Efficient Product Management With Knowledge Graphs

August 30, 2019 by Stefan Summesberger

Klaus Reichenberger co-founder and director of SEMANTiCS 2019 Gold Sponsor intelligent views. The Darmstadt/Germany-based company was started in 1997 as a spin-off of the Fraunhofer Society, and has become one of the leading providers of semantic technologies, knowledge graphs and AI-applications in the German-speaking region. In this interview, Klaus talks about the i-views Smart Data Engine, its contributions in the area of explainable AI and the way these interesting areas will be covered in his talk at SEMANTiCS 2019.

Intelligent views’ Smart Data Engine i-views enables people and businesses to create value from their business data even if they have only very little software engineering competences. How does this work?

At i-views, we see the real power of semantics and Knowledge Graphs in

  • Providing a rich representation that can adequately and comprehensibly capture the connections and relationships of a business domain.

  • And in putting this possibility to model their business into the hands of the subject matter experts.

Therefore, we have equipped our software i-views with user friendly modelling tools, a powerful visualisation, easy configuration of searches and views. And we offer all that in one integrated software suite – no need to screw together a whole zoo of components. This way, i-views unleashes the potential of data and turns it into economic value.

Are you interested in learning how to create value from your business data? Register for SEMANTiCS 2019!

What are the typical business areas where your approach  is the most helpful and valuable? Please give us an idea of your top success stories.

Knowledge Graphs are also automatically transforming and representing knowledge in the required context for any users. They have the capabilities to integrate structured and unstructured data and capture hidden knowledge. Therefore, an Enterprise Knowledge Graph should be the heart of a modern digital workplace. For many of our customers in big organisations, where knowledge workers are constantly looking up information in different systems – some report working with up to 40 different applications a day, the Knowledge Graph as a navigable index of the main information sources is already a huge benefit, let alone the knowledge sharing capabilities it offers.

From our point of view a knowledge graph is perfectly suited to manage a Compliance in a complex environment. Knowledge Graphs store the entire business logic of an enterprise, integrate external sources and save costs caused by regulatory and contract conflicts as well as unforeseen risks in the ecosystem. They also can assess the far-reaching impacts of decisions or changes and their consequences for any part of the company. We are working with companies who have tried to solve the problem manually, companies who have invested heavily in their employees and all of them struggle with quality issues and outdated information when bringing regulatory knowledge into the operative processes.

One of our favourite use cases evolves around intelligent product information. By using a Knowledge Graph you can build, manage and leverage networked product knowledge and thus create the best offer for each of your customers. It's about enriching your products with knowledge, your interaction, your work and the needs of your customers.

Product Knowledge Management with a Graph enables you to better meet customer needs and language with smart, connected product information. It allows you to come up with better recommendations, improve your configurator, improve the cross an up sell, and to compare your product with the competition in an intelligent way.

One of our customers uses the Knowledge Graph to represent information about 50.000 B2B electronic components with a lot of variants. Here, the possibility to represent product information redundancy-free in a Knowledge Graph using inferences and rules to pull all properties of an article together or to calculate possible variants dynamically, saves vast amounts of manual work and significantly enhances the consistency of the information. Having modelled the logic of the products in the Graph, it was easily possible to derive cross selling recommendations just by adding additional rules. Now we are working on feeding a configurator, feeding a chatbot is next to come.

i-views was named by Gartner as a company offering solutions in the area of knowledge graphs along the likes of Google, Facebook and Microsoft. Please explain from your perspective, when and why are knowledge graphs use- and helpful?

We are proud that Granter mention again intelligent views gmbh as a software vendor for knowledge graphs in the following five new hype cycles:

  • Hype Cycle for Artificial Intelligence, 2019

  • Hype Cycle for Emerging Technologies, 2019

  • Hype Cycle for Digital Workplace, 2019

  • Hype Cycle for Utility Industry IT, 2019

  • Hype Cycle for Digital Grid Transformation Technologies, 2019

More and more decision-making processes are relying on artificial intelligence. A Knowledge graph like i-views supports highly complex decision making by considering subject matters and  expert knowledge from several domains. Real world dependencies and cross-correlations are considered before recommendations are derived and explained to humans. If you like to substantiate a decision or need to provide an explanation you need an enterprise knowledge graph. Examples for explainable AI use cases are:  HR decisions on applications, loan commitment or rejection and therapy decisions in medicine. Also, think about products. Electricity Components, hazardous product or toys for kids. Here you need reliability, compatibility, transparency and traceability.

However, knowledge graphs are not only a hype but an industry proven technology since 15 years at least and there are a lot of addition use cases for the knowledge graph. Intelligent views, for example realized more than 160 knowledge graph projects in midsize and large enterprise companies to help them to maximize the value of their data and build knowledge-based AI Solutions.

What are the greatest benefits of engineering knowledge graphs with i-views?

With i-views, you can build ai - applications and adapt them in an agile manner without any special programming language or knowledge. Complex interrelationships can be easily visualized. Whether you use it as a database, an integration platform, or as a Enterprise knowledge graph – with i-views, you can let your data do the thinking.

The greatest benefits are:

  • Intelligent linkage – use semantic graph technologies to generate smart data: agile, simple, and valuable.

  • Modelling instead of programming – the i-view Knowledge Builder is intuitive to use and makes it easy to import and link data together.

  • Smart integration – using a clear and flexible data model that provides the ideal basis for integrating heterogeneous information.

  • Maximum security – a patented access rights system and access management system ensure that your precious data is protected.

  • Explainable AI –   avoiding ‘black box’ effect by solving complex problem and enabling AI to answer the question “WHY”

  • Industry-proven – by providing open interfaces, a sophisticated data deployment, a multilingual Data model and powerful meta-model capabilities with a clear focus on the business user friendliness

Among other improvements, you laid a focus on explainable AI in the latest version of i-views (5.3). This is currently a very hot topic. What can we expect from i-views 5.3 in this area and what is the main benefit for users that originates from the updates in this version?

Knowledge Graphs can offer transparency and interpretability as part of the solution, so accountability and fairness are promoted. i-views offers an AI with which humans and machines can handle each other equally confidently - transparently networked semantically. The latest version of i-views makes it even easier to combine the benefits of machine learning with Knowledge Graph. In this way semantic reinforcement learning graph can be partially automated. The extraction of information from unstructured documents in the latest version of i-views is now even more generalized - the extracted information is made transparent in the Knowledge graph. In addition, the new designer makes it easy to build AI-driven applications from a business perspective - with explain functions and visualization.

Your Talk at SEMANTiCS 2019 is entitled “Industry-proven AI applications based on an Enterprise Knowledge Graph.” What can we expect at your talk and why must we not miss it?

My talk is about getting out of the sandbox with knowledge graphs: How do we deliver real value and how do we integrate ourselves into the enterprise landscape and into the processes? And how do we sell the idea to business experts, management and IT? So, if you want to go beyond (academic) experiments in your company / with your clients you will find a wealth of real-life experience and examples in this talk.

Discuss the potential of AI and Knowledge Graphs with Klaus.Register for SEMANTiCS 2019!

About SEMANTiCS Conference

The annual SEMANTiCS conference is the meeting place for professionals who make semantic computing work, and understand its benefits and know its limitations. Every year, SEMANTiCS attracts information managers, IT-architects, software engineers, and researchers, from organisations ranging from NPOs, universities, public administrations to the largest companies in the world. http://www.semantics.cc