Semantic Modeling of Events (SCRIBE)

The SCRIBE project completed at IBM Research had a computer science aspect and a Smarter Planet aspect.

As a computer science project, we are exploring methods to build sets of modular ontologies (using OWL) starting from existing open standards, and we also specify the appropriate methods to compose these sets of ontologies with specific deltas (not built by us) that add and remove from the general purpose ontology for a general well-run client (in our case, a general well-run city), the specifics that could apply to one specific instance of a client (in our case, city) and its current processes and IT infrastructure. This research relates to Knowledge Engineering as well as to Model-Driven Development.

As a Smarter Planet project, we are testing the methodologies and tools being constructed on one very rich problem area: the information flows between discrete agencies and applications and citizens and businesses in cities that choose to collaborate in order to maintain quality of city life at reasonable cost and with minimal disruption.

There is so much that does go on in a city, so many resource flows, so many asset classes, so many roles, so many kinds of sensors and actuators, so many measured Key Performance Indicators, so many unplanned failures or demand surges, so much disruption from severe weather, so many kind of analytics whose results matter to different stakeholders, and so many policies about what information may be collected, must be stored, must not be stored, as well as policies which reflect as constraints on city operations, that this rich domain allows us to study how different parts of our modular ontologies can be built by different teams, at different times, but still be able to be used together.

And our ontologies do not correspond to any single city, but rather support the most common variations from nation to nation as to what services are provided by government, what services are not provided by government, which ones require user fees, which ones do not, which ones occur "on customer demand" and which on "a scheduled basis" etc. This variability is another issue we wish to address.

SCRIBE Ontologies will eventually support many open standards, but critical to supporting inter-agency interchange is the Common Alerting Protocol and the National Information Exchange Model (NIEM).

The SCRIBE project was begin in 2011 by Bob Schloss, Biplav Srivastava, Rosario Uceda-Sosa and we are fortunate to have periodic collaborations with our Research colleague Martin Stephenson in Dublin and our Research colleague Kalapriya Kannan in India.

We can now share the results of this work with academic colleagues, under the name Intelligent Semantic Model Palette. An introduction to our approach was presented at the Semantic Technologies conference held in June 2012 in San Francisco. See slides.

Publications

    • The D-SCRIBE Process for Building a Scalable Ontology, Robert Schloss, Rosario Uceda-Sosa, Biplav Srivastava, in Semantic Cities workshop held with Twenty-Seventh Conference on Artificial Intelligence (AAAI-14), Quebec City, Canada, July 28, 2014. [Describes D-SCRIBE process of building ontology]
    • SemEnAl: Using Semantics for Accelerating Environmental Analytical Model Discovery, Kalapriya Kannan, Biplav Srivastava, Rosario U.-Sosa, Robert J. Schloss, and Xiao Liu, Big Data Analytics (BDA 2014), New Delhi, India, Dec 20-23, 2014. [SemWeb-Envt Models]
    • Utility-driven Evolution Recommender for a Constrained Ontology, Pramod Anantharam, Biplav Srivastava, Amit Sheth. In 3rd International Conference on Web Intelligence, Mining and Semantics (WIMS'13), Madrid, Spain, June 12-14, 2013.[Ontology]
    • Building a Highly Consumable Semantic Model for Smarter Cities, Rosario Uceda-Sosa, Biplav Srivastava, Robert Schloss, in AI for an Intelligent Planet (AIIP) workshop held with International Joint Conference on Artificial Intelligence (IJCAI 2011) conference, Barcelona, Spain, July 18, 2011.[Area: Ontology-SmarterCities]