Tackling phenomena such ascyber bullyingor analyze thespread of rumors on the internetare the objectives ofCyberguardian, an application capable of connecting to a social network and automatically extracting information from web pages. “This tool is capable of behaving like a person browsing the Internet and when it finds interesting information, it stores it in a database and makes it available for later analysis,” explains Luis López Fernández, professor of Telematics Engineering at the Rey University. Juan Carlos (URJC) and person in charge of the project.
Thanks to this analysis, you can perform tasks such as periodically checking what new friends a person has, what messages have appeared in their account, what new events they have been called to … But with a limit: the information that each user makes himself publicly available. “The legality and thelimits of privacythat people want to impose on their own information, “says the researcher.
Ciberguardian technology is based on two modular components. Firsta crawlercapable of connecting to a web server, downloading documents from it and building a computerized representation of the information that appears in the HTML pages that users see. Thanks to this representation, the robot can retrieve and store texts, images, videos or any other element that may appear on a browser screen. The second of the components isan intelligent module that analyzes textsthat the robot has retrieved to determine, from the words that form it, if the associated message contains information relevant to the objectives of the program.
In this way, if you want, for example,detect harassing or threatening messagesSimply add to the system a dictionary of words that appear frequently in this type of message. Furthermore, due to the modular nature of the application, it is easy to integrate natural language computerized analysis libraries that could be used to classify the extracted messages or to extract entities, opinions or relationships between them.
In addition, CiberGuardian can be very useful for fields such as sociology or marketing. And this application is capable of analyzing aspects such aswho knows whom within an internet social network. That is, the structure of said network can be reconstructed and visualized by means of a graph, in which the nodes are the people who are part of the network and the edges represent the relationships between them (friendship, message exchange, etc. ). When this graph is represented in a data structure, it is easy to analyze aspects such aswhat roles appear in the network, who are the leaders, how information and rumors spread, what communities of interest are there, or in what ways does the network structure evolve in the face of external events.
