Full text graph search
With Elasticsearch as the backend, EG can provide full text "graph" and "multilingual" search.
Feature: Takes ONE HIT for complex search queries with joins
Graph search means - searching over relationships. EG can do everything that Elasticsearch can do and it can do more. For ex. Show me companies of friends of friends of person X, who like India or have read books with description or title containing text "India". This query would take in a typical database or Elasticsearch setup, multiple queries to the database - 1 for friends of person X + 1 for books whose description or title contains text "India" + 1 for count wise breakup of city ids of people whose id is one of the friends of X's ids and whose likes contain either the id of India or the ids of the books whose description of title contains India + numCities queries for finding names of the cities that matched (numCities). This means (3 + numCities) queries for the above case - increasing the response time due to multiple round trips and also increasing developer's overhead of writing complex queries and in-memory joins while adding load to system resources like network, RAM, CPU. Using Elasticgraph this can be executed in a SINGLE HIT to elasticsearch, thanks to index time joins - a unique feature of Elasticgraph, very easy to set and use.
Feature: Allows Multi-lingual search
Multilingual search means the capability to store, index and search text fields in multiple languages. Suppose your database contains data in English, Hindi, Arabic, and Chinese. Now you can search and query independently over any language through the API, in a very neat way. This is also a unique feature offered by Elasticgraph, worldwide.
Domain: Search engines
This can be used to make search engines of different kinds, or to empower search within your own product, or to set up a structured knowledge base which presents knowledge in a neatly organized and searchable fashion.
Domain: Semantic knowledge graph
Semantic knowledge graphs that present knowledge in structured sets are a great way to organize and present information. They allow browing of content through relationships, along with a powerful search. Further, they can be used to extract semantic information from any content, and organize & search content by that. For ex. Extracting the person entity - Donald Trump when the text Donald Trump shows up in content.
Imagine you want to make a Data lake. With this, you can crawl content from any domains like research, technology, spirituality, etc. or social networks, or whole world wide web, and provide search and analytics on top of it. Of course, this technology is not as advanced as Google's, but it is GOOD ENOUGH for 99% of use cases across the industry. You can crawl, extract and organise HUGE AMOUNT of content (webpages, blog posts, social network posts, images, videos, audio...) with its metadata (date, author, keywords, description, sentiment, location, domain...), index the same and make it available for search and analytics at scale.