Facebook buys UK-based AI Startup Bloomsbury

IndraStra Global

Facebook buys UK-based AI Startup Bloomsbury

By IndraStra Global News Team

Facebook buys UK-based AI Startup Bloomsbury

Image Attribute: iStock

On July 2, 2018, TechCrunch reported the latest acquisition news of Facebook buying a British artificial intelligence (ai) startup named Bloomsbury AI. According to the same report, the acquisition would cost between US$23 to US$30 million.

The tech pundits are claiming this acquisition to beef up human editors and problem-spotting AI  at Facebook, as it addresses concerns about election meddling and abuse of the platform.

Bloomsbury AI is an UK-based data analytics company that has created - Cape - an API based AI tool that reads text documents and answers questions about their contents. Its proprietary algorithms will prove a useful tool for Facebook's battle against fake news.

Cape API enables developers to convert "text" to "knowledge" by applying powerful deep learning neural network models in an easy to use API. Simply paste text or upload text documents and ask complex questions like you would to a human and get immediate answers.

So, how does it works?


STEP 1: Login at https://alpha.thecape.ai/

STEP 2: Go to documents tab.

Cape AI - Screenshots
STEP 3:  Just copy and paste a text of an article or the full article or upload a document itself.

Cape AI - Screenshots

STEP 4: We selected the 1st option, copy and paste a text from an article -  A Brief Note on India's Defense Manufacturing Sector

Cape AI - Screenshots

STEP 5: Copy and Paste.

Cape AI - Screenshots

STEP 6: Save

Cape AI - Screenshots


STEP 7: Go to interactive mode, and type in a question! 

Cape AI - Screenshots

STEP 8: "What is the main subject of this article?" (Follow the simple natural language flow, without wondering about the grammar)

Cape AI - Screenshots


STEP 8: Cape AI pulls the subject line/crux of the article, semantically. 

Cape AI - Screenshots


STEP 9: The result can be further elaborated with the opening of the whole article in a separate pane. 

Cape AI - Screenshots


So, how it's going to be used?


The question is how Facebook is going to use Bloomsbury's Cape API?, Well, as per our understanding of this tech, this particular tool can be used by former's moderation teams to identify "fake news" or any content which violates its content policy at first place. 

It would be purely a mixture of AI-based automation with the timely human intervention - which can reduce the existing response time between the identification of such contents to completely weeding it out by taking necessary actions.

So, what's the future?


The more we connect the content over the web with semantic ontologies, the more we are able to deploy AI at unprecedented levels. As of now, we are looking at content monitoring part but there are other examples too from automated content - Associated Press, The Washington Post, and several other uses machine learning to create basic sports and financial news - through to analyzing social media.

Currently, the whole AI-based reporting (as well as the monitoring) is in a nascent stage and also very much hard to quantify how much engagement, ad revenue and subscriptions can be attributed to those AI-reported stories. Publishers, as well as the content syndicators like Facebook and others, have to still figure out what the economic impact's going to be when they automate end-2-end cycle of the media industry.