Analytics On Entity Resolved Knowledge Graphs

Some things just seem meant to go together.

By their nature, graphs are already powerful and visually appealing analytic tools. But something special happens when you mix graphs with entity resolution…

Suddenly, the same graphed data becomes insightful and actionable.

Entity resolution is the missing ingredient that brings true utility to knowledge graphs. It’s one of the very first things organizations should invest in to maximize their knowledge graphs – creating what is known as an Entity Resolved Knowledge Graph.

At the Senzing 2023 Global User Conference we heard over and over how much benefit our end users are getting from their entity resolved knowledge graphs.

If you weren’t able to attend, you might want to check out this keynote by an independent technical consultant (who has delivered graph based analytic solutions for a variety of organizations):

Entity resolved knowledge graphs will turbo-charge the world’s graph based machine learning, graph analytics and human-facing graph visualizations.

And, as the speaker above makes clear, “You need an entity resolution process that occurs at the data layer as you ingest the data.”

Senzing, as a purpose built real time AI for entity resolution, allows organizations to quickly and easily integrate entity resolution into their knowledge graph projects.  You can test it out for yourself, by following this quick 30-minute tutorial written by Paco Nathan and posted on Neo4j: Entity Resolved Knowledge Graphs: A Tutorial.

“This is one such sample architecture I’ve deployed before” ~ Mel Richey

When it comes to entity resolution – Nothing is more accurate than Senzing.
Nothing is easier. Nothing is more cost-effective.

Nothing.

But don’t take our word for it – you can get started with Senzing in minutes.

If anything seems hard – then something is wrong – so shoot us an email here support@senzing.com and we’ll help you, stat.

01:58 – “Regardless of which synonym you use and regardless of your discipline and domain, I contend that everyone here is working on the same basic challenge, which is how do we distill understanding and prediction from the vapors of the human experience and the data sets that our sensors collect within it.”

02:30 – “We’re all trying to do this aspect of reasoning and sensemaking, and that’s where our focus is.”

03:31 – “And I by no means mean to suggest that entity resolved graphs are, you know, the answer to this. But I do think they fill a gap. Entity resolved graphs fill a gap in the knowledge representation domain because fundamentally the power of a graph […] in your organizations is synthesizing data and making that synthesis available to analysts and to algorithms. That’s really where you get the bang for the buck in graph space.”

10:50 – “[…] you need an entity resolution process that occurs at the data layer as you ingest the data. So, entity resolution is your first step. Many people think it’s like the last step – bring all the data in and then entity resolve it. It’s actually when you’re building a graph-based capability, any entity resolution is your first step because it’s your key to deriving relationships and connections within your data sets.”

13:22 – “When we talk about entity resolution and graphs, this is pretty important. I do want to make the distinction between graph-based entity resolution and entity resolution that occurs at the data layer, because many people, many of my customers come and say, ‘Well, you know, I heard so-and-so does their own entity resolution,’ or ‘We’re going to do graph-based entity resolution because it’s better.’ […] But it does not scale. It’s very hard. It’s something you do after the fact. So, the answer for graph-based entity resolution is yes, and. You have to do entity resolution at the data layer first, as you bring all of the data into the graph environment, because that is the process that scales and that is the process that connects your data in a way that an entity-resolve graph is then rendered. Then after the fact, if you’d like to have a couple of data scientists go at the graph-based entity resolution problem, it will probably add to the quality of your entity resolution.”

16:30 – “[…] if you are looking for an entity resolution solution that is a scalable developer library, that gives you the explainability behind the resolutions, that gives you the ability to tweak your thresholds for what you want to consider a resolution or a connection. There is no other option. The other solutions for entity resolution are either baked into large platforms that you have to buy and adopt that entire platform within which entity resolution is a service, usually a more lightweight service than the type of sophisticated and robust entity resolution Senzing provides.”

17:27 – “When it comes to data storage in the national security community, we use a lot of Neo4j.”

17:47 – “And then we get to the visualization and analysis piece. And this is where you know, JavaScript isn’t going to cut it. If you actually want a tool and you need a tool that gives you that robust front-end ability to iterate through your graph as an analyst, click through query the tool you’re looking for something like one of the platforms up here. ” […]

“I’m a big fan of a platform called Hume. It’s made by a company called GraphAware who may actually be wandering around here today. […] it is designed for investigative workloads and case management. So, it’s a very analytically-minded tool and it’s something that I’ve been using in my deployments for the past couple of years.”