Mapping Human Dynamics with Social Media for Disaster Alerts · 2015-07-09 · Mapping Human...

Mapping Human Dynamics with Social Media for Disaster Alerts

Dr. Ming-Hsiang Tsou mtsou@mail.sdsu.eduProfessor of Geography, Director of the Center for Human Dynamics in the Mobile Age, San Diego State University

Dr. Chin-Te (Calvin) Jung, Chief Data Scientist, the Center for Human Dynamics in the Mobile Age, SDSU.

What is Big Data?

Image source: http://visual.ly/big-data (definition from IBM, and WIPRO) Is this a good definition of “Big Data”?

Big Data is Human-Centered Data

• Big Data is dynamic datasets created by or derived from human activities, communications, movements, and behaviors (Tsou, 2015).

• The term, big data, should refer to big ideas, big impacts, and big changes for our society rather than only focusing on big volume.

Big Data Category (Tsou, 2015).

Social life data include popular social media services (Twitter, Flickr, Snapchat, YouTube, Foursquare, etc.), online forums, online video games, and web blogs.

Health data include electronic medical records (EMR) from hospitals and health centers, cancer registry data from state and local communities, official disease outbreak tracking and epidemiology data

Business and commercial data include credit card transactions, online business reviews (such as Yelp and Amazon reviews), supermarket membership records, shopping mall transaction records per store, credit card fraud examination data, enterprise management data, and marketing analysis data.

Transportation and traffic data include GPS tracks (from taxi, buses, Uber, bike sharing programs, and mobile phones), traffic censor data (from subways, trolleys, buses, bike lanes, highways), social media data (from check-ins, Waze, and other social media platforms), and mobile phone data (from data transmission records and cellular network data).

Scientific research data include earthquakes sensors, weather sensors, satellite images, crowd sourcing data for biodiversity research, volunteered geographic information, and census data.

Disaster Data Layer

Value of Big Data: Integration (Data Fusion)

Explore their spatiotemporal relationships in both network space and geographical space.

Image provided by Dr. Atshushi Nara (Associate Director of HDMA Center).

PlaceTime

Big Data

(information)

Geography (place and time)is the KEY for Understanding and Integrating Big Data

(Tsou and Lietner, 2013)

Human Dynamic in the Mobile Age (HDMA)

NSF project website http://socialmedia.sdsu.edu/

Two Main Goals:

1. Improve the Alert Warning Effectiveness in Multiple Social Media Channels

2. Monitor Social Media Messages and Potential Help Requests/Ground Truth Observation.

NSF IBSS Award (2014-2018)This project will enable the collaboration between SDSU and San Diego OES to build and test an multi-channel, geo-targeted disaster warning system:

1. Alert Broadcaster to effectively disseminate official alerts and warnings messages from OES staff via social media channels.

2. Social Media Monitor (SMART Dashboard) to monitor the diffusion of the official alerts in terms of mentions, re-tweets, and followers.

3. Volunteer Collaborator to ensure each OES volunteer will re-tweet the official announcement from OES. The platform will identify and recruit 1000 social media volunteers in San Diego based on their social network influences

4. Social Media Analytic Viewer (GEO Viewer) for OES staff to query specific keywords in social media and map the relevant geo-tagged social media messages.

Web-based Real-time GeoViewer Tool v.2.2 (Video demo)EC2: http://vision.sdsu.edu/ec2/geoviewer/sanDiego (Live)

New Functions and Improvement:1. Online Tutorial and YouTube video2. Add new map layers from OES and NWS CAP maps.3. Login-User labeling functions.4. Save Search result function.5. Dynamic cluster texts and hot spot radius change visualization.

Use ESRI ArcGIS Online Basemap Layers(Light Gray Map, Satellite, Street, National Geographic,

NASA Night View, Open Street Map)

How to find out critical information from thousands of GPS-tagged tweets or hundreds of thousands of Non-GPS-tagged tweets? Nepal Earthquake Example: (keyword search: “trap”)

One Possible Solution: Manual labeling (first 1000 tweets by volunteers) + Machine Learning Classification (built-in).

Need Some programming and design help from OES, RedCross, and 211:

1. How to combine multiple volunteers’ Inputs and Integration Systems (ranking system).

2. Which category and color schemes/labels should we use for each types of disasters (flooding, wildfires, earthquake, hurricanes).

3. Which tags might be useful? 4. Who are the target users? What

kinds of “Output” system should we create? (for OES staff? For RedCross staff?)

5. Other suggestions?

Digital Volunteers may help us identify and select important Tweets (for machine learning) during and after the disaster events.

SMART Dashboard for Nepal EarthquakeSocial Media Analytic and Research Testbed

http://humandynamics.sdsu.edu/NepalEarthquake.html

Design the Alert Broadcaster(to disseminate OES alerts effectively)

• Recruit 500 top influencer and 500 selected sample group representatives• Create the Alert Broadcaster for them to retweet and re-send alerts.• Analyzing the Spreads (Speed, Scale, and Range) of Social Media Messages

in Different Social Networks. (following, retweets, and mentions relationships)

@10News @SanDiegoCounty

@KPBSnews

@ReadySanDiego

@UTsandiego

Human Dynamic in the Mobile Age (HDMA)

@SanDiegoCounty Twitter Account: All Followers Analysis (31,000: 2014, 40,500: 2015)

@SanDiegoCounty Twitter Account: Newly Added Followers (350 - Changed between 5/16 to 5/18, 2014, During the Carlsbad wildfire)

395

-45

350

-100

0

100

200

300

400

500

Follo

wer

s

Change in Followers (@SanDiegoCounty - 05/16-05/17)

followGain

followLoss

followChange

Most influential followers (using followers_count) are not from San Diego

Most new followers are from San Diego.

Situation Awareness

San Diego County: Office of Emergency Services (OES)

Thank You!

Dr. Ming-Hsiang TsouSan Diego State University

mtsou@mail.sdsu.edu

Spatiotemporal Modeling of Human Dynamics Across Social Media and Social Networks Interdisciplinary Behavioral and Social Science Research, # 1416509 National Science Foundation