A Framework for Analyzing Twitter to Detect Community Suspicious Crime Activity

Abstract

This research work discusses how an integrated open source intelligence framework can help the law enforcements and government entities who are investigating crimes based on statistical and graph analysis on Twitter data. The solution supports a real-time and off-line analysis of the tweets' collections. The framework employs tools that support big data processing capabilities, to collect, process and analyze a huge amount of data. The outline solution supports content and textual based analysis, helping the investigators to investigate a person and the community linked to that person based on a tweet. Our solution supports an investigative process composed of the following phases: (i) find suspicious tweets and individuals based on hashtags analysis; classify the user profile based on Twitter features; (iii) identify influencers in the FOAF networks of the senders; and (iv) analyze these influencers' background and history to find hints of past or current criminal activity.

Date of Award	May 2017
Original language	American English