Imagine actively taking steps to reduce information gaps, without having to hunt for every last inefficiency out there in your knowledge management. What if you could overcome organizational silos, one regular CI/CD build at a time?
Introducing the SikaMa, a SIKA wide knowledge hub to overcome information barriers, with accurate and trustworthy information. Our approach is composed of 4 main components that are built with a view on their modularity, so as to enable future extensions.
- Firstly, the various stakeholders define what gaps are unacceptable. The construction industry, which is still in the nascent stages of digitalization could be missing critical information, like the product heirarchy numbers of SIKA products. The formulation of the gaps are used as key performance indicators (KPI), which can be objectively measured and quantified.
- The test engineer can then convert these requirements into SPARQL queries using a controlled natural language (like Behave framework).
- The third component is comprised of the knowledge base and the indexer. The indexer maintains a record of which Sika webpages, manuals and other information sources have already been indexed during the previous build and should keep track when something changes. The indexer points to the next product page to be included in the SikaMa knowledge base, which uses Ontology Web Language (OWL) representation that can be exploited by computer programs. The consistency of the current knowledge base are verified with the SPARQL test queries, thereby highlighting missing entries and other inconsistencies in the KB.
- At this stage, the fourth component, the ML-based data crawler identifies the named entities, its relationships with other entities and all the information that can be gleaned from the indexed item. This gets added to the knowledge base along with the metadata about its original source. As a final step, the improvements made to the knowledge model become auditable for the stakeholders.
As future improvements to the framework, the knowledge base can be connected to a chatbot to answer customer queries. This has to pay special heed to the confidentiality and sensitivity of the information available in SikaMa.
So, instead of firefighting when inefficiencies and information gaps arise, you stay ahead of the digital transformation curve with SikaMa.