10 New Market Intelligence Queries From Diffbot's Knowledge Graph [Webinar]

Summary of "10 New Market Intelligence Queries From Diffbot's Knowledge Graph [Webinar]"

Short Summary:

This webinar explores the capabilities of Diffbot's Knowledge Graph (KG) for market intelligence. The KG is a massive database of entities, facts, and relationships scraped from the web and structured for querying. It includes billions of entities, including organizations, people, products, and more. The speaker demonstrates 10 new queries that leverage the KG's advanced features, such as granular industry categorization, employment analysis, revenue estimation, look-alike queries, and sentiment analysis. These queries provide valuable insights for market research, competitor analysis, and trend identification.

Detailed Summary:

1. Introduction (0:00-2:10):

The speaker introduces Diffbot's Knowledge Graph, emphasizing its vast size and comprehensive coverage of entities and relationships.
The KG is described as a powerful tool for organizing real-world data, offering flexible schemas and mimicking human thought processes.
The speaker highlights the KG's relevance to market intelligence, emphasizing its ability to automate fact accumulation from the web.

2. Exploring the Knowledge Graph (2:10-10:30):

The speaker introduces the visual query editor in Diffbot's app, demonstrating how to build queries using a user-friendly interface.
The first query showcases the KG's granular industry categorization, allowing users to identify computer vision companies with specific criteria like employee count and investment value.
The speaker emphasizes the KG's ability to provide detailed information on organizations, including their revenue, investment history, and location.

3. New Queries for Market Intelligence (10:30-23:30):

The speaker presents a series of new queries, demonstrating the KG's capabilities for market intelligence:
- Query 2: Analyzing employment data by job function and category, allowing users to identify hiring trends within specific organizations.
- Query 3: Distinguishing between reported and estimated revenue for organizations, providing insights into financial performance.
- Query 4: Using "look-alike" queries to identify companies similar to a given organization, aiding in competitor analysis.
- Query 5: Faceting queries to explore data coverage and identify trends, such as investment activity and article sentiment.
- Query 6: Analyzing investment activity by industry and time period, revealing investment patterns of major companies.
- Query 7: Identifying hiring trends within organizations, including job functions and employee turnover.
- Query 8: Analyzing entity-level sentiment, providing insights into the sentiment surrounding specific entities within articles.
- Query 9: Utilizing the "description" field to identify organizations related to niche offerings, expanding market research beyond traditional industry categories.

4. Conclusion (23:30-24:00):

The speaker summarizes the webinar, highlighting the power of Diffbot's Knowledge Graph for market intelligence.
The speaker encourages viewers to reach out for further information and collaboration.

Notable Quotes:

"The knowledge graph is a majority of the web scraped for entities, facts, and relationships between entities and then structured into a queryable format."
"Knowledge graphs are a great way to organize real-world data or data in the wild for a few reasons. First off, they have flexible schemas and they can adapt to different fact types. They also sort of mimic how humans think about the world in terms of entities, how they're connected."
"We recently employed some machine learning to get a bit more granularity so we have all these subcategories of companies that are beyond the standard industry category."
"This is not current employees. So what we're going to do here is we are going to make this a slightly more specific query."
"This is another machine learning computed field. Every organization of the 250 million plus in the KG has a similar to score to every other organization looking at a wide variety of metrics."
"This is document level sentiment, but we also have entity level sentiment, which is much more rare and a more powerful tool."
"Description can be used for a lot of things that it's it's not it's organizations are not neatly pegged into every single category you would think every single time."