Example of energy industry ontology with external references in Fluent Editor

Modern energy sector is a wide area of industry, that concerns numerous aspects, like energy efficiency in different regions, renewable energy sources, energy companies' specializations and many others. Due to the variety of information, it is often difficult to get comprehensive answers to questions about specific fields. Semantic Technology allows to manage this knowledge in a simple and flexible way. It provides versatile description of reality, that is understand and can be adjusted to give complete, comprehensive information in chosen areas. This article presents simple ontology written in Fluent Editor, describing energy industry. It contains information about energy companies, regions and ecological aspects of their activities. You can download this sample ontology through the following link: EnergeoOnt.encnl

Outline

First we show how it is possible to adapt ontology for using external vocabulary sources. Then we present example of modelling concepts and relations from energy branch. Next we proceed to building more complex rules that allow to automatically infer more information basing on existing knowledge. Finally we present how created ontology can be used for getting answers to variety of queries concerning:

• energy companies, 
• their areas of activities, 
• specializations,
• assets, 
• different aspects of energy efficiency, 
• renewable energy sources. 

One can ask for precise information about given organization as well as more general, statistical data. Ontology construction also allows to limit area of results to specific regions of the world if needed.

Following example references external ontologies, including:

• reegle (http://reegle.info/schema) - used to describe clean energy actors, projects and technologies;
• fea (http://vocab.data.gov/def/fea) - business-based vocabulary framework;
• rov (http://www.w3.org/ns/regorg) - vocabulary for describing legal organizations in national registers;
• place (http://purl.org/ontology/places) - vocabulary connected with geographic places.

Usage of external vocabulary assets allows to extend knowledge base as well as make it more universal.

The ontology is divided into several parts, describing different aspects of expressed knowledge.

Part 1: Relations and attributes synonyms

This section defines equivalent roles (relations) and properties (attributes) which one can use to reference external ontologies' names.  For example role operates-in is declared as synonym for "profile"[reegle], therefore query

Who-Or-What is energy-company that operates-in Italy? 

will give the same result as

Who-Or-What is energy-company that "profile"[reegle] Italy?

Such ontology mapping can be expanded to concepts and instances. It allows to conveniently use own introduced names and at the same time maintain consistency with external vocabulary.

Part 2: Introducing new concepts, individuals and roles to ontology

Here are defined some basic concepts used throughout ontology, like "energy company". By using external references one can show connections between defined concept and already existing vocabulary. Stating that

Every energy-company is a company[fea].

implies that it is also a formal-organization within [org] namespace (referenced by [fea]). 

There are also some statements which will help with identifying organizations by unique names or inferring individual's properties without explicitly expressing them.

Next section of this part contains declarations of geographic locations and energy sources, especially including renewable energy sources and also some additional statements used during reasoning procedure.

Using statements presented above one can ask reasoner and materialized graph variety of queries about energy providers and renewable energy generation in different regions of the world. In example query:

Who-Or-What is renewable-energy-provider and operates-in South-America?

reasoner uses mentioned statements and knowledge about energy companies provided in further sections to infer company identity.

Another query example could be

Who-Or-What has-energy-provider something (that has-technology Solar-Power) and "in"[place] Europe?

Question above concerns countries located in Europe that have energy providers using solar power technology. Notice that parenthesis position here is important, because without them query would have different meaning and would give no answer, as role "in" is not declared for usage with energy companies.

Part 3: SWRL rules utilization

Following section concerns  usage of some SWRL rules for deducing new data attributes and object relations from declared attributes values.

If a thing(1) has-carbon-footprint equal-to the value(1) 
and a thing(1) has-total-energy-efficiency equal-to the value(2) 
and the value(1) divided-by the value(2) is equal-to the value(3) 
and the value(3) times 1000 is equal-to the value(4) 
then the thing(1) has-carbon-emission-ratio equal-to the value(4).

If a thing(1) has-carbon-emission-ratio equal-to the value(1) 
and Carbon-Emission-Ratio-Limit has-value equal-to the value(2) 
and the value(1) is greater-than the value(2)  
then the thing(1) has High-Carbon-Emission-Ratio. 

Carbon-Emission-Ratio-Limit has-value equal-to 800. 

First statement indicates that energy company has carbon emission ratio proportional to it's carbon footprint value divided by total energy efficiency. In second statement asserts that companies with carbon emission ratio above value represented by carbon "emission ratio limit" should be marked as having "high carbon emission ratio". Third sentence defines value of carbon emission limit. Example queries one can ask after graph materialization are:

Who-Or-What has-carbon-emission-ratio lower-or-equal-to 550?

which selects energy companies with carbon emission intensity not higher than 550 and

Who-Or-What has High-Carbon-Emission-Ratio and operates-in U-K?

which selects energy companies that have high carbon emission intensity and operate in UK.

Now any change in carbon emission ratio will influence query result. For limit value equal to 500 answer to the last query will be following:

Some presented values, like carbon footprint or carbon emission ratio, must be expressed in precise units, e.g. Mt CO2 or g CO2/kWh. To provide that information one can use annotations for key elements of ontology, as showed below.

Part 4: Describing individuals

This section contains variety of statements describing different energy companies, their roles and properties, such as activity area, operating sectors, specializations, energy efficiency in terms of total energy and renewable energy as well as total organization assets. This provides general data about individual companies and combined with statements included in previous sections is used to infer more valuable information.

Part 5: Queries

As shown previously, created ontology makes it possible to ask queries directly referring data, as well as concerning inferred information. User can get answers to question e.g. about:

• energy companies operating in specific regions;
• energy providers by energy source, especially including different kinds of renewable energy;
• different aspects of companies' energy efficiency, including total energy efficiency, renewable energy efficiency etc.
• ecological impact of industrial activity.

It is also useful, that one can formulate complex queries to receive very specific answers, e.g. about energy companies operating in given areas, utilizing solar power technology, that have specific energy efficiency. Much information can be inferred automatically from ontology, which otherwise would be difficult to access directly.

Summary

Presented ontology provides basic information about energy industry actors and green energy generations in different areas. It can be further develop to contain more complex and complete knowledge. Included references makes it easier to use by external users and provided metadata allows to receive variety of  information from queries to reasoner and materialized graph. Semantic model of data provided in ontology is very convenient and versatile way to manage knowledge about vast field of energy industry. It can be easily extended to cover more branch connected areas, like energy market information, or contain more specific information about individual power stations and companies' properties.

More Resources:

• Fluent Editor 2014 R4 direct download
• Ontology Examples (FIBO, GoodRelations, etc.)