Leading Canadian News Agency Leverages the Progress Semaphore Solution to Increase Content Discoverability, Enhancing the Customer Experience
Products:
SemaphoreChallenge
The multilingual national news agency distributes news content in French and English in many formats, including text, photos, videos, graphics and more. It’s mostly manual system of tagging, where authors tag their own content, resulted in inconsistencies, providing a less-than-optimal search experience for its customers.
Solution
- Leverage the Semaphore Knowledge Management platform for richer metadata to replace existing siloed technologies and simplify their architecture.
- Deploy an auto-tagging solution to replace a mostly manual process to increase consistency and expand tagging categories.
- Integrate the Semaphore platform with its CMS and DAM solutions for new levels of efficiency and better content delivery.
Result
- The agency’s auto-tagging solution has transformed the search experience, tying tags through relationships to other content and relevant topics and delivering superior results.
- Richer metadata has resulted in a better news product with easier, richer search experiences for its customers.
- The news agency is enjoying the benefits of a fully hosted, secure solution that is efficient, easy to use and deploy.
Full Story
Challenge
A multilingual national news agency, serving as a vehicle for Canadian media outlets, offers a wide variety of text, audio, photographic, video and graphic content to websites, web outlets such as Yahoo and MSN, radio, television, commercial corporations, government agencies, newspapers and the Associated Press (AP).
Like most news organizations, this news agency has faced numerous changes and disruptions over the years. Meeting those challenges requires staying abreast of the latest technologies. Recently, the agency installed a new content management system and a new digital asset management system. To complement those upgrades, the agency was committed to integrating knowledge management into its technology infrastructure.
As the Associate Director of Content Operations at the agency explained, “Having good metadata means that you can get the story out there to the right customer, so that it doesn't get lost, it gets searched and also gives us the opportunity to create new products and better products for customers.”
With this technology we can create new products and new revenue streams by exploiting metadata to its fullest.
Associate Director
Content Operations at Canadian News Agency
Solution
The Canadian news agency chose the Progress Semaphore™ semantic AI platform to create an automated and consistent content classification process that would include unstructured data in the form of text, photos, videos, graphics and more across both English and French. It replaces a mostly manual system of tagging that provided a less-than-optimal search experience for its customers.
Leveraging an open-source CMS purposefully built for news organizations, the system essentially consults the Semaphore APIs to help the agency auto-tag stories, search for tags and suggest new tags.
“Previous to this, journalists would use a field and an index and they would get a hierarchical set of tags that they could dig through until they found the one that they wanted,” the Associate Director, Content Operations explained. “It was all handpicked by the writer and mostly inconsistent from writer to writer and of varying degrees of granularity. We have now built and integrated an auto-tagger with Semaphore semantic AI in the background.”
Result
Today, the news agency finds itself with an advanced auto-tagging solution that is already delivering superior results.
“We have our taxonomies set up and ontology of entities, events, organizations, people and places which we previously were not tagging our stories with,” the Content Operations Associate Director said. “And we have tags in both French and English which are in a place where we can manage these terms, expand them and have relationships between them. When we talk about tagging, it’s how we've transformed our whole approach to knowledge at what is essentially a knowledge organization.”
Looking ahead, with auto-tagging of news stories in place, the news agency is planning to automate tagging for other media types like photos and videos and develop a strategy to auto-tag its vast archives dating back to the 1980s.
“As we refine our model, we'll be in a state where we're going to run the archives through, but need to have a different approach,” the Associate Director, Content Operations explained. “We're going to have to start using entity extraction because we're not going to be able to have tags ready for everything that happened in the '80s, but it would be a boon to this organization to have all our archives properly tagged and classified.”
He also explained that events are another important category to auto-tag.
“We've got events, feeds, news of events and conventions. This is part of our media monitoring service for our corporate clients,” he explained. “So, we want to relate news to events and photos and other media. We're going to be classifying these events and pulling them into the same controlled taxonomic and ontological structures that we've established in [the] Semaphore [platform].”
To sum up, he said, “With this technology we can create new products and new revenue streams by exploiting metadata to its fullest.”
We have tags in both French and English which are in a place where we can manage these terms, expand them and have relationships between them. When we talk about tagging, it’s how we've transformed our whole approach to knowledge at what is essentially a knowledge organization.
Associate Director
Content Operations at Canadian News Agency