Tracking Quantity Fluctuations using STT

We describe an approach to extracting and tracking events in which measurable quantities such as economic indicators undergo a change. The causes and consequences of oil price fluctuations are an example of such events, and are tracked in the STT (Situation Tracking Testbed) prototype. We propose a representation of these event types in the logical representation language CycL and apply this representation to the problem of identifying and extracting information from news sources in RSS feed repositories. Conceptually, these event types often represent changes in quantities that denote capacities, rates of supply or demand, inventory levels and similar measures. We pro- pose to call these event types quantity change events. The ability to identify and track such events—and the chains of causal relationships that they represent—is an important tool for intelligence analysis. Such event chains can be of strategic importance, especially where energy prices are concerned. In adding this class of situation modeling, STT is applying insights from qualitative process research (Forbus 1984) to the problem of situation tracking. The structure of the remainder of this document is as follows. We first give an informal characterization of the language used in source documents to describe the infor- mation to be extracted. Our primary example domain will be quantity changes related to oil price fluctuations. Since STT's research activities are guided by its qualitative models, we next propose a representational approach to modeling the quantity change events discussed in our sources using the logical representation language CycL (Matuszek et al 2006). We then describe how STT will use this representation in extracting and tracking chains of quantity change events in document corpora such as RSS feed repositories. We conclude with a prospective look at how we have begun to evaluate the performance of the tracking system—and thereby the appropriateness of the representation— and will continue to in the near future.