Structure and Understand Natural Language
Convert text into structured facts about entities, their sentiment, and their relationships.
Schedule a DemoPrefer to jump right in? Start a free 2 week trial.
We enrich all of our monitored posts with Diffbot
"I’ve been working on replacing our former solution with NER in favor of Diffbot, and it has been working out great. Today, we enrich all of our monitored posts with Diffbot and it is utilized heavily in the base part of our platform. This fall we will complete the replacement of the last front-facing part and so far, it has really delivered."
Chief Technology Officer
| Entity Recognition | Entity Linking | Relation Extraction | Entity Sentiment | Commercial API? | |
| Diffbot | 92.38% | 91.21% | 83% | 71% | Y |
| 92.80% | 89.80% | 77% | 62% | Y | |
| Microsoft | 92.75% | 88.90% | N/A | N/A | Y |
| IBM Watson | 91.30% | 87.40% | 16% | 46% | Y |
| Amazon | 92.86% | N/A | N/A | N/A | Y |
| OpenAI | 88.67% | 90.85% | N/A | N/A | Y |
| 93.50% | 90.87% | N/A | N/A | N | |
| U of Washington | 92.22% | 86.80% | 75% | N/A | N |
| Stanford | 92.61% | 82.90% | 40% | N/A | N |
Breakthrough Performance
Our Natural Language API that goes beyond just identifying keywords and connects you with structured information about entities, facts, sentiment, and relationships Trainable for your custom domain of interest. Try any text in our demo now.
Non-Vaporware Demo
No cherries to be picked here. Enter your own text, or use one of our examples for inspiration.
Make Better Investment Decisions
A global Fortune 500 financial services firm uses Diffbot's Natural Language API to identify and track the sentiment toward companies related to an investment theme. Using the entity-directed sentiment, their analysts can determine how to best allocate the fund and identify risks to the investment area from government economic policy and regulation.
Monitor the News
Our Natural Language API that goes beyond just identifying keywords and connects you with structured information about entities, facts, sentiment, and relationships Trainable for your custom domain of interest.
Bring Text Documents to Life
Academic publisher JSTOR worked with HBO to create an interactive site for video transcripts of the Martin Luther King Jr documentary King In The Wilderness. By processing these rare primary interviews with the natural language API, they were able to connect the spoken word to structured knowledge in context, creating a rich new experience.
Recommend Content
Dianomi is the largest native ad network in finance with over 350 publishers. They use Diffbot to monitor topics being discussed in the pages of their network to find the best pages for each ad. Dianomi uses Diffbot’s sentiment analysis to make sure their ad placements are brand safe.
Build Your Own Knowledge Graphs
Graphs provide a compelling way to explore the complex relationships in large datasets. We analyzed the Game of Thrones anthology using our Natural language API to build the largest geneology of the Game of Thrones world built to date.
Add Your Own Entities And Relationships
Tracking a specialized set of entities and relationships? We can train our NL API around entities and relationships you care about. Our customization options can take you from lacking out-of-the-box textual analysis to 100% coverage. Customization occurs quickly within the course of general introductory discussions.
