Project Proposal Shu Chen Project Topic Events are extremely important in daily human life, and present in diverse areas of engineering and science. A lot of applications emerge in both academic and industrial areas. Examples are simulation, monitoring, business process, knowledge representation, environmental modeling, and active database systems. Events have two characters in information systems are: (1) events are triggering conditions rather than being stored and queried as traditional data, such as events in active databases [1]. (2) Events are subjects of data representation and query processing, such as events in spatiotemporal databases [2]. We can consider events in the context of spatial-temporal databases, where events establish something that happens at a specific point or region and time instant or interval, but then it is gone. Unlike entities in traditional databases systems, relationships between events, play an important role for querying events, like temporal order, spatial connectivity or causality. An example of such stream system can be a Sensor-based Activity Recognition system which has raw sensor data stream as input, and labeled activity as output.
We are going to implement a few new data types and operators in SECONDO database system [3] in order to support event query on spatial-temporal database to answer the query like: which event happened after event E1 happened in the same area before T1 time. Traditional database does not support the event query like given before by either lacks the spatialtemporal support or cannot represent the relationship between events. We decided to choose extend SECONDO database system, which is modular, extensible and clearly documented. The SECONDO system features different algebra modules to support different models, such as relational, graph model, moving objects etc. It also supports data visualization with the extendable model viewer object and optimizer rules. We will build a new module called “Event
Algebra” to support event models. This algebra includes two data types (Event and Edge) and one operator.
Background and Related Work 1. Event Query There is no clear consensus in the meaning of the term event. Events are also referred as occurrents, as opposite to continuants or objects. Occurrents happen or occur and are then gone. Depending on the type of events, they can occur at a time instant or last for a time interval. A single event that occurs at a time instant or lasts for a time interval is atomic, whereas a composite or complex event lasts for a time interval and is defined by an aggregation of events, atomic and composite events [4].
In spatio-temporal databases, tree basic concepts have been used to model applications in this domain: objects, events, and processes [5]. They have spatial and temporal locations, also called settings, and can be related by means of involvement and participant relations. Paper [6] gives a description on how to model the event in graph-oriented manner, and it is the main reference of our project.
2. Secondo System Secondo [3] database system uses a data-type based approach to implement the data model. Each data type is defined as separate modular algebra. The system is designed for a general purpose DBMS that fit different application requirements, including spatial, spatiotemporal, graph model or streaming applications. The Secondo system has three major components: kernel, optimizer, and GUI interface. The kernel is implemented using C++, and specific data models are defined. To extend data types, new algebra modules need to be defined. Meanwhile, it also provides basic query processing over the algebras using a special syntax. The underlying storage manager is
based on BerkelyDB, an open source database system. The optimizer is implemented using Prolog, and support conjunctive query optimization. This optimizer gives user the possibility to use SQL-like syntax for different queries. The third component is GUI interface implemented using Java. It supports different viewers for different spatio-temporal and moving object applications. The three components can work independently or collaborate with each other. Project Outline In order to support event query in SECONDO, we plan to add several new data types and several basic operator and combine them into a new algebra module “Event Algebra”. We will build a full-fledged event system. And the proof of the closure of this algebra will not be issued in our project. For example, we are only implementing a few data types and operators. The data types we are implementing include: Event (a list of the name, location, time) Edge (the relationship between events) The operator we are implementing is: events, edges: returns a set of events or edges get_event, get_edge: returns a particular event or edge. union, intersection start, end: time point of interval
Timeline: Implement Event and Relation Data type (now till Apr. 18) Implement Operators (Apr. 18 till Apr. 25) Extend current viewer (optional, now till Apr. 25) Debug and Testing (Apr.25-Apr.27)
Writing Report (Apr.27, 28) References:
[1] Zimmer, D., Unland, R.: On the semantics of complex events in active database management systems. In: ICDE, pp. 392–399. IEEE Computer Society, Los Alamitos (1999) [2] Worboys, M.F.: Event-oriented approaches to geographic phenomena. International Journal of Geographical Information Science 19(1), 1–28 (2005) [3] Seconod Database System. http://dna.fernuni-hagen.de/Secondo.html/ [4] Bry, F., Eckert, M., Patranjan, P.-L.: Querying composite events for reactivity on the web. In: Shen, H.T., Li, J., Li, M., Ni, J., Wang, W. (eds.) APWeb 2006 Workshops. LNCS, vol. 3842, pp. 38 –47. Springer, Heidelberg (2006) [5] Galton, A., Worboys, M.F.: Processes and events in dynamic geo-networks. In: Rodr′ıguez, M.A., Cruz, I., Levashkin, S., Egenhofer, M.J. (eds.) GeoS 2005. LNCS, vol. 3799, pp. 45–59. Springer, Heidelberg (2005) [6] Miguel Romero and M. Andrea Rodríguez R.: A Graph oriented Model and Query Language for Events, First International Workshop on Semantic and Conceptual Issues in GIS (SeGoGIS) Lecture Notes in Computer Science Vol. 4802, pp. 358-367, Springer -Verlag