SemTechBizThe Financial Industry Business Ontology (FIBO) was a main topic of interest at last week’s Semantic Technology & Business Conference – which took place in New York City, the capital of the financial services industry. FIBO, as The Semantic Web Blog has previously discussed, is both a business conceptual ontology and an operational ontology delivered together, designed to be useful both to the financial industry and the regulatory community in understanding the complex patterns and relationships of information characteristic of the sector, with the goal of driving greater transparency. The FIBO initiative is a joint effort underway by the Object Management Group and the Enterprise Data Management (EDM) Council.  But many other different standards will be useful to solve the industry’s issues, as well.

At the presentation, Semantics in Finance, Thematix Partners’ principal Elisa Kendall – self-described standards wonk and member of the OMG Architecture Board and co-chair, Ontology Definition Metamodel (ODM) Revision Task Force – pointed out that the amount of regulation in the financial services sector has increased over 400 percent in the last two to three years. She argued for a little more sympathy for the financial services industry, too, which hasn’t been on the receiving end of a lot of that since about 2008 – even though some of these players stepped up to buy companies that were knocked flat by the mortgage market meltdown.

“In five years they are supposed to understand all the data exposure of all the entities they politely took on,” Kendall said. “We should give them a bit of a break. So we need better standards for banks and others in the industry to identify all the hooks and connections, and move some of those silos into at least being more integrated from a very high level set of knowledge bases, in order to do some of the queries the government is asking them for.”

To get to the point where banks can figure out the connections across their assets, including the ones they acquired, and regulators can understand the data about how different banking groups hook into each other, requires “global standards for reporting and interchange across institutions. These standards need to include higher level terminology and definitions so that the regulators and bankers can have confidence in the meaning for the terms used in reporting or information exchange. And the vocabularies and schema used must be much more expressive and evolvable to address these challenges.”

kendallnewAs Kendall described in an interview following her presentation, the biggest pieces of the puzzle revolve around developing the standards to support the kinds of queries the banks get from the regulators, as well as to support their own internal analysis requirements around their level of risk, risk mitigation and compliance with government policies. “Internally they have to do that given how huge some are and how far-flung some of their legal entities are, and externally they have to report that to regulators,” she said. “And regulators want to be able to aggregate some of the information, whether banks want them to or not, to understand what goes on in the systems and to identify issues before they get as big as they got before.”

FIBO fits in with its focus on standardizing the language used to precisely define the terms, conditions, and characteristics of financial instruments; the legal and relationship structure of business entities (what is a corporation, trust, a legal person and so on); the content and time dimensions of market data; and the legal obligations and process aspects of corporate actions, based on the legal structures and obligations contained within the myriad of contracts that form the foundation of the financial industry. The OMG at the end of September published for comments the FIBO Foundations ontology, which covers the basic semantic abstractions needed to support the forthcoming FIBO Business Entities specification, and which will be further updated to provide the semantic abstractions needed for securities, derivatives, loans and so on.

Modifications included an increased use of metadata and provenance, and “modularize it in a big way,” she said, separating concepts along topical dimensions and commonly used properties into separate modules. “We added richness from a semantic perspective so that we can leverage reasoning and rule engines to make the connections and answer the kinds of queries we anticipate will be needed, to do the job that an ontology is intended to do for you.” Comments to this are due by Nov. 11, the same timing for publishing the FIBO Business Entities, for which comments will be open until February.

“We’ll stage the pieces for securities, derivatives, equities that are parts of FIBO too, with ontologies for each of those, over the next couple of years,” Kendall says, but a chunk of that work already is done. It’s important to address these areas as they’re the ones where the most risk is involved, as evidenced by the 2007-2008 meltdown, she explained.

The OMG also is considering other standards that can move forward independent of FIBO but integrated with it in a standard way, Kendall notes: That includes mapping to FIBO the Bloomberg Open Symbology, a flexible and open system for identifying securities across all global asset classes, OMG Property & Casualty, ACORD insurance standards, ISO 20022/BIAN (at least for corporate actions), and parts of  XBRL.  Taking FIBO as inspiration, she added, is the Financial Industry Regulatory Ontology and Financial Industry GRC Ontology (from the Governance, Risk, and Compliance Technology Centre, College University Cork), which is work in progress to enable efficient access to the spectrum of regulations by relying on formal semantics to capture and represent the knowledge embedded in such regulations.

“The first step is a standard language to talk about financial contracts, and then let’s extend that to build systems that understand the regulations and help us understand whether or not we are in compliance with these things that FIBO describes, in our business processes,” she says. There’s been a lot of involvement from the financial services industry in keeping the FIBO momentum going forward, but it’s not done yet and the more that get onboard, the merrier. “We are still building a broader community of practice and interest around this,” says Kendall.

More Financial Sector Momentum

The interest in FIBO, and in helping the industry to use it,  was evident in another session at the conference, The Perfect Match – Linking FIBO and Business Processes Through Semantic BPMN, led by Lloyd Dugan, co-author of BPMN 2.0 and chief architect for Business Management Associates and Mohamed Keshk, senior semantic architect at Semantic BPMH who built the first ontology-based query engine for BPMN (Business Process Model and Notation) 2.0. Semantic BPMN, they noted, is a standard-based, model-driven platform to enable the capture, retrieval and management of interoperable process information and enterprise corporate data.

“FIBO is great for transparency but it lacks the processes,” Keshk says, as it’s all data-oriented. Bringing the FIBO standard together with the BPMN standard is what their work in semantic BPMN delivers, They showed to the audience their working prototype on linking process models and FIBO-based data by allowing business owners to turn their business process diagrams from static documents into a run-time application that enables them to integrate, publish, query, and infer over business process models and FIBO-based data.

“No one else has the answer on how to match legacy data to the ontology stuff,” says Dugan. “What they do is to map to a data source — a row in a table or document or XML file to an ontology concept, but that is just re-representing data to a format you can more easily navigate. We align that instance data to process context, so you can use the vocabulary of the business process to ask about the data.” Using process models with semantic technology to make everything queryable in the same way, they explained, delivers the process context to make the data context dynamic. Their prototype is not limited to the financial arena and FIBO, either, but supports a plug-and-play approach for use with any ontology.