Here at Brightspot we pride ourselves on the fact that we save reporters, photographers and editors time when they adopt our Brightspot solutions. Leveraging Brightspot to accomplish their everyday tasks is easier and faster compared to other solutions they are migrating from.
We also continue to push the boundaries of the platform to benefit users that depend on Brightspot to accomplish great work. One of the initiatives we’re driving forward is providing in-context help at different stages of the publishing process for editors and reporters. We started that review by focusing on one of the key actions that newsrooms rely on Brightspot for, authoring an article.
We asked ourselves this question: How do we provide in-context help during the article authoring workflow? The answer is found in our decision to integrate with the artificial intelligence/machine learning service from Amazon Web Services (AWS), Amazon Comprehend.
What is Amazon Comprehend?
So, you might be wondering: What is Comprehend? This is how Amazon explains it:
Comprehend uses machine learning to help you uncover the insights and relationships in your unstructured data. The service identifies the language of the text; extracts key phrases, places, people, brands, or events; understands how positive or negative the text is; analyzes text using tokenization and parts of speech; and automatically organizes a collection of text files by topic.
Applying Comprehend to Brightspot
We realized those types of insights would be most useful to digital teams in two places: tagging and taxonomy.
In the article publishing process, any given asset will pass in front of the eyes of several different people—from reporters and editors to copy editors to SEO editors—and somewhere along the way, tags are added. These tags have several different jobs, from boosting SEO performance to “pushing” the article onto relevant tag pages, making it available for users who are interested in that tag.
The action of adding a tag requires both fully understanding what the article is about and how users who are interested in the topic (or related topics) are going to search for it. On top of that, in order to tag an asset appropriately, Brightspot users need to be knowledgeable of their organization’s entire taxonomy in order to even know which tags are most relevant. That process grows more difficult over time as the taxonomy expands and more tags are added. Editors are then forced to go back and forth referencing the article text and searching for all relevant tags.
This is where our implementation of Amazon Comprehend comes in. By leveraging Comprehend’s power to analyze and extract key phrases and entities from text, Brightspot suggests* tags that are contextual to the story, and the editor or reporter can simply click to add them to the story. That means even more time saved for teams to focus on producing high quality content. (*We say suggests because we understand that humans know better than the machines when it comes to curation.)
Now Brightspot can offer Suggested Tags functionality on any text-based content asset that newsrooms need to manage in Brightspot.
And our drive to continue improving the functionality of Brightspot means we’re expanding the idea of suggested tags to other content assets in Brightspot as well. So teams that contend with significant numbers of visual assets, like images or video, can also utilize suggested tags, saving them time and effort.
Artificial Intelligence/Machine Learning in Brightspot’s future
When it comes to working with AI/ML tools for your use-cases, they are just that, tools. Tools don’t simply solve your problems for you, it takes truly understanding your use-case and the problem you are trying to solve.
At Brightspot, understanding our customer’s use-cases and problems to provide efficient and valuable solutions is at the core of what we do. As we continue to work with our partners to explore the possibilities of AI/ML, we’ll discover even more solutions.