| 摘要 |
A rule based alert investigation system, which is able to identify anomalous trading behavior as satisfying a predetermined pattern of trading transactions and events on real time data feeds, filters out eligible trading transactions and events and retains the unique identifiers of the transactions and events for inclusion in one or more data sets associated with the predetermined pattern. The system generates an alert when the eligible transactions and events satisfy the predetermined pattern rules, and retains the eligible event and transaction identifiers in an alert data set along with the generated alert. The system utilizes the retained identifiers to retrieve the underlying actual transactions and events in their chronological order, which eliminates manually locating the transactions and events that triggered the alert. The system further provides commentary articulating the anomalous trading behavior along with the relationships between the eligible transactions and events contained in the alert data set. |
