Faculty Contact: Jianwen Su
Abstract: A temporal database can act as a log for larger database systems, tracking processes and activities executed in a system over many time intervals. These time-ordered databases, or process logs, are subject to queries in built without explicit timestamps, as query languages need not quantify over the implicit ordering on the structures over which they are evaluated. Given that linear temporal logic is effective for expressing changes in the truth of propositions over time, I investigate the usefulness of FO-LTL (first order linear temporal logic) as a query language for expressing properties of these process logs. Because these logs are sequences of instances of a database schema, the addition of first order logic increases the capacity of the query language to describe changing properties of data fields, a desirable extension to traditional temporal logics. This project involves an explicit construction of process logs and of the semantics which define their satisfaction of FO-LTL formula, as well as an examination of their use in querying workflow logs, logs of executions of workflow executions in business processes.
- Fall 2017: Isaac Mackey