Galen Jones, The Open University
The 3rd Taxonomy Boot Camp London was held on the 15-16 October and proved to be an invaluable opportunity for networking with people with very mixed specialisms. System vendors, content providers, consultants and practitioners – from varied sectors, including health, education and agriculture were in attendance. The two days featured more conversation than you usually see at conferences and this helped surface new ideas for future work and collaboration. The theme for this year’s boot camp was Making taxonomies go further – and for me this was truly delivered with an agenda packed full of empowering, thought provoking and informative sessions.
In his opening keynote speech Paul Rissen, Product Manager – Springer Nature, UK demonstrated that whilst the web removes barriers to communication, conversations often degenerate into opposing sides and that this can open the doors to malicious exploitation intended to distract, discourage, drown – to overwhelm a population with different narratives and generate apathy.
Paul explained that the notion of questioning everything and trusting no-one results in the positive force of curiosity becoming endangered, and warned that new strategies to combat this are required. Fact checking does not always work for audiences who feel inundated with information, so there is a very real challenge to update the methods and models on how we manage our information environments. To this end, he set out 13 rules for designing better information environments. These ranged from tolerance, to identifying different forms of falsehood (categorized as disinformation, misinformation and mal-information), and valuing and reflecting diversity. I found this presentation both useful and hugely encouraging.
As part of my job, I am currently investigating the options for implementing SharePoint Online (SPOL) managed metadata to support enterprise search and information retrieval. For this reason, an appraisal of this functionality and outline of the options and design configurations by Agnes Molnar, CEO – Search Explained, was especially useful. For anyone working in large knowledge based organisations, and particularly for newer members of staff, it is easy to relate to the feeling of being overwhelmed by the masses of content; not knowing what is stored where and how to access it.
Enterprise Search tools are only part of the solution to this issue. In SPOL hierarchical, managed metadata term sets, site collections and columns with defined properties are integral to delivering the kind of search and retrieval we desire. It was also interesting to note the option for hybrid taxonomies for combined on-premises and cloud customers which can be a necessity for large and complex organisations. Randy Perkins-Smart, Director – Qaixen, UK reminded us how creating taxonomies and manually applying metadata to content can be a costly and time intensive process, often with imperfect results. There are significant gains to be achieved through automation. Randy described that the use of PowerShell scripts to assign tags for example, can also be flawed, so the use of specialist software is often required. Many providers were at the event and helped make it so successful.
It was fascinating to hear the work Ben Miller, Senior Platform Capability Manager – Wiley, UK is leading to challenge the company’s existing business models by harnessing the power of knowledge models and linked data. With over a billion triples in the knowledge store, Ben has been developing prototypes that utilize this store to demonstrate a range of new use cases. These emerging services are very different to the licensing of content that we are familiar with from Wiley and I feel have the potential to offer significant benefit to the research community. The ability to locate people working in specific areas of specialist activity, to build knowledge about funding organisations, to identify people who may be able to co-author publications or conduct peer reviews were some of the potential uses demonstrated in beta systems. These novel opportunities to disseminate knowledge through innovative new service platforms are the product of linked data, and I feel confident this area will see dramatic growth.
On the second day we were all in high spirits – although I imagine some were nursing a fuzzy head following the drinks reception and ISKO chill out event at the neighboring pub the previous night! Fortunately, delegates were gently eased into the second day’s proceedings which began with a book giveaway raffle, with some great titles up for grabs including the excellent Digital Literacy Unpacked by my colleagues Katherine Reedy and Jo Parker from The Open University.
Soon enough, the second keynote speech was underway. Tom Reamy, Chief Knowledge Architect – KAPS Group, USA and author of ‘Deep text’ covered the benefits of taxonomies and various techniques using numbers, stories, and text analytics to help advocate and enhance our work. Whilst there may be an increasing understanding of what taxonomies are, there are real concerns about the quality, application, and maintenance of too many taxonomies. So, as a profession we still have an important role to play to educate and address these failures.
With so many different kinds of taxonomies, internal advocacy is a key activity which needs to be tailored to different audiences within our organisations. Numerical studies that demonstrate the benefits returned on investment in taxonomies are a good place to start. Tom outlined the financial cost of 1) time used doing ineffective searching, 2) unnecessary recreation of documents and 3) the consequences of bad decisions and poor quality work (based on not finding relevant resources), which has been calculated to being in the region of $22.5 million per year per 1000 employees in recent IDC studies.
Belief factor is difficult to achieve in numbers. As Tom explained, for the human brain, positive events persuade better than negative, statistical losses. Stories are one of the best methods to get our messages across and are heavily used in my organisation’s planning activities. Case studies demonstrating the role of taxonomies in achieving business objectives; increasing richness in discovery; delivering better, more complex search, enabling more granular reporting, developing a common knowledge across company even where different languages are used are some of the possible scenarios outlined in the presentation.
Social media has become a key communication channel – useful case studies could look at how taxonomies can be used to understand what customers are saying; gauge levels of satisfaction, monitor the effectiveness of marketing campaigns and manage brand perception. This presentation really drove home the theme of making taxonomies go further -with other potential benefits from the application of taxonomies ranging from enhancing customer support to fraud reduction; achieving competitive advantage to reputation management. We were, however, reminded of the challenge to demonstrate that benefits are derived from both the application and the taxonomies.
The major flaw though is the gap between taxonomies and content. Amongst the obstacles Tom discussed were the cost of tagging documents with taxonomy nodes and the differing level of human understanding of subjects and/or business processes, variations in categorisation skills, etc. Text analytics was presented as the solution. Combining text mining, entity extraction, sentiment analysis and summarisation to perform auto categorization do seem to offer a comprehensive and workable solution.
Theresa Regli, Director – Vox Veritas Digital Ltd, UK gave a presentation that drew inspiration from the annual migration of the Monarch Butterfly from Canada to Mexico. She used this epic migration to highlight themes of regeneration and rejuvenation during the journey which served as a powerful analogy for those of us who are involved in migrating metadata, content and other assets. Next up was an image of emigrants departing on a ship (the painting ‘Parting cheer’ by Henry Nelson O’Neil) which was used to further exemplify the theme of migration and its correlation to our work – the need to leave things behind; the difficulties encountered during the journey and how we adapt to these challenges.
Theresa reminded us that whatever it is we are migrating, we should always be thinking about how it will be rendered and used once it has arrived at its destination. Her own relocation from the US to the UK and how Amazon’s Alexa content and metadata failed to recognize her new geographic context following migration served to demonstrate that technology doesn’t solve this problem.
Our second migration story from nature featured the Arctic tern’s annual voyage from the Arctic to Antarctic. A team somehow managed to attach to the birds and then retrieve a year later a small geolocator in order to learn more about their incredible voyage. The major discovery was that the birds do not travel directly to their destination, they stop and enrich part way through their journey.
As a result of the research into the bird’s migration, parts of the ocean were identified for conservation measures. Theresa compared this with the need for metadata stewardship: determining a metadata strategy, understanding the destination storage structure, mapping old structures to new, planning for and communicating about system transition, validation and quality assurance.
And so, as our assets are moving between systems, Theresa suggests we too need to consider how we track and enrich them during transit. Moving everything isn’t wise; her advice is to clean out the R.O.T. –(redundant, outdated or trivial content) before moving between systems; run the two systems in parallel if feasible; conduct thorough testing; and drive adoption by making sure people know how to search and use the new system. Migration doesn’t happen quickly: mapping metadata and system consolidation are rarely straightforward in my experience.
This has been a brief outline of just a few of the many excellent presentations that featured over the two days. The event was largely informal in nature, thanks to the expertly tailored programme and mix of professionals attending. It was great to see so many people enjoying themselves, in deep, enthusiastic conversations. My thanks go to organisers of TBCL2018. Having been designed to satisfy an unmet need for professionals in this field, the learning opportunities and focus on what the people in our community are working on, our challenges and opportunities ensured this goal was delivered.
K&IM Refer 35(1), Winter 2019