The Radicalization Watch Project (RWP) is based on a multilingual monitoring system focusing on radical network activities worldwide. The project maintains an anonymized database of documents (text, audio, video) generated by extremist groups and radical organizations. It is intended to provide professionals with situational awareness of hostile environments through indicators and markers, video exploitation and content analysis, applying predictive linguistics, and quantitative and qualitative methodologies.
The RWP-SAD (Suicide Attack Database)
The development of the Suicide Attack Database (SAD) marked the initial phase of suicide research. It was our desire to build a “corpus” or collection of “testaments”, and analyze the words, phrases and sentences in them according to the aggressive emotions or violent actions they convey. The Radicalization Watch Project (RWP) maintains a multimedia database of documents on major suicide attacks since 2003. The database includes the original written, audio or video “testament” (suicide note or letter). The database is multilingual and uses native language sources (eg. Arabic) that are likely to have the most extensive relevant information.
Ref. Guidere M. (2010), Predictive Profiling of Lone Bombers
The RWP-VEM (Violent Extremism Monitor)
The VEM is a tool developed to help security experts screen more efficiently for those subjects that are undergoing religious radicalization that might lead to violent extremism. The online posts are electronically monitored and automatically analyzed based on a particular ontology. Natural language processing and artificial intelligence techniques (e.g., deep machine learning) are used to build a predictive model that could be used to automatically identify those online posts containing linguistic markers of radicalization leading to violent extremism.
Ref. Guidere M. et al. (2009). Rich Language Analysis for Counterterrrorism. Berlin, London, New York: Springer-Verlag.
The SNOW Database (Suicide Notes Worldwide)
After the RWP-SAD, the second phase of research was the development of the Suicide Notes Worldwide Database (SNOW). It was our desire to build automatically a collection of suicide notes, posted on the World Wide Web. The database contains over 1,000 notes, written by people who attempted or completed suicide. These notes were identified based on linguistic markers (semantic matrix and patterns) with an emotional connection to the subject of suicide (such as pessimism, negativism, guilt, sorry, fear, anger, blame, hopelessness, etc.).
Ref. Guidere M. & Howard N., (2012). LXIO The Mood Detection Robopsych. The Brain Sciences Journal, 1(1), 98-109.
The SRIP Tool (Suicide Risk Preventer)
Suicide is the one of the leading cause of death among young people (15-34 year olds), and many of those who attempt or complete suicide (25%-30%) leave a suicide note (letter or message) posted online, often on blogs and social networks. These notes contain information relevant both to the study of pathologies associated with suicidality and to the prevention of suicide and recidivism. The SRIP Tool monitors posts on the web and social networks and tracks linguistic markers of gravity and psychic pain, as well as signs of depression related to suicidal behavior. It goes beyond the mainstream sentiment analysis (simple binary positive-negative classification) by applying machine-learning classifiers on data captured from heterogonous sources (blogs, websites, forums, social networks, etc.). Ultimately, the goal is to carry out detection and prevention actions upstream, with innovative means resulting from predictive linguistics such as electronic monitoring of suicidal intentions, and remote monitoring of suicidal patients.
Ref. Guidere M. (2019), Predictive Linguistics Applied to Suicide Notes Posted Online by Suicide Attempters and Suicide Completers
6th World Congress on Mental Health, Psychiatry and Well-being, New York, 20-21 March, 2019.
Related Patent (Predictive Linguistics)
Method for Cognitive Computing: https://patents.google.com/patent/US20120064493A1/en
By connecting to this Internet platform, I commit to treat all the information contained in it as strictly confidential, to not divulge in any way this confidential information, and to take all the provisions necessary to protect them against such disclosure.