Knowledge Graph Dataset

Knowledge Graph

Knowledge Graph connects, organizes and normalizes consumer data into constructs that are superior for high-performance information retrieval. Consumers typically engage with brands and the internet using a wide array of mobile apps and networks. Given the scale and diversity of data, it would be difficult to personalize mobile app experiences and understand an audience if data were not organized into consumer profiles that are accessible very quickly.

Knowledge graphs represent knowledge in a form that is usable by machines and provide a foundation for Artificial Intelligence (AI). Cleansing, organizing and integrating data typically consumes 80% of the time for a data scientist to deliver an AI project. Knowledge Graph accelerates the delivery cycle by delivering connected data that can be accessed on leading graph platforms.

Types of Graph Models

Graphs contain "vertices" (nodes) and "edges" that connect vertices. There are two types of open graph models in the market and Phunware supports both.

  • RDF Graph - The Resource Description Framework (RDF) is a W3C standard for data exchange on the Web. It is open, standards based method for sharing and querying data. RDF graphs are stored as "triples" in a triple store (semantic graph database).
  • Property Graph - Property Graphs contain nodes (edges) and relations that connect nodes together. What is unique here, is that property graphs can store key-value pairs on the node and relation between the nodes.

Based on our ontology, Phunware can deliver connected data that can be hosted within leading graph database platforms.

Use Cases

High performance, connected data can enable a wide array of use cases.

  • Data science - influence future business outcomes by using Knowledge Graph as a high quality data foundation for machine learning to predict customer churn, optimize engagement and increase revenue through product recommendations.
  • Intelligent user experiences - personal, user-specific graphs of connected data enable experiences to be tailored to each user.
  • Intelligent business processes - use the high performance knowledge store to power business processes that leverage insights at the audience or user level.
  • Behavioral analysis - custom queries can quickly derive unique insight about a specific group of users using interactions across apps, networks and media.
  • Mobile audiences - custom queries can derive a list of mobile advertising ID's that represent a audience that you want to engage and deliver targeted media to.
  • Data monetization - a graph is a rich source of information that could be monetized for secondary purposes. Specific structured feeds could be exported from Knowledge Graph and used by other businesses to enrich their data infrastructure.
  • Data pre-processor for a DMP - since Knowledge Graph is a normalized, clean and enriched data set, the connected data can be used to derive unique segments from large volumes of user data. Marketers can then use a DMP to design audiences with Phunware and other data.
  • Structured web content and optimization - one strategy for optimizing search engine relevancy within Google and Bing is using structured HTML. These search engines understand how to consume and index markup. Google surfaces web content for a "thing" in their "knowledge graph panel". Phunware's data model leverages the vocabulary and therefore graph data could be easily exported to HTML and retain the markup for search index crawlers.