Categorise this! Attend the online Bite-sized Taxonomy Boot Camp in June

The latest version of Bite-sized Taxonomy Boot Camp includes talks on optimising taxonomy use for e-commerce customer experiences by Pivotree's Chantal Schweizer and by Tom Alexander on Cancer Research UK's looking at content migration. The role of the taxonomist working with an agile team is addressed by Jo Kent, Data Architect, and Jonathan Engel, InfoArk, discusses taking a hybrid approach to information management. The Boot Camp begins with a talk on autocategorisation from Synaptica’s Sarah Downs, who asks, “What does your content have to tell you, if only you could hear what it has to say?” Here’s her take on that question:

"We design taxonomies at the outset with a content audit, but once a taxonomy is in production, how often are you able to test and improve your taxonomy with real-time content feedback?

Often the ability to build a constructive feedback loop—from taxonomy to content and back again—is limited by technical constraints. Taxonomists may require data science or engineering solutions to sample content, analyse it, and deploy taxonomy improvements. That, or taxonomists themselves face learning coding and scripting languages – a daunting task on top of every day workloads.

Enter the ability to extend your SKOS taxonomies for machine tagging, paired with the ability to build a content-aware knowledge graph – a path to positive change for better connecting your taxonomy to your content.

By extending enterprise taxonomies for autocategorisation, enterprises can support human-in-the-loop machine-driven tagging. When you integrate a transparent text analytics service that can be adapted with rules, and front it with a usable UI, you can enable non-technical taxonomist users to power this process: no coding or scripting skills required; no black-box algorithms.

This approach has the advantage of addressing gaps in human tagging workflows. Many enterprises invest significant resources in human tagging of content, but the tagging is incomplete or inaccurate and limited in its ability to support further insight or functionality.

Storing the outputs of this autocategorisation in a content aware knowledge graph allows you to better explore and answer questions like the following: 

• How well does my taxonomy match my content? How frequently are specific concepts appearing in the content? Should I change aspects of my taxonomy design based on content trends I see? 
• Is my content compliant with corporate guidance or style guides? Can I detect non-compliant content through the autocategorization process? 
• How relevant is my content to business strategy? Are there concepts that should be more represented in the content base but aren’t? 
• What content is most similar? Similarity indexing can power recommendations or similar analysis."