12
Dr Weisi Guo Assistant Professor School of Engineering Warwick Institute for the Science of Cities (WISC) University of Warwick, UK Social Media Data for Planning and Monitoring Services Exchange Assistant Professor Centre for Urban Science and Progress New York University, USA Visiting Professor SCIE Shanghai University, China

Dr Weisi Guo, University of Warwick

Embed Size (px)

Citation preview

Page 1: Dr Weisi Guo, University of Warwick

Dr Weisi Guo

Assistant Professor School of Engineering

Warwick Institute for the Science of Cities (WISC)

University of Warwick, UK

Social Media Data for Planning and

Monitoring Services

Exchange Assistant Professor Centre for Urban Science and Progress

New York University, USA

Visiting Professor SCIE

Shanghai University, China

Page 2: Dr Weisi Guo, University of Warwick

School of Engineering | Warwick Institute for the Science of Cities

A bit about me

Brief Bio: I graduated with MEng, MA, and PhD degrees in

information engineering and computer science from the

University of Cambridge.

I am currently the joint coordinator in Smart City research theme

at the School of Engineering. I have worked in academia and

industry for over 7 years.

I currently run a research team (3 doctoral and 4 graduate

researchers) working at the inter-section of big data, wireless

networks and smart cities. I want to design solutions that can

integrate big data analytics into traditional ICT systems.

Awards in 2014/15:

IET Innovation Award 2015: Communications Category

Bell Labs Prize Finalist 2014 (only UK recipient)

IEEE Best Paper Award 2014

IEEE Communication Society 2014 Best Project 2nd Prize

Page 3: Dr Weisi Guo, University of Warwick

Activities at the University of Warwick: WISC &

• Warwick is home to the only UK government funded Doctoral

Training Centre in Smart Cities (training 50-75 PhD students

2014-2023). The students combine research skills in big data,

urban planning, engineering, and social sciences. The centre

is called Warwick Institute for Science of Cities (WISC).

• Warwick is also part of a global 5 university alliance on smart

city research: New York University, Carnegie Mellon, Toronto

University, and IIT-Mumbai. The headquarters is called CUSP

(Centre for Urban Science & Progress), funded by ex-NYC

mayor: Michael Bloomberg.

• CUSP is opening its 1st overseas expansion campus in

London which sees Warwick and KCL join forces to examine

the challenges related to health and big data in cities.

• Warwick is also a core partner in the new big data Alan

Turing Institute.

Page 4: Dr Weisi Guo, University of Warwick

Why Cities

• Cities are permanent human settlements with a history of almost 10,000 years.

Typical attributes: high population density, specialist economy, public

infrastructure, strong local governance, high import & export volumes.

[Ur City (modern Basra) – 3800 BC]

• Cities occupy 2% of land surface, but account for up to 60-80% of the global

energy consumption.

• In the past decade, first time in history that more than 50% of the world’s

population live in cities. In developed nations, this value is between 70 to 95%.

A third of the most densely populated cities are in the developed world.

[Population density in Paris is comparable to density in Delhi]

• According to the United Nations (2012 Habitat Report), more than 70% of the

world will live in a city by 2050.

• What are the metrics that gauge a city’s performance?

Page 5: Dr Weisi Guo, University of Warwick

Activities at the University of Warwick: HAT

• The United Nations has published a set of 5

metrics to gauge the performance of cities:

Productivity, Infrastructure, Quality of Life,

Equity (Equality), the Environment.

• Global rankings of cities use metrics such as:

Connectivity, Competitiveness, Power, and

Influence.

• Quality of Living rankings of cities use metrics

such as: Environment, Safety, Public Services

and Stability.

• Such metrics are seen as complex indicators to the

performance of cities in competing for human and

material resources.

Top Global Cities: New York and London

Top Quality of Life Cities: Vienna and Zurich

Page 6: Dr Weisi Guo, University of Warwick

Scaling Law of Cities

• Cities grow like organisms, and as they grow in

size, they also experience more problems and

convey more benefits.

• The scaling law of problems and benefits is of

interest to us, as many of our cities are growing

in size, whilst some (i.e., Rust Belt of USA) are

shrinking rapidly (20% loss in population in

recent years).

• Research has shown that whilst mammals

experience a sub-linear growth (everything

gets less efficient per kg of weight), cities

experience super-linear growth (everything

gets more per capita).

Page 7: Dr Weisi Guo, University of Warwick

Challenges Faced by Cities

• Cities face ancient and new challenges, but never

has the scale of the problem been so big, and never

have we been in a better position to use technology

to solve them.

• Examples of universal challenges include: pollution

(air and water), traffic congestion (inter- and intra-

city), crime, energy efficiency, public order, balance

between green space and commerce, acoustic

noise (highest complaint in NYC), and shocks in

temperature (heat is the highest killer in NYC).

• What we have to foster is to allow cities to grow in a

sustainable and prosperous way (i.e., growth of

benefits outweigh problems), otherwise some of our

cities may one day be a historical landmark.

Page 8: Dr Weisi Guo, University of Warwick

School of Engineering | Warwick Institute for the Science of Cities

How can social media data act as a senor and help us understand cities and services?

• High Resolution: in the past 5 years, the growing penetration of

smartphone usage and social media usage has led to a wealth of

data across a wide range of hardware and application orientated

research. In particular, social media offers high resolution

compared to survey/census approaches:

- Spatial Resolution: wireless assisted GPS (~< 10 metres)

- Time Resolution: seconds

- Scale: Twitter has 316 million users with 500 million

messages/day

• Detailed Context: Not only is the quantifiable data of interest, but

the unstructured text and multimedia data is also of interest.

- Text: what are people saying / feeling and how does

information spread

- Community: how do people connect and follow each other

- Habits: what do people do and what behavioural patterns

emerge

Page 9: Dr Weisi Guo, University of Warwick

School of Engineering | Warwick Institute for the Science of Cities

Sentiment Mapping of Services

• Sentiment: natural language processing words and

phrases into sentiments

- Real time mapping of emotions on individual and

regional level

- Identify areas of sadness and correlate it to real

challenges in business and services for targeted

prioritised intervention

• Case Study of London: converted 600,000 tweets into

geo-tagged sentiments

- Blue = Sadness

- Red = Happiness

Unhappiest Wards: Barking, Newham

Happiest Wards: Westminster, Hillingdon, Camden

Page 10: Dr Weisi Guo, University of Warwick

School of Engineering | Warwick Institute for the Science of Cities

Creating Networks from Data

• Relationship between Stakeholders: it is important to

analyse the relationship between stakeholders, rather than

treat them in isolation.

• Complex Network: As an example, we model short-ranged

trade network across Europe to reveal the following attributes:

- Areas of redundancy (benefits: robustness against

failure, cons: inefficiency)

- Areas of strategic importance / influence or areas of

vulnerability

- How the network can improve or adapt subject to a

constraint or a perceived threat

This analysis can be adapted to small-scale networks (within a

company) or large-scale multi-level systems (i.e., transport

network within a city or country).

Page 11: Dr Weisi Guo, University of Warwick

(b) No. of Connecting Links

A

C

B

(g) Modularity Class

1

2

3

(c) Average Link Distance

D E

F

(d) Cluster Coefficient

F

G

H

I

J

(a) No. of Critical Links

A B

School of Engineering | Warwick Institute for the Science of Cities

Analysing a Real Trade Network

Critical

Nodes

Importance

A Critical & Important Links

Connections

Influential

B Critical & Important Paths

Connections

C Important Paths

Connections

Influential

F Central

Cluster

(e) Influence

A

C

(f) Page Rank

A

C

B

Page 12: Dr Weisi Guo, University of Warwick

Dr Weisi Guo

[email protected] School of Engineering

Warwick Institute for the Science of Cities (WISC)

University of Warwick, UK

Thank you for Listening