If you have data you very likely have entity resolution challenges. Entity resolution is the task of figuring out who is who and what is what in your data. It encompasses finding duplicates, correlating, disambiguation and understanding the relationships between people, objects and key metadata.
Entity resolution is a key part of many workflows including use cases for marketing, sales, finance, IT, supply chain management, customer support, fraud detection, law enforcement, counterterror and cybersecurity, including the insider threat.
We recently put the new capabilities of Senzing through the paces and found its entity resolution capabilities to be absolutely amazing. It is easy to start, easy to connect to your data, easy to view and interact with results, and easy to drive decisions based on iterative examinations of the data. The Senzing app runs on your Windows or Mac (this is not a SaaS app, which keeps you in control of your data).
Here is a quick orientation and walk through of Senzing.
When you first launch Senzing you are presented with a easy to follow overview of how to get started:
Senzing is designed for entity resolution around people and organizations, so it is best over data that has names, addresses, phone numbers, identifiers and related attributes. You can try Senzing over data from your CRM tools (like exports from Salesforce, ACT, SugarCRM, Microsoft Dynamics), your address books (like Gmail, Microsoft Outlook), your marketing systems (like Mailchimp, Constant Contact), your web systems (like WordPress, WooCommerce, Stripe), your accounting and HR systems, or even just spreadsheets.
Once you have exported your data into .csv files all you need to do is click on the icon of the database in Senzing and then click where it says “add data source.”
For our test we already have some .csv files we are going to load.
Many data sources will require you to review your mapping of fields before loading just to make sure the columns in your data source are appropriately described. Clicking on the “review mapping” button will show you a screen where you review the mapping and then load the data.
After loading your data you can review your results in the dashboard visualization portion of Senzing. If you only have one data source loaded you will get insights like the number of duplicates found as well as entries that seem to be related in some way. If you have more than one data source loaded you can compare entities both within and across the data sources.
You can also run searches in Senzing and find all the records that related to a single entity. This is not to query for a group of people who live in the same city or have the same last name, it is to find entities that appear to be the related.
Here is a look at our current Senzing dashboard, which shows six data sources loaded.
Note we have files for a company that include total employees, terminated employees, employees that have had internal investigations, employees terminated with cause, and job applicants. We also loaded a file that contains information on 10 key vendors to the company.
Now lets do some quick analysis.
In the review section of the dashboard, after selecting the datasource for “vendors” and the datasource for “employees terminated with cause” we see something interesting.
There are no exact matches, and no possible matches, but two that are possibly related. Seems like we should take a close look.
Well now that is interesting. In two cases there are employees who were terminated who seem to have data in common with firms that are now in our supply chain. This is just test data but is a clear example of how Senzing could lead to an investigation that could mitigate some risks.
But these many data sets hold so many other examples. Here is a shot of analysis comparing job applicants to employees who have been terminated for cause.
Here too we see a correlation that would need to be considered, this time in support of the hiring process.
These were some very simple examples but we have shown how data from corporate systems can be rapidly ingested and analyzed and lead to actionable conclusions that in this scenario could help mitigate insider threats.
Overall: I love how easy Senzing is to work with. You can go from start to actionable insights driving decisions in no time at all.
Our recommendation: Since Senzing is free to download and free to use for your first 10,000 records, download it and give it a try on your own data. Developers can also access the core entity resolution engine via a real-time API to easily add this capability to Java, Python or C projects.
Learn more and download the app at: Senzing.com