What are the system requirements and dependencies of WissKI?
The WissKI system is a set of modules which extends the current stable Version of the Content Management System Drupal (at the time of writing this FAQ this is 8.4.5, however we always use the most current one for testing) by a virtual research environment. The dependencies for Drupal 8 are documented at the Drupal website (https://www.drupal.org/requirements). Additionally, the WissKI system has as semantic backend a triplestore instead of Drupal’s database.
To put it in a nutshell, the WissKI system has got the following dependencies:
- Drupal 8
- Any XAMPP package with the following features or a manual installation of:
- PHP >=5.6
- MySQL 5 or higher respectively alternative PostgreSQL 7.1, MariaDB, MSSQL or other supported database systems
- Any valid triplestore which has a SPARQL 1.1 interface. We use for testing:
What programming languages, development environments, repositories are used?
The code of the WissKI software currently amounts to ca. 40000 lines.
How is the current status of the documentation?
The code documentation is based on Doxygen and is continuously updated and extended. As a user documentation there is the website http://wiss-ki.eu with detailed information about the software infrastructure inclusive Installation Guide. Furthermore, the theory and the used technology around WissKI are described.
What are you planning for the future development of WissKI?
In the scope of the current project the virtual research environment will be developed further based on the WissKI software and in cooperation with the partner institutions a competence center will be established. In this regard, the museums using WissKI (Germanisches Nationalmuseum, Zoologisches Forschungsmuseum Alexander Koenig) provided positions to guarantee the utilisability of the software. On this basis, even after the end of project the maintenance and support of the software are ensured. For long-term proposal and support of non-profit properties the association IGSD - Interessengemeinschaft für semantische Datenverarbeitung e.V. (http://www.igsd-ev.de) - was founded. There will be a spin-off to provide further services. Moreover, the WissKI software is Open Source. Therefore the source code can be completely reviewed and also edited by a third party. You can find the complete WissKI development on Drupal GitHub (https://www.drupal.org/sandbox/fichtner/2304911).
To what extent is the future support for single and individually adapted WissKI instances ensured?
Users of single and individually adapted WissKI instances are primarily personally responsible for the maintenance of the instances. However, further future support can be managed with contractual agreements.
How do you guarantee in the future to adapt WissKI instances promptly after updating external basic modules?
Adapting the WissKI modules after updating external basic modules is in the interest of all WissKI users and especially of the museums, and therefore prioritized.
For adapting to external third party modules the users are personally responsible. Additional services can be handled by contractual agreements.
How will current and future changes of the basic model (CIDOC CRM) be added to WissKI?
During the installation of the WissKI system the ontology to be used is loaded dynamically. The administrator of the WissKI system is able to update the ontology at any time manually.
Maintenance and further development of the ECRM implementation is task of the competence center mentioned above.
How do you guarantee security on running systems?
The operator is personally responsible for the security of the operating system and installed server software like e. g. Apache, MySQL, Drupal.
Regarding WissKI, it is in the interest of the museums to provide for a running system so that WissKI security measures obtain top priority.
What about the response time to eliminate safety gaps?
Being in the interest of the museums to provide for a running system security measures obtain top priority. Consequently, security holes will be eliminated promptly.
How does the schedule look like for updating the out-dated system?
Old WissKI Systems (based on Drupal 6) can be upgraded by simply installing Drupal 8 and providing the system the same triplestore interface. WissKI stores all important data in the triplestore - so it can be transfered easily. It typically takes us about an hour to upgrade an old system.
What storage mechanisms do you use?
The WissKI system stores its data selectively in a triplestore which provides a SPARQL 1.1 endpoint. The physical data management is ceded to the user and thus is at this point flexible.
What did you experience regarding consumption of resources and time response with mounds of data in which dimension?
Consumption of resources and time response of the WissKI system primarily depend on the underlying basic system Drupal and the used triplestore. Drupal demonstrates its power as content management system e. g. on the websites of the White House (https://www.whitehouse.gov/) and MTV (http://www.mtv.de/). In the scope of the project Synat (http://www.synat.pl/) the Poznan Supercomputing and Networking Center integrates the data of the polish libraries, archives and museums into a SPARQL end point using the triplestore BigOWLIM which answers live despite this amount of data. Based on Drupal and by means of an appropriate data back end, e. g. BigOWLIM, WissKI is also applicable for great amounts of data.
Did you experience system crashes?
The WissKI system consists of a set of modules which extends the Content Management System Drupal Version 8 by a virtual research environment. Respectively, the data and data acquisition forms are provided via the browser to the computer of the user. The WissKI system is not influenced by crashes of the user’s computer. In case of a server crash the current mechanisms based on Apache and MySQL provide data consistency.
Are there recovery methods?
According to the system configuration of the server the current recovery methods of the data base system respectively server system are used.
Who is modelling the custom-designed ontology for a WissKI instance?
The user is responsible for the modelling of the custom-designed ontology. Generally, it is possible to handle additional services by contractual agreements.
What is the relation between the custom-designed ontology and the Erlangen CRM?
The respective custom-designed ontology is a sub-ontology of the Erlangen CRM (ECRM). The custom-designed ontology specializes the concepts and properties of the Erlangen CRM, so that they are of use for the domain of the application.
How open and expandable is the data model?
The CIDOC CRM (ISO 21127) provides for all concepts and properties of the custom-designed ontology to be modelled as sub concepts and sub properties of the CIDOC CRM. Since the CIDOC CRM perceives itself as a top level ontology, there is wide scope left. In line with the custom-designed ontology, the user is able to decide by himself which concepts and properties he wants to model for the domain.
The WissKI software does not necessarily depend on the CIDOC CRM, however, the use of the ISO standard is recommended due to long term availability and security as well as data portability.
How do you control the consistency of the underlying ontology?
The user is responsible for controlling the consistency of the underlying custom-designed ontology. The consistency of the ISO standard is guaranteed by the CIDOC Special Interest Group (CIDOC SIG). For checking the consistency of the ECRM and the custom-designed ontology tools like e. g. Protégé (http://protege.stanford.edu/) can be applied.
Who can modify the ontology (Administration of user rights)?
Who can modify the ontology (Administration of user rights)? The ISO compliant version of the CIDOC CRM ontology can only be modified by the CIDOC SIG. Modifications of the ECRM are carried out via Open Source on GitHub (https://github.com/erlangen-crm). Each user can make modifications of the custom-designed ontology by himself. The custom-designed ontology is reloaded via the administration interface of the WissKI system.
What standard interfaces provides the WissKI system?
The WissKI software is able to import data via the ODBC interface of any SQL server. Furthermore, data can be loaded via the import interfaces of the triple store e. g. in OWL/RDF, RDF/XML, N-Triples, Turtle, SPARQL + SPOG, Legacy XML, HTML tag soup, RSS 2.0, Google Social Graph API, JSON (see https://github.com/semsol/arc2/wiki). The system is able to import normative data in SKOS format.
As export interfaces WissKI provides natively all export formats of the triple store, e. g. OWL/RDF, RDF/XML, N-Triples, Turtle, SPARQL + SPOG, Legacy XML, HTML tag soup, RSS 2.0, Google Social Graph API, JSON. Additionally, data can be exported in Excel. Normative data are exported in SKOS format.
Is it able to use external templates, vocabularies and how are they synchronized?
External templates can be integrated via the custom-designed ontology. As vocabularies you can use any normative data like Name Authority Files or Getty TGN (subject to license) in SKOS format. These are manually synchronized by the user.
What possibilities to import or export large amounts of data are there?
Large amounts of data can be imported via the ODBC interface and the triplestore interfaces and exported via the triplestore interfaces.
What tools for modifying large amounts of data provides WissKI?
Basically, due to semantic modelling of data in the WissKI system it is generally non-essential to make modifications on large amounts of data sets, because each identity only exists once in the system. However, modifications on large amounts of data can be made via the triplestore by means of SPARQL.
Which testing mechanisms are there for input/import?
The user inputs data via the current Web widgets. In forms the auto-complete function supports data input. The user is self-responsible for checking import data.
What does the data storage for the text editor look like?
The text editor writes the data optionally into the Drupal database as well as into the triplestore.
What is the relation between free text input and structured data, normative data, vocabularies?
The WissKI system integrates free text with structured data just like the known Wiki approach. Moreover, the WissKI System supports tagging entities in free text. Data storage of structured data and entities tagged in free text is kept consistent and saved into the triplestore. In this way, saving facts twice is avoided. Gathering structured data as well as tagging entities is supported by loaded normative data and vocabularies.
For the old FAQ on Drupal 6 please go here: WissKI on Drupal 6 FAQ here.