A datacenter story...

... so I set it up to run the 4000 question eval on Diffbot LLM overnight and went to bed. The next morning, we beat Sonar Pro.

Approximately 1.6 billion articles in the Diffbot KG, of which, 271 million have authorship attached. About 16.9%.

Every few years or so, someone tries to reinvent Wikipedia. The honor system works for a bit, but they all inevitably run into trolls and spammers. The solution to more reliable facts is a knowledge graph that links a single fact to several sources.

To the kind stranger who left us a helpful note on our docs — It's fixed! Thank you.

Looking for a chance to get some hands-on experience with Diffbot products? Come join us Jan 21-23 virtually, or at our Menlo Park office! RSVP at diffbot.com/hack

Stay tuned to find out more things you can't Google (or consumer search generally). Try it out yourself!👇 app.diffbot.com/get-started/ #knowledgegraph #semanticsearch #marketintelligence #newsmonitoring #searchDifferently

Want to learn about similar uses to the world's largest Knowledge Graph? Check out our webinar on finance, VC, and investment uses of our KG tomorrow at 1:30 PST! Register here: my.demio.com/ref/tvONd7YjBci… #knowledgegraphs #finance #marketintel #vcfunding

More data, more problems... Not that fresh, cleaned, and applicable data is bad. But data has a way of quickly becoming stale, corrupted, or just being wrong in the first place. youtube.com/watch?v=Mn2QOJtO… #datascience #knowledgegraph #ai #datacleaning #salesdata #crmcleanup

Crawlbot can be used to spider through 50 pages or 50,000. But how does it process the pages it visits? You simply tell Crawlbot which extraction API to use. And you don't have to be technical to use them. #webscraping #webdataextraction #datascience youtube.com/watch?v=PdNL9qfe…

Each fact within our billions of knowledge graph entities provides explicit origins for where this fact is from. ML enables sites with bad fact records to be skipped. And our record improves with each crawl! #knowledgegraph #nlp #dataprovenance youtube.com/watch?v=gotLKDmI…

Answers to open-ended questions, transcribed conversations, reviews, and news mentions are a few examples of unstructured data. While there can be a ton of valuable intel in these documents. They're hard to wrangle! Check out how our NL API can help: youtube.com/watch?v=2Whu9gbE…

Facet searching in the Knowledge Graph lets you return a summary view of any of the billions of entities in the KG based on a given property. Check our how in this short fundamentals video! youtube.com/watch?v=ii3CNboQ… #knowledgegraph #analysisskills #ai #publicwebdata #marketintel

Let's say you wanted to find all green energy startups in Egypt. Or what about all negative news mentions of a corporation. How would you do it? #knowledgegraphs #nlp #marketintelligence #newsmonitoring #AI youtube.com/watch?v=RmSRUqrp…

Our Knowledge Graph is comprised of billions of entities. But what are entities? And how do they fit into your public web data sourcing? youtube.com/watch?v=ft0BGYTn… #knowledgegraph #publicwebdata #ai #nlp

What do global investment indicators like news mentions, regulatory announcements, and sentiment have in common? You can find them all in Diffbot's Knowledge Graph search. Check out how in this quick vid #newsmonitoring #futures #knowledgeGraph youtube.com/watch?v=4fxIq1I7…