Analysts reconstruct the data state of processes from an event log that records the process execution trails. Efficient analysis of these logs requires the storage of event data throughout the lifetime of a process. Traditional trace-based logs store the traces generated by (isolated) process instances. These call for reconstructing case variables after each event in the trace. Object-centric process mining pushes this complexity one level up, as each event can create, delete, or update multiple artifacts like objects, relations, and attributes. Besides, object-centric event data alone is not good enough to reconstruct such updates. In this paper, we tackle this pressing problem by augmenting object-centric event data with a lightweight form of domain knowledge, consisting of condition-effect rules. These rules explicitly capture the semantics of events contained in the logs. We then propose a method that takes an object-centric event log and a set of condition-effect rules as input, and produces a timeline accounting for the step-by-step evolution of the artifacts. Cypher queries can then be used to inspect the timeline and retrieve which facts hold at a given time point, or the time intervals of existence for the artifacts.