and use the data. When it came to the mapping with the data, he insisted that it is necessary to make analytical and meaningful maps rather than something fancy or colourful. In addition, it is essential to acknowledge that some errors could be made by way of ‘generalisation’ in the process of research, therefore, setting up the range and the level of the research will enrich the quality of it.

 

After one hour lecture, students had a training session that mapping population data on the map of London Borough with R. (Image 4)

Image 4

 

GIS Course Note 01: Spatial is Special



Image 1. Dr.Adam Dennett introduced the course outline on 2nd October, 2013

From this academic term, Networking City is doing a teaching assistant role for ‘GEOGRAPHIC INFORMATION SYSTEMS AND SCIENCE’ course which is set up by CASA for their provocative master programme ‘MResAdvanced Spatial Analysis & Visualisation’ and Bartlett students. In this year, the course is opened to Urban Planning and DPU students of Bartlett, so thirty students registered, while fifteen students who were mostly from the CASA had an opportunity last year.

Dr. Adam Dennett, the lecturer, briefly showed the outline of the course and explained the meaning of studying spatial analysis, definition of Geographic Information System, linkage between GIS and scientific research, the difference between GISystems and GIScience, and short history of GIS.

During one hour his lecture, the most impressive part was what the meaning of information is in Geography and Urban studies, and how it can make an impact on decision making process. When he illustrated the structure of how one spatial data could be developed to information, knowledge and wisdom, and could be the initial point which change our environments, he emphasised not to make a graphic image by GIS programmes but to consider the meaning behind the data.

After the lecture, the students had two-hour practical session. They operated the main programmes of the course: Arc-GIS, QGIS and R on UCL computers, and checked how they can set up the programmes on their own laptops. In order to learn basic knowledge and functions of Arc-GIS, Adam recommended registering My Virtual CampusTraining on ESRI homepage and complete its modules.

 

CASA Seminar: Parameterizing land use planning by Talia Kaufmann



Image1.Talia Kaufmann gave the presentation at CASA Lunchtime Seminar. The image was taken by Networking City.

 On 10^th July 11, 2013, CASA Lunch Seminar was commenced in The Rockefeller Building. Talia Kaufmann, who is a visiting student from MIT Master in City Planning, provided her current research ‘Parameterizing land use planning’. After she finished her bachelor degree in Architecture Tel-Aviv University, Israel and worked as an urban planner for the Tel Aviv-Yafo City Planning Department.

Throughout the presentation, she showed brilliant ideas and cool images. Also, some arguable points were emerging in her presentation, and the considerable discussion between audiences and Talia was continued.

For example, one of her main idea is to randomly capture Google Street View images from several targeting cities and ask people to answer their sense of images. And it becomes the score of each city. The audiences questioned about the precondition of the idea. Can we measure the feeling of the city by just watching some images? Would there be the gap between the feeling of Street View images and the true feeling of cities? Is there any delicate setting for cultural backgrounds and individual subjectivity of participants?

This seminar could be a good opportunity for her to listen others’ comments, articulate the ideas and develop many possibilities of her research.

 

 

Lauréat Prix International de Géographie Vautrin Lud

The founder and chair of CASA, Prof. Michael Batty has been awarded the prestigious Lauréat Prix International de Géographie Vautrin Lud, the highest award that can be gained in the field of geography. This prize is widely known as the “Nobel prize of geography”. From 1991, only 22 remarkable researchers were awarded this prize such as David Harvey, Yi-Fu Tuan and Sir Peter Hall.

He has been running his own blog, A Science of Cities, which provides valuable theories, papers and presentation files. Also, he launched a special online class to introduce his complexity theory and methods within Santa Fe Complexity Explorer teaching site. This site is systematically organised with good materials and should be helpful to those who want to learn more about complexity theory and application methods.