Fusing social media and mobile analytics for urban sense-making
The project was motivated by the observation that urban environments are increasingly characterized by a variety of non-traditional “sensors”, whose data streams can be harnessed to infer a variety of latent events and urban context. For example, users spontaneously generate huge amounts of content...
Saved in:
| Main Author: | |
|---|---|
| Format: | text |
| Language: | English |
| Published: |
Institutional Knowledge at Singapore Management University
2017
|
| Subjects: | |
| Online Access: | https://ink.library.smu.edu.sg/sis_research/6162 https://ink.library.smu.edu.sg/context/sis_research/article/7165/viewcontent/AD1037489.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Singapore Management University |
| Language: | English |
| Summary: | The project was motivated by the observation that urban environments are increasingly characterized by a variety of non-traditional “sensors”, whose data streams can be harnessed to infer a variety of latent events and urban context. For example, users spontaneously generate huge amounts of content (text, images and video) on social network channels, whereas GPS & other sensors on taxis and buses increasingly provide near-real time traces of their movement throughout the city. Similarly, advances in Wi-Fi based sensing allow us to passively capture the individual and collective movement of visitors across various public spaces, such as college campuses, museums and convention centers. The goal was to develop analytics algorithms and techniques to fuse the information from these diverse sources, to more accurately characterize urban events, both at city-scale (such as sports activities, and protests) and within specific campuses (such as university events). Through our work, we have developed a variety of novel algorithms for fusing such non-traditional sensor data. Our work has included (i) the use of Wi-Fi sensing data, optionally coupled with video monitoring, to detect events; (ii) the use of social-media and/or city transportation data to detect where and when events happended and (iii) the development of new crowdsourcing techniques that help increase the participate rate of city residents in reporting on such events. Our results and techniques provide foundational capabilities in the area of urban event sensing and anomaly detection. |
|---|