Professional Services
Structured Dynamics LLC provides these professional services. Please contact us directly for a formal Statement of Qualifications.
- structWSF and Web Services Development
- conStruct and Drupal Modules and Installations
- Project Evaluation and Planning
- Legacy Structure and Data Conversion
- Ontology Development and Mapping
- Instance Record and Dictionary Creation
- Information Extraction
- Architectural Design and Deployment
- Linked Data Training and Education
As
its original developer, Structured Dynamics is well positioned to
provide further development assistance for the platform-independent structWSF Web services framework. Such services might include:
- Development of entirely new Web services
- Expanding framework support to other CMS systems or RDF triplestores
- Wrappers around existing software to integrated it with the framework
- Enhancing the scope or parameters supported by current services, or
- Adding more supported conversions or serializations.
In addition, Structured Dynamics is available to consult or assist structWSF development efforts by others.
Structured Dynamics is available to develop new conStruct modules or to assist in designing and deploying Drupal installations utilizing the system.
Custom module development may extend to any of the modules or
functionality within the existing pool of contributed Drupal modules.
Also likely is exposing the unique aspects and capabilities of
structured data to the Drupal system.
Certain areas in keeping with Structured Dynamics' own development
interests may be cost-shared with the client. Present priorities
include content harvesting, structured data extraction, graph analysis
and visualization, and integration with natural language processing.
Structured Dynamics combines its knowledge of linked data resources and structures with detailed evaluations of existing data resources and structure. Public Web resources -- in existing linked data format or convertible to it -- are evaluated in conjunction with private or limited-access sources (such as from trade associations or standards bodies).
Enterprise source evaluation may extend into many databases and schema, akin to traditional evaluations for large-scale federation or warehousing efforts. Project evaluations can either be limited to the diligence associated with an initial proposal development or extend to a full-blown consultation and report.
A linked data deployment may require conversion of legacy structure or data. Existing and meaningful structure may be found within:
|
|
|
Any of these may occur in various serializations (data encodings and syntax) and range from private and local to public. Some may arise from third-party APIs.
Converting legacy sources to linked data can be viewed as a pipeline involving both syntactic and semantic conversion, similar to a software driver. In a linked data context, these converters are sometimes called “RDFizers”.
Structured Dynamics routinely develops new converters and welcomes unusual formats. We have experience with nearly 100 forms of existing converters, many of which we initially designed.
Ontologies supply the structure for relating information to other information in the semantic Web or the linked data realm. Ontologies play a similar role for the organization of data that is provided by relational data schema. Because of this structural role, ontologies are pivotal to the coherence and interoperability of interconnected data.
Ontology development starts with re-use of
“standard” ontologies, supplemented with quality
domain-specific ontologies, if such exist. Only after re-use
is fulfilled does Structured Dynamics recommend to build a
new ontology.
Structured Dynamics is a leading developer of current
ontologies (see related tab)
and is fully conversant with all mapping and conversion
techniques. A key aspect of ontology development is coherent
design to assist interoperability. We are experienced with
many tools, including use of the Cyc knowledge base, for coherency
testing.
Please contact us for a white paper on our ontology building process.
Instances and named entities are the notable objects, persons, places, events, organizations and things of the world. Each instance is characterized by multiple attributes and is related to one or more classes (concepts) to which it is a member. Instances do not set the structure of the domain, but populate that structure.
Structural Dynamics converts existing records into conforming instance records and “dictionaries” – lightweight data record structures – for each contributing source of instance data. These dictionaries also contain the attribute relationships (predicates) for each instance type. A relational database (RDB) is often the best performing system for this data.
In a linked data setting, these structures are known as
ABoxes, and can be
easily mixed and matched. Mapping entities between
dictionaries enables attribute information for the same
individuals to be combined. For presentation purposes (and to
aid disambiguation), each type of instance is also assigned to a
single SuperType, of which there are about 30 kinds.
These instance dictionaries are used in lookup and matching when
entities are encountered (via information extraction or
otherwise). The processing of these entities from new sources
must also account for whether that entity is already in the
system, but in a different dictionary, which may be a public
or private source of entity instance data.
Structured Dynamics extracts either subject concept or instance information from content “on the fly”. This process may work from defluffed complete documents or Web pages (that is, with navigation and other boilerplate fluff removed leaving desired content) or from window- or sentence-length text ‘snippets’.
Information extraction may also apply to semi-structure within documents, such as within tables or patterned text. The company has a variety of wrappers and extractors for these purposes, some of which work within the DOM structure used by browsers.
These Structured Dynamics information extraction and named entity recognition services are used when creating structured data from unstructured text or mapping between datasets. A variety of disambiguation approaches or algorithms may be employed.
Structural Dynamics is an advocate of Web-oriented architecture and modular design. In WOA, discrete functions are packaged into modular and shareable elements (services), then made available in a distributed and loosely coupled manner using Representational State Transfer. REST provides principles for how resources are defined and used with simple interfaces without additional messaging layers. REST is a foundation to the HTTP protocol and a key reason for the success and scalability of the Web.
As first articulated, WOA was simply the RESTful expression of services. However, with linked data, we now see a consistency of design and architecture across services and data.
In terms of modularity, Structured Dynamics designs its systems into separate concept-only ontologies (TBox) and instance record dictionaries (ABox). This split provides better performance, modularity, flexibility and scalability than standard approaches that blur these differences.
Structured Dynamics also provides deployment assistance for a number of platforms, with a special emphasis on OpenLink Software's Virtuoso server.
Structural Dynamics principals are noted thought leaders and bloggers and communicators in the linked data space. We can provide training and education in RDF and linked data from introductory briefings to multi-day courses.