Blendee's Identity Graph: The Heart of the Identity Resolution Process
Posted: Wed Feb 12, 2025 8:59 am
As we have seen, the identity resolution process is based on the possibility of merging , but above all of resolving the different identifiers assigned to the user in the different interactions with various devices , into a single ID . The objective of an identity resolution process is therefore to create a complete and up-to-date 360-degree customer view , which is essential for customer experience personalization activities.
The crucial phase of an identity resolution process therefore concerns the construction of an Identity Graph , which could be considered as a large database, where the different identifiers are collected and linked to each other.
Coming back to the graph model, the different IDs will constitute vietnam mobile database the nodes of the graph and the relationships created will define the edges.
But what identifiers are we talking about?
Let's imagine a user who browses from multiple devices (PC, smartphones, smart TV) and uses more digital content: each time he connects, he is identified, after giving his consent, thanks to a code provided by the device or by the application environment.
In particular, the identity graph collects:
Device generated identifiers (MAID, TVID);
Identifiers generated by the application context (PPID, WEBID based on proprietary and third-party databases);
Strong identifiers (email, phone number) that are issued by the user during registration and authentication;
In the identity graph, the different identifiers collected are correlated as information about connections is also acquired, so we are able to define associations between the identifiers of people and devices , when the user logs in, the device association of the AdTech platform (ID Matching), the association between devices taking into account contextual attributes such as connection, geographic location that provide interesting data for profiling (let's imagine how the information about the shared wi-fi connection in a house between the PC, the smartphone and the smart TV, allows us to infer data about the interests, the age of the household members).
The identifiers thus organized within a graph allow me to identify both a single person and the devices associated with them (Device IDs associated with strong ID), and a group of people, allowing them to collect information or send personalized communications.
The crucial phase of an identity resolution process therefore concerns the construction of an Identity Graph , which could be considered as a large database, where the different identifiers are collected and linked to each other.
Coming back to the graph model, the different IDs will constitute vietnam mobile database the nodes of the graph and the relationships created will define the edges.
But what identifiers are we talking about?
Let's imagine a user who browses from multiple devices (PC, smartphones, smart TV) and uses more digital content: each time he connects, he is identified, after giving his consent, thanks to a code provided by the device or by the application environment.
In particular, the identity graph collects:
Device generated identifiers (MAID, TVID);
Identifiers generated by the application context (PPID, WEBID based on proprietary and third-party databases);
Strong identifiers (email, phone number) that are issued by the user during registration and authentication;
In the identity graph, the different identifiers collected are correlated as information about connections is also acquired, so we are able to define associations between the identifiers of people and devices , when the user logs in, the device association of the AdTech platform (ID Matching), the association between devices taking into account contextual attributes such as connection, geographic location that provide interesting data for profiling (let's imagine how the information about the shared wi-fi connection in a house between the PC, the smartphone and the smart TV, allows us to infer data about the interests, the age of the household members).
The identifiers thus organized within a graph allow me to identify both a single person and the devices associated with them (Device IDs associated with strong ID), and a group of people, allowing them to collect information or send personalized communications.