Systems and methods for determining real-time visitor segments

Patent metadata

Field	Value
Publication number	US-2017193382-A1
Application number	US-201614986811-A
Country	US
Kind code	A1
Filing date	Jan 4, 2016
Priority date	Jan 4, 2016
Publication date	Jul 6, 2017
Grant date	—

Abstract

Systems and methods disclosed herein compactly store representations of segment-specific interaction data from a real time data stream of data interactions by multiple entities to facilitate segment-specific analytics for particular time periods. Segment rules defining characteristics of entities within a segment are received. A first probabilistic data structure is created representing unique entity IDs included in instances of interaction data in the real time data stream during a first time period. A second probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a second time period different from the first time period. The first probabilistic data structure represents only entity IDs of entities within the segment and the second probabilistic data structure represents only entity IDs of entities within the segment. The first and second probabilistic data structures are indexed and stored.

First claim

1 . In an environment in which a real time data stream of data interactions by multiple entities is tracked, a method for compactly storing representations of segment-specific interaction data to facilitate segment-specific analytics for particular time periods, the method comprising:
receiving segment rules defining characteristics of entities within a segment; creating a first probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a first time period, the first probabilistic data structure representing only entity IDs of entities within the segment; creating a second probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a second time period different from the first time period, the second probabilistic data structure representing only entity IDs of entities within the segment; and indexing and storing the first probabilistic data structure and the second probabilistic data structure, wherein the indexing allows unique entity IDs in the first probabilistic data structure and the second probabilistic data structure to be counted and compared to facilitate segment-specific analytics for particular time periods.

2 . The method of claim 1 , further comprising:
determining a first number of unique entity IDs from the first probabilistic data structure and determining a second number of unique entity IDs from the second probabilistic data structure.

3 . The method of claim 1 , further comprising:
receiving segment rules defining characteristics of entities within a second segment different than the segment; creating a third probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a first time period, the first probabilistic data structure representing only entity IDs of entities within the second segment; creating a fourth probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a second time period different from the first time period, the second probabilistic data structure representing only entity IDs of entities within the second segment; and indexing and storing the third probabilistic data structure and the fourth probabilistic data structure, wherein the indexing allows unique entity IDs in the third probabilistic data structure and the fourth probabilistic data structure to be counted and compared to facilitate segment-specific analytics for particular time periods.

4 . The method of claim 1 , wherein the entity IDs are user IDs.

5 . The method of claim 1 , wherein the entity IDs are device IDs.

6 . The method of claim 1 , wherein the first probabilistic data structure and the second probabilistic data structure are created using a probabilistic algorithm.

7 . The method of claim 1 , further comprising creating, indexing and storing multiple probabilistic data structures associated with multiple time periods.

8 . The method of claim 6 , wherein the probabilistic algorithm is at least one of a HyperLogLog, Bloom filter and a Count-Min sketch

9 . The method of claim 1 , further comprising generating a report for display providing a first number of entity IDs of entities within the segment during the first time period and a second number of entity IDs of entities within the segment during the second time period, wherein the report is generated using an HTTP streaming API.

10 . In an environment in which a computer receives a real-time data stream, a method for determining a first segment count and a second segment count from the real-time data stream, the method comprising:
receiving first segment rules defining characteristics of entities within a first segment; receiving second segment rules defining characteristics of entities within a second segment; creating a first probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a first time period, the first probabilistic data structure representing only entity IDs of entities within the first segment; creating a second probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during the first time period, the second probabilistic data structure representing only entity IDs of entities within the second segment, wherein the first probabilistic data structure and the second probabilistic data structure are created in parallel; indexing and storing the first probabilistic data structure and the second probabilistic data structure, wherein the indexing allows unique entity IDs in the first probabilistic data structure and the second probabilistic data structure to be counted and compared to facilitate segment-specific analytics for particular time periods.

11 . The method according to claim 10 , further comprising:
creating multiple probabilistic data structures representing entity IDs for the first segment and the second segment for multiple time periods; and indexing and storing the multiple probabilistic data structures.

12 . The method of claim 10 , wherein the entity IDs are user IDs.

13 . The method of claim 10 , wherein the entity IDs are device IDs.

14 . The method of claim 10 , wherein the first probabilistic data structure and the second probabilistic data structure are created using a probabilistic algorithm.

15 . The method of claim 14 , wherein the probabilistic algorithm is at least one of a HyperLogLog, Bloom filter and a Count-Min sketch

16 . The method of claim 10 , further comprising generating a report for display providing a first number of entity IDs of entities within the first segment during the first time period and a second number of entity IDs of entities within the second segment during the first time period, wherein the report is generated using an HTTP streaming API.

17 . In an environment in which a real time data stream of data interactions by multiple entities is tracked, a system for compactly storing representations of segment-specific interaction data to facilitate segment-specific analytics for particular time periods, the system comprising:
a processing device; a memory device communicatively coupled to the processing device, wherein the processing device is configured to execute instructions including in the memory device configured to perform operation comprising:
receiving segment rules defining characteristics of entities within a segment;
creating a first probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a first time period, the first probabilistic data structure representing only entity IDs of entities within the segment;
creating a second probabilistic data structure representing unique entity IDs included in instances of interaction data in the real time data stream during a second time period different from the first time period, the second probabilistic data structure representing only entity IDs of entities within the segment; and
indexing and storing the first probabilistic data structure and the second probabilistic data structure, wherein the indexing allows unique entity IDs in the first probabilistic data structure and the second probabilistic data structure to be counted and compared to facilitate segment-specific analytics for particular time periods.

18 . T

Assignees

• Adobe Systems Inc

Inventors

• Villa Natalee
• Pead Scott
• Nyland James
• Mathis Craig

Classifications

• G06N7/01 Primary

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F16/2477 Primary

  Temporal data queries · CPC title

• G06N7/005 Primary

  Physics · mapped topic

• G06F17/30289

  Physics · mapped topic

• G06F17/30321

  Physics · mapped topic