Xu, Z., Zhang, H., Sugumaran, V., Choo, K., Mei, L.,

From Digital Culture & Society

Jump to: navigation, search


Participatory sensing-based semantic and spatial analysis of urban emergency events using mobile social media. Review by Ben Blanchard

Recent advances in information communication technologies(ICT'S) it is critical to improve emergency management through modern data processing techniques in order to generate response and recovery plans during non disaster times. This article introduces a sensing based model for mining spatial information of urban emergency events in order to explore the use of data mining, statistical analysis, and semantic analysis to obtain information on public opinion. Chinese microblog data gathered from Weibo with Typhoon Chan hom(2015) being used an an example of an emergency event. The abstract if effective in that it introduces the topic of the article clear and concisely and provides a detailed outlined of what each individual sub section of the article will cover and what it hopes to accomplish with its study.

With advances in ICT's it is critical that emergency management systems are improved through the use of modern data processing techniques. Lack of Information, constantly changing situations and short time for decision making are many of the common challenges faced in any disaster scenario. The four stages of emergency management are planning and mitigation, preparedness, response and recovery with geospatial applications(GIS) being utilized in each stage of emergency management. Social media platforms such as twitter and weibo allow used to provide geographic information being referred to as volunteer geographic information, allow for the cost effective and timely data to emergency planners. This article introduces a participatory sensing-based model for mining spatial information of urban events by first providing basic definitions of proposed methods, secondly positive samples are selected to mine the spatial information or urban emergency events, thirdly location and GIS information are are extracted from positive samples. Lastly the real spatial information is determined based on address and GIS information.

The following section introduces a brief literature review on the subject matter. Earle et All examined how fast tweeters reacted to the earthquake in Morgan Hill Ca march 2009. It is noted that many researcher have used data from social networks such as Twitter and Weibo with Sakaki investigating the real time nature of twitter with a focus on event detection. Crooks et All analyzed the spatial and temporal features of twitter feeds responding to an earthquake, arguing that twitter used represented a hybrid form of sensor system. Longquville et all examined twitter in response to a forest fire in order to demonstrate its role in supporting emergency management. This brief literature review works well to provide a brief summary on prior research which helped establish what this article will add to the existing literature.

The proposed method is then illustrating beginning with basic definitions of key terms within the study such as urban emergency event, word set of an urban emergency management event which includes 3 different layers. (1) the social user layer, (2) the crowd-sourcing layer, (3), Spatial information layer. Next positive samples of the study are defined by 3 different heuristics. A semantic analysis is then used in order to extract useful data from mass data along with location data an GIS information which together can be used in order to detect whether a word is a location name or not. Some background information is also given on the typhoon Chan-hom which services as the case study for this article. Chan-hom made landfall in Zhejang, China on July 11 2015 with maximum winds speeds of 45 m/s, it was tracked using physical and GIS data. The total number of event related microblog posts was 3321. The semantic analysis on this microblog data found that citizens paid more attention in the area the disaster happened than in other provinces where a greater concern was put towards having information on future situations. An explosion event was also examined in the area as due to the increasing threat posed by terrorism these disaster are more likely to occur. The explosion that was examined was an explosion at “Taoyan Apartment” on January 17 2015, searching Weibo using the keyword “explosion” 478 messages are returned with 45 being providing location information. Using this information a time-line of the event and response was also established. The established method was also applied to this event using sensor networks and data collection systems based on cloud storage platform. This allows for more rapid and accurate leakage location with GIS mapping. In the conclusion of this article Zheng et all provides a summary of his proposed research method, further reiterating that the aim of the study was to explore data mining, statistical analysis, a semantic analysis methods to obtain valuable information which can be used to enhance situational awareness and help governments offer more effective assistance.

This article provides an extremely detailed and thorough analysis on how mass data from social media platforms can be extracted and refined into useful data than can be used to improve emergency response and management. The sensing based model established within this article provides an original and effective method of analyzing the mass amount of data generated during an emergency event and extract useful information from it. The method developed in this paper for studying the use of social media platforms in emergency situations and how it can contribute to the improvement of emergency response present an important contribution to the ongoing work being done to determine how best these services can be utilized. It highlights considerable challenges that face emergency response networks in using information communication technologies, such as the difficulty in extracting useful information from the large influx of data that emergency events typically generate. This article was however extremely technical and confusing in certain sections, the application of the proposed method on page 7 was difficult to understand and make sense off. This article could have perhaps benefited from having a more concise argument that could have been reflected upon in its conclusion. This would serve to give the article some much needed focus in certain areas.

Personal tools
Bookmark and Share