ChemConnect2017 is an advanced state-of-the-art combustion database and data repository, derived from mechanistic, kinetic and thermodynamic chemical data (both experimental and modelling), organised into a network of interconnecting concepts. ChemConnect2017 goes beyond traditional repositories by providing a platform promoting data availability, searching and exchange. Through the parsing of the data sets, the pieces of information within data sets not only available to state of the art keyword searching algorithms, but through the use of semantic web techniques, also provides interconnections between independent data sets for efficient data exchange and comparison.
Data within the database has several levels of representation. At the top level is the data set in its original unmodified form and complimented with its source and reference. Data is kept in its original form to protect its integrity and reduce the introduction of errors. At this level, it is a traditional repository of information.
The second level consists of the individual pieces of information parsed from the original data. For accountability, there is a link between this parsed data and the original data from which it came. Individual pieces of data can be, for example, the thermodynamic and kinetic constants associated with the species and reaction data extracted from a CHEMKIN text file. At this level, it is a traditional searchable database.
However, the third level advances ChemConnect2017 beyond the state-of-the-art searchable databases. The third level transforms the database structures into a set of relationships based on the Resource Description Framework (RDF) from the World Wide Web Consortium’s (W3C) semantic web specifications. Keyword objects within the database objects are represented as a triple: a subject, a predicate and an object. The object keyword is related through a predicate keyword to a subject keyword. For example, a species label is connected to a reaction through the relationship saying that it is a product of (ProductOf) that reaction.
Transforming the database objects into this form provides a rich searchable interconnected network of fine-grained combustion data. This is the basis of advanced searching techniques and provides a platform in which data in the original independent data sets can be compared and contrasted.