Usability
- PDF / 959,964 Bytes
- 32 Pages / 547.087 x 737.008 pts Page_size
- 73 Downloads / 200 Views
Uncertain Databases ▶ Probabilistic Databases
Uncertain Information ▶ Incomplete Information
database applications, the events of interest were not the results of single queries, (e.g., insertion or deletion of data), but rather could be deterministically inferred from several such queries. To facilitate such inferences, event inference languages were defined. Initially such languages were specific to active databases (e.g., SNOOP and ODE). However, more general languages were developed suitable for implementing such deterministic inferences in any event-based system. These general languages resulted from the need to enable event driven behavior in a wide variety of application domains and became part of a wider area known as Complex Event Processing (CEP). As event driven applications became more complex, it became necessary to handle uncertainty regarding the occurrence and inference of events.
Foundations
Uncertainty in Events S EGEV WASSERKRUG IBM Research, Haifa, Israel
Synonyms Event uncertainty
Definition Uncertainty in events is uncertainty regarding either the occurrence of an event, or uncertainty regarding the data values associated with an event. This uncertainty is a result of a gap between the actual occurrences of events in the real world, and the availability of knowledge regarding the events.
Historical Background The first event-based systems were active databases, in which automatic actions were carried out as a result of database queries. This was done using the ECA (EventCondition-Action) paradigm. However, in many such #
2009 Springer ScienceþBusiness Media, LLC
For a system to implement event driven behavior, the system must be able to recognize all events of interest. However, in many cases, there is a gap between the actual occurrences of events to which the system must respond and the data generated by monitoring tools regarding these events. This gap results in uncertainty. To understand this, consider a thermometer that generates an event whenever the temperature rises above 37.5 C. The thermometer is known to be accurate to within 0.2 C. Therefore, when the temperature measured by the thermometer is 37.6 C, there is some uncertainty regarding whether the event has actually occurred. Another gap between the actual occurrence of events and the information available to the system is caused by the following: The information regarding the occurrence of some events (termed explicit events) is signaled by event sources (e.g., monitoring tools such as the thermometer described above), while for other events, explicit notification is never sent (non-explicit events). An example of a non-explicit event is insider trading. Although insider trading
3222
U
Uncertainty in Events
either does or does not take place, no explicit signal regarding such an event is generated. For an event-based system to respond to non-explicit events, in many cases the occurrence of these events must be inferred based on the occurrence of other events. (The events based on such inference are termed inferre
Data Loading...
