A novel approach for detection and ranking of trendy and emerging cyber threat events in Twitter streams

Abstract

We present a new machine learning and text information extraction approach to detection of cyber threat events in Twitter that are novel (previously non-extant) and developing (marked by significance with respect to similarity with a previously detected event). While some existing approaches to event detection measure novelty and trendiness, typically as independent criteria and occasionally as a holistic measure, this work focuses on detecting both novel and developing events using an unsupervised machine learning approach. Furthermore, our proposed approach enables the ranking of cyber threat events based on an importance score by extracting the tweet terms that are characterized as named entities, keywords, or both. We also impute influence to users in order to assign a weighted score to noun phrases in proportion to user influence and the corresponding event scores for named entities and keywords. To evaluate the performance of our proposed approach, we measure the efficiency and detection error rate for events over a specified time interval, relative to human annotator ground truth.

Semantically Modeling Cyber Influence Campaigns (CICs): Ontology Model and Case Studies

“…This study has focused on developing and evaluating an ontology model for analyzing cyber influence campaigns in conflicts conducted in social media networks. Social media networks can also give indications of emerging cyber security threats [119]- [123]. One interesting future work direction is to adapt our ontology model to uncover the sources and agents behind emerging cyber threats.…”

Motion-Based Handwriting Recognition and Word Reconstruction

“…Dataset for Auto-Correction Since our auto-correct model is implemented based on SymSpell [21], we utilize the standard corpus from SymSpell for auto-correction lookup. The corpus itself is a frequency dictionary created by combining the Google Books Ngram data and Spell Checker Oriented Word Lists (SCOWL).…”

“…The last puzzle piece of our word reconstruction system pipeline is to summarize the top K predictions from the trajectory search and combine the corresponding information from the auto-correction model to finalize the prediction for the word. Our auto-correction model is based upon Sym-Spell, which is proven [21] to be an effective non deep learning approach for correcting misspelled word.…”

Detecting Trending Terms in Cybersecurity Forum Discussions

“…Once a trending topic is identified, topic ranking is needed, to avoid overwhelming a user. This is used to highlight current important topics, including Bose et al (2019) who use this to detect and flag known serious threats.…”