Case Study 3: Assault Investigation – Test Retrospective Claims
| Background | A number of complainants asserted that they did not know each other. Defence alleged complainants conspired to change statements. Social media was used to show probable real-world links between complainants and dialog in relation to the investigation and trial. |
| Collection | We used a four different programmatic tools to download public and credentialed (from cooperating party) account information from a number of social media accounts that appeared to be operated by complainants. A cooperating party allowed us to use their Facebook account to view data from friends and friends of friends. |
| Collation | In the case-based investigation data repository we collected many thousands of items including posts, chat, photos, check-ins, likes, friending requests etc. In separate repositories- using both public and “friend” account credentials – we were able to deduplicate and highlight second level connections between contacts (friends of friends). Data from several social sites was downloaded, normalised and combined into a single case repository. |
| Search and Analysis | Clustering and visualisation tools showed different forms of connections and communications indicating comments, connections, likes etc. We searched the collection for keywords and highlighted communications relating to key people and subjects involved in the case. Social network activity was analysed for function, density and reciprocity. |
| Product | Social network diagram – showing interactions of several different types (likes, dialog relevant to case and other communications) Date / communication / friend matrix Timeline Report Court appearance explaining data collected and presentation to the jury of detailed trial graphics Cross examination of material by Crown prosecutors. |
| Outcome | Our material contributed to the objective categorisation of complainants’ statements. |