Titre : | Network analysis for social media intelligence exploitation |
Auteurs : | Yannick Maekelberghe, Auteur ; Wim Mees |
Type de document : | Thesis |
Editeur : | Brussels [Belgique] : Royal Military Academy, 2016 |
Article en page(s) : | 68 |
Langues: | Anglais |
Tags : | SSMW 151 ; Master's dissertations ; Defense (E) ; Social media |
Résumé : |
The widespread use of the Internet and the ever-increasing accessibility of mobile technologies have been the driving forces that make social media intertwine with daily life. These platforms enable the creation and sharing of content between individuals on an unprecedented scale and thus provide a voice to those who would not be heard otherwise. Recently, social media are used by terrorist organizations to radicalize, recruit, and disseminate propaganda. This contorted use of social media implies a need for targeted data gathering and analysis to assist the efforts made countering current security threats.
This work aspires to be an added value to the domain of social media intelligence in two complementary ways. The first contribution being the establishment of a firm theoretical framework that merges the related aspects of intelligence with the concepts of social network analysis. The second contribution is the elucidation of the data mining process used to gather and exploit user data collected from Facebook. A dedicated hardware structure and several custom-built software scripts were conceived in order to collect and store the user data relating to the amount of individual likes on pages with a strong affiliation to a certain subject. This user interaction is examined visually in tandem with the application of social network analysis measures. Additional methods of analysis, such as the comparison of interaction patterns and the study of user activity over time, were trialled throughout the chapters and attained varying levels of success. The outcome of these inquiries not only allow for the identification of influential users and users that are critical for network connectivity but also provide a certain insight into the social networks formed by the Facebook interest groups subjected to research. The practical value of targeted and automated Facebook user data collection and the resulting potential for analysis is illustrated by a case study pertaining to Islamic radicalism. This collection infrastructure and associated analysis platform could potentially provide the means to detect and dissect social networks that facilitate the process of radicalisation and recruitment. The true power of this setup lies in the fact that the setup and procedures can be used on an unlimited variety of datasets. In order to capitalize on these promising opportunities the crawler infrastructure requires some improvements. Firstly, the collection operation needs to be split up into several sessions that use different IP addresses with the purpose of staying under the user request limit defined by Facebook. Secondly, the server needs more RAM so that larger datasets can be extracted from the database. Alternatively, this issue can be resolved by writing the data to disk memory for extraction. |
Promotion : | 151 SSMW |
En ligne : | http://units.mil.intra/sites/UBDef-BUDef/_layouts/DocIdRedir.aspx?ID=UBDEF-6-21700 |
Exemplaires (1)
Code-barres | Cote | Support | Localisation | Section | Disponibilité |
---|---|---|---|---|---|
119405R | RMA Mast SSMW 151/81 | Thesis | Royal Military Academy | Bibliothèque ERM | Exclu du prêt |