For this interview, we talked with Ian Davis, a knowledge management professional with extensive experience in information architecture and data science. Ian is currently Global Technology and Information Manager at SGS.
When did your interest in Information Architecture begin?
ID: From an early age I enjoyed organizing information. As an avid reader, I always liked to have my books in order on the shelves of my bedroom! This enjoyment of organizing information led me to think library or information work was worth exploring. After a bit of digging, I was advised to take a degree then top up with a specific post-grad qualification. I chose History then a Master of Information Studies.
I assumed I would move into academic librarianship. However, a lecturer connected me to a friend who was setting up a project team working for a new company Bill Gates was funding to digitize images from around the world. I took a one year contract, based in central London, to work as part of a team manually cataloguing 60,000 digitized images.
Following this, I worked for a variety of companies including Corbis, Photonica, BUPA, and Dow Jones. I first met Dave around 2000 when I licensed the Synaptica KMS for Photonica and I went on to work for him at Dow Jones. I believe we were one of the first subscribers when the KMS software was first launched in the early days.
My time at Dow Jones was especially interesting. I worked as a taxonomy consultant developing and delivering various projects to key clients. This work took me to many places including Singapore and Cape Town – where I was stranded for a while due to the Icelandic volcanic explosion of 2010.
”Graphite worked better. I find it more practical, user-friendly, and flexible. Graphite allows you to be more person-centered on the front end.
Tell us about your experience in using Graphite.
ID: During my career I’ve spent time investigating how Graphite from Synaptica produces rich semantic structures that create knowledge graphs and more. I am a great advocate that modelling the underlying semantic structures of business domains benefit companies and their customers by producing a more data-driven, customer-centric approach.
I’ve looked at a number of tools to build ontologies, such as the free tool Protégé. The downside for me was that particular one felt too academic and not easily accessible. Graphite worked better. I find it more practical, user-friendly, and flexible. Graphite allows you to be more person-centered on the front end.
Modeling an ontology is a complex business. It’s an intellectual challenge that needs to be solved in a way that creates something that works for an organization. Graphite’s flexibility is certainly significant. The tool is pretty easy to use with lots of potential to connect ontologies to various business models or domains.
I would like to see ontologies and knowledge graphs adopted more widely. I believe now is a great time for businesses to start to work in this space – creating concepts and making specific relationships between them so stakeholders can start to see real-world business benefits.
Are there any challenges related to ontology development and adoption?
ID: The reality is when we talk taxonomies and ontologies people can glaze over; the subject feels pretty dull to some. It can be complex and difficult to explain what it is. The first reaction when you explain it is often to ask, “What’s it going to do for me?”
Trying to find the right people – those who will benefit from it – then explaining the concept, selling it to them, working out how it will benefit them, and then setting up a trial and getting things up and running: these are difficult but rewarding challenges of adopting these technologies.
Synaptica is well placed to help with ontology development. The Synaptica team and myself go back a long way and I’ve always found them knowledgeable and easy to work with. They care about what they do, and I find that they deliver. You don’t have to navigate multiple layers of company structures or complex support ticket routing rules to get to people who can help. You can send questions directly to the people that count. When you are shaping projects at any scale it’s a real benefit to send queries and receive a quick response from the experts.
What’s your favorite feature of Graphite? What do you like best about it?
ID: Often when you work with a new tool the experience can be frustrating, and a lot of the functionality can be confusing. I like the way Graphite is laid out, it’s easy to understand and simple to navigate. It’s nice and easy and doesn’t require a massive amount of time commitment to learning how it works. There is still a learning curve but I got the hang of it pretty quickly.
I would say it’s easy to set up and being web-based, it is simple for people to access and review work as it progresses. Starting to build schemes, create concepts and predicates, then link these concepts and predicates together can all happen very quickly. I also love the fact that Graphite can be easily connected to linked open data – creating links between what is being done internally and what others are working on. Essentially, pulling relevant data that somebody else has developed. Ontology is key to knowledge graphs, but you only really get a knowledge graph when you link out to external data. Graphite is perfectly positioned to do that.
”Ontology is key to Knowledge Graphs, but you only really get a Knowledge Graph when you link out to external data. Graphite is perfectly positioned to do that.
What advice would you give to someone who was developing an ontology project?
ID: Take time to plan before you build. You will benefit if you spend more time planning and preparing at the start. Avoid jumping in or making it up as you go along. Think logically and as clearly as possible about what you’re working with. Consider the entities, relationships and concepts, what they are and more importantly, how they relate to each other and why they need to.
Try to be accurate and consistent and document your thought processes and plans. This way you can monitor what you’ve done, where you’re going, and where you are. It also makes it easier to hand over to someone else if you ever need to.
Make sure you try and connect it to your business. Try to learn as much as you can about the areas of your business that relate to the work you’re doing. Develop good relationships with subject matter experts and business people. Don’t forget that the ontology is a business tool; it must work for the business, not the other way around. You can’t change the business to fit the ontology, your ontology has to fit the business.
Lastly, don’t be too hung up on making everything intellectually perfect. That’s rarely what’s needed. It needs to be good, but more importantly, it needs to be used, and then you can improve it.
”You can't change the business to fit the ontology, your ontology has to fit the business.
What do you think are the biggest challenges for the industry in the future?
ID: I think the issue is ensuring others understand what we do and the value of our work. Generally, people don’t often appreciate what we do and why they need to care about our work. When a business is formed, it’s accepted that you need Finance, Marketing, Sales, and Legal. Information and knowledge are not automatically considered at the same time. With the digital revolution, we can see that IT has become normal – an accepted part of every business, but the information sector has not reached that point. It’s still seen on the edges of too many organizations.
We need a growing number of business people in senior roles to see information as a key function at the heart of the business. Information and knowledge need to be perceived as an investment rather than an additional cost. Some businesses are beginning to comprehend all this but too many senior staff still think they solve business problems by buying pieces of magic technology without long term investment in making these technologies work for customers.
”In the industry, we need to see information as a key function, at the heart of the business. Perceived as an investment rather than an additional cost.
What makes a good taxonomist?
ID: A taxonomist needs to be able to consider and approach things in a logical way. It’s useful to have an interest in the information and a love of order and detail. You need to care about your projects and getting them right. If you’re someone who doesn’t like detail, then you will struggle to do this type of work. Similarly, you need to understand how one thing relates to another – to be able to break things down to a granular level but not lose track of the bigger picture.
A consultant’s mind and skill-set are also very useful. You need the skills to be able to listen, understand requirements, and present useful solutions. It’s an interesting mix, a combination of introverted intelligence with extroverted people skills.