jump to navigation

Semantic Technology and Crystal Reports June 30, 2011

Posted by kirbyr in Summer 2011 1st Session.

What would a standard reporting tool look like using semantic technology?  In this post I will describe my thoughts on this topic.  I will base my model upon the Crystal Reports software, since I am familiar with that tool.

Like relational databases, semantic technology also involves a data source and a data model for how data interacts or relates to other data.  When using a semantic technology framework, instead of building a formal table structure to show the data relationships, information about data is encoded and stored along with the information.  There are no data tables, or data keys suggesting how information should be linked together.  The storage structure commonly used is called a triple, and the most commonly used language of implementation is called Resource Description Framework (RDF: http://www.w3.org/TR/2004/REC-rdf-primer-20040210/ ).  Within a triple, there are three pieces of information: a subject, predicate, and object.  Triples are used to encode and describe information.

In order to interpret the meaning of a triple, an ontology, or knowledge system needs to be defined for the described data.  An ontology maps out the connections for a data system, and is usually hierarchical.  For example, an ontology for mammals might denote that the class cat has subclasses of housecats, lions, and tigers.  The ontology might also provide attributes describing the cats, such as fur color or minimum and maximum weight.  The ontology provides the meaning for the data described by RDF triples.  When using the semantic web, ontologies are referred to by a URI (Universal Resource Indicator), and can be used to compare or translate one data set to another.

Another important component for semantic technology systems is a query tool.  A data set is not very useful without having a way to search through the data.  A common query tool used is SPARQL (http://www.w3.org/TR/rdf-sparql-query/).  The syntax used for SPARQL queries is similar to SQL queries used to search relational databases, with the exception that database tables are not specified since there are no tables defined for the dataset.

So where does Crystal Reports fit into this?  Crystal Reports is a reporting tool that interfaces with traditional relational databases as the information source for a report.  Reports are built by linking tables together using key data fields.  Each table is a collection of related information.  Table information is mapped out and designed in advance.  This method of reporting and interacting with data is very different from how semantic technology works.

One advantage of Crystal Reports is that the average user does not need to know how to write complex queries to be able to create reports.  Instead, the user employs a graphical interface to link tables together, extract information, and arrange the data into the desired output.  If the data source used is semantic instead of relational, the Crystal Reports user would need to know how to build SPARQL or similar queries to create reports.  Semantic data does not use the standard relational table layout, meaning the user would have no guide for how to link data together.  This would prove to be an obstacle for inexperienced report writers.

I have been thinking of a way to solve this issue.  One idea is to replicate the table structure of relational databases with the structure of the dataset’s ontology.  When importing a dataset into Crystal Reports, the ontological structure could also be downloaded, and used as a guide for linking together data to create a report.  The user could then use the structure of the ontology for report purposes.

Overall I would say that the Crystal Reports tool is not ready to harness the power of semantic technology.  More thought needs to be put into solving the problem of linking together data triples in a consistent and easy manner.  Also, greater adoption of semantic technology in datasets will aid in the development of semantic technology reporting tools.



No comments yet — be the first.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: