This study investigates the use of Twitter in the emergency domain by developing a Twitter-integrated web system capable of collecting and processing emergency-related tweet data. The objectives of the project are to collect voluntary emergency tweet data from public sources via the Twitter API, process the data based on geolocation information and syntax into organized information entities relevant to an emergency situation, and then distribute the information deliver to a map. -like interface. It is hoped that the project can ultimately contribute to the existing knowledge domain on social media-enabled emergency applications.
The greatest thanks are also expressed to Universiti Teknologi PETRONAS (UTP), especially the Final Year Committee of the Department of Computer Information Science (CIS) for their excellent organization and management of this course. Last but not least, the writer would like to thank all respondents who participated in the survey initiated by the writer as part of the project.
INTRODUCTION
- Background
- Problem Statement
- Objectives
- Scope of Study
Studies have been conducted into the possibilities of utilizing volunteer emergency data on Twitter in a more useful, structured way. This project provides a proof-of-concept model on how Twitter can be leveraged in emergency communications, by developing a Twitter-integrated web system that will aggregate and process emergency-related data on the Twitter platform to provide a geovisualization of emergencies. situations. A solution is needed that can automate the aggregation of emergency-related tweet data available on social media platforms and organize it to produce an emergency-specific information report with a geographic perspective.
The scope of tweet data to be collected is tweets that satisfy a predefined hashtag-based syntax for a specific emergency event, with the format prescribed by the emergency management organization. The system is generally aimed at emergency management organizations seeking to acquire and organize the volumes of volunteered emergency-related data available on the Twitter platform, and although it can potentially be a tool to support formal decision-making processes in aspects of management in an emergency, when an emergency occurs, it is not useful for emergency preparedness, emergency recovery, or policy communication.
LITERATURE REVIEW
- Overview
- Leveraging Social Media in the Emergency Domain
- Twitter Use During Emergencies
- Collecting and Processing Emergency-Related Data from
Twitter has been described as a way to provide information not covered on radio and television, such as details and first-hand accounts immediately after an event (Mills, Chen, Lee, & Rao, 2009). While the broader Twitter audience tends to be interested in high-level accounts of an emergency event, retweets by locals affected by an emergency tend to be more specific, which can help dissemination of relevant local information (Starbird & Palen, 2010). Although the focus of this particular research is not on disaster preparedness, TED nevertheless demonstrates the rich possibilities offered by social media in the context of integration with the emergency domain, such as the ability to join together and make easily accessible information contributed by citizens. , as well as improving public outreach regarding timely information in the field (Guy, Earle, Ostrum, Gruchalla & Horvath, 2010).
It is necessary to develop an understanding of potential strategies that can be applied in the collection and computational processing of Twitter data to maximize the utility of extracted emergency-related information. In the context of this research, information extraction is focused on hazard-focused data that is delivered immediately after an emergency.
METHODOLOGY
- Project Development Methodology
- Project Activities
- Tools Required
- Key Milestones
- Gantt Chart
The project activities include the software development methodology processes that make up the project. The design phase relies heavily on the documentation from the requirements planning phase to validate whether an effective model of the system is being built. It will continuously contribute to the refinement of the system architecture, which will serve as the basis for the system construction.
It provides a high-level overview of the various components and data flows within the system. The 140dev Twitter framework is the backbone of the system that aggregates and analyzes raw data from the Twitter platform, and the tweet data is further processed with custom algorithms for the system's custom needs before being stored in a MySQL database . The construction phase is performed iteratively with the design phase, and focuses on the programming/application development and testing aspects of the project.
The separate functional modules of the system are developed in accordance with the system architecture, namely the aggregation, analysis and display functions. The system usability scale (SUS) is the method of choice, as it is technology independent and a general yet well-known measure of the perception of system usability. The results of the SUS questionnaire carried out in this project are tabulated in Chapter 4.
The advantages of the XAMPP solution stack are that it is free, open source, and cross-platform. The streaming API's low latency means there is minimal delay between the creation of a tweet and its delivery to the API. The core module of the framework is the Twitter database server, which connects to the Streaming API to collect Tweets based on specific keywords and cache them in a MySQL database.
RESULTS & DISCUSSION
- Use Case Diagram
- Aggregation Process
- Parsing Process
- Display Process
- Sample of Functional Testing Results
- System Usability Scale Survey Results
- Limitations
Monitoring Agent: The monitoring agent represents a member of the emergency management organization that implements the system and that monitors incoming Tweets for events. If the information is confirmed by reliable sources, the corresponding tweet will be marked as verified in the results on the web interface. The current design relies on the Twitter platform as the main source of data input for the system.
The syntax, in turn, is determined by the emergency management organization implementing the system based on the nature of the emergency and the type of data required. The purpose of the #loc location tag is to allow geocoding to be performed on the location text if the geotagging feature is disabled by the user of the tweet. The processing algorithm performs programmatic operations on the tweet text to extract relevant informational entities within the tweet.
In short, a web application built on top of a data collection framework can safely rely on a database of tweets without being subject to any issues with the operation of the Twitter API. The location of the tag depends on whether the tweet is geotagged. In short, it covers the entire process of feeding a tweet to the Twitter platform and ensuring that the tweet is displayed on the system's web interface with the right information.
On receipt of the above tweet by the system, the tweet is displayed on the notification sidebar of the web interface on a default screen for. The screenshot of the output tweet displayed on the message sidebar is as in Figure 4.13. From Figure 4.13, we can see that each element of the tweet has been correctly processed by the system.
A system usability scale (SUS) survey was conducted to measure the perception of usability for the web interface of the developed system as described in Section 4.4. The system is also limited in that it still has to rely on Twitter users to adjust their tweets so that the tweets conform to the extremely rigid specifications of the tweet format that are set in advance.
CONCLUSION & RECOMMENDATIONS
Conclusion
Recommendations
Geo-Twitter Analytics: Applications in Crisis Management, Proceedings of the 25th International Cartographic Conference, Paris, France.
SUS Questionnaire
One of the prominent trends that have contributed to said phenomena has been the emergence of social media platforms. This makes it easier for the system developer to continuously refine the system model and produce improved iterations at different times during the project. 140dev Twitter Framework is the backbone of the system which collects and analyzes raw information from tweets, an alternative approach to “train” Twitter users to create such a readable format is proposed [12].
A potential weakness of the methodology is that certain performance may be required to complete the project in the specified time. The 140dev Twitter framework is the backbone of the system that collects and analyzes raw. On-Site Citizen: On-Site Citizen may be a part or casual observer of the emergency that provides input from the Twitter platform and the tweet data is further processed with customized algorithms for the system's custom needs before being stored in a MySQL database.
With the exception of the mandatory hash tags, the tweet format above can be adapted to accommodate different emergencies that have different alert types, with different descriptions and location. The alert type hash tag determines the category of alert within the event, while the #dct tag precedes a text description of the alert. The monitoring agent represents a member of the emergency management organization that implements the system and that monitors incoming Tweets for events.
The purpose of the #loc location tag is to allow geocoding to be performed on the location text if the geotagging feature is disabled by the user of the tweet. The purpose of the #loc location tag is for the encoding to be performed on the location text tagging feature to be disabled by tweet activation, however, the geotagged location takes precedence over the location specified in the tweet text when processed by the system. The length of the location description and text is solely at the discretion of the user, as long as it is within the 140 character limit enforced on Twitter messages.
Technical Report
