US 10,373,074 B2Granted

Adaptive correlation of user-specific compressed multidimensional data profiles to engagement rules

AssigneeIndiggo Associates LLC

InventorsVijay Perincherry, Janine Gelbart, Marc Inzelstein

Life of this patent

Estimating the term…

Loading patent…

Abstract

Some embodiments relate generally to the processing of compressed multidimensional data and selection of engagement rules based on the compressed multidimensional data. In some embodiments, a method includes retrieving, via a processor, a multidimensional data profile that includes a set of first inclination distributions, each associated with a data dimension. The processor matches a first set of engagement rules to the multidimensional data profile to define a matched set, each engagement rule of the first set of engagement rules having a corresponding confidence level and a corresponding set of second inclination distributions. The processor selects an engagement rule from the matched set that has a corresponding confidence level no less than a corresponding confidence level for each remaining engagement rule from the matched set, and sends a signal causing display of a stimulus to a user according to the selected engagement rule and not according to the remaining engagement rules.

Claims17

Dependent claims

A method, comprising:

retrieving, via a processor, a compressed multidimensional data profile including a plurality of first inclination distributions, each first inclination distribution from the plurality of first inclination distributions being associated with a data dimension of a plurality of data dimensions;
matching, via the processor, a first plurality of engagement rules to the compressed multidimensional data profile to define a matched plurality of engagement rules, each engagement rule of the first plurality of engagement rules having a corresponding confidence level and a corresponding plurality of second inclination distributions;
selecting, via the processor, an engagement rule from the matched plurality of engagement rules that has a corresponding confidence level no less than a corresponding confidence level for each remaining engagement rule from the matched plurality of engagement rules;
and
sending, via the processor, a signal causing display of a stimulus to a user according to the engagement rule of the matched plurality of engagement rules and not according to the remaining engagement rules,
the matching the first plurality of engagement rules to the compressed multidimensional data profile including one of:
ID
⁡
(
X
,
I
)
⁢
(
m
)
⁢
ID
⁡
(
X
,
R
)
=
(
Σ
⁡
(
min
⁡
(
lean
⁡
(
X
,
I
)
i
,
lean
⁡
(
X
,
R
)
i
)
)
,
⁢
i
=
1
,
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
⁢
(
Σ
⁡
(
lean
⁡
(
X
,
I
)
i
)
,
i
=
1
,
⁢
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
;
and
ID
⁡
(
X
,
R
)
⁢
(
m
)
⁢
ID
⁡
(
X
,
I
)
=
(
Σ
⁡
(
min
⁡
(
lean
⁡
(
X
,
I
)
i
,
lean
⁡
(
X
,
R
)
i
)
)
,
⁢
i
=
1
,
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
⁢
(
Σ
⁡
(
lean
⁡
(
X
,
R
)
i
)
,
i
=
1
,
⁢
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
,
where (m) is a matching operator, X is a data dimension of the plurality of data dimensions, ID(X,I) is an inclination distribution of the plurality of first inclination distributions, ID(X,R) is an inclination distribution of the plurality of second inclination distributions of an engagement rule R of the first plurality of engagement rules, inc(X, I) is a time increment of the inclination distribution of the plurality of first inclination distributions, inc(X, R) is a time increment of the inclination distribution of the plurality of second inclination distributions, lean(X, I) is a value of the inclination distribution of the plurality of first inclination distributions, and lean(X, R) is a value of the inclination distribution of the plurality of second inclination distributions.

243 more elements

An apparatus, comprising:

a user interface configured to display a plurality of stimuli to a user and to receive from the user a plurality of responses associated with the plurality of stimuli;
a memory operably coupled to the user interface and configured to store a compressed multidimensional data profile associated with the user;
and
a processor operably coupled to the memory and configured to:
retrieve the compressed multidimensional data profile from the memory;
match a set of engagement rules to the compressed multidimensional data profile to produce a matched set of engagement rules;
select an engagement rule of the matched set of engagement rules that has a corresponding confidence level no less than a corresponding confidence level for each remaining engagement rule from the matched set of engagement rules;
and
send a signal to cause the user interface to display a stimulus of the plurality of stimuli to the user via the user interface based on the engagement rule of the matched set of engagement rules,
the matching the set of engagement rules to the compressed multidimensional data profile being performed according to:
ID
⁡
(
X
,
I
)
⁢
(
m
)
⁢
ID
⁡
(
X
,
R
)
=
(
Σ
⁡
(
min
⁡
(
lean
⁡
(
X
,
I
)
i
,
lean
⁡
(
X
,
R
)
i
)
)
,
⁢
i
=
1
,
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
⁢
(
Σ
⁡
(
lean
⁡
(
X
,
I
)
i
)
,
i
=
1
,
⁢
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
,
where (m) is a matching operator, X is a data dimension of a plurality of data dimensions, ID(X,I) is an inclination distribution of a first plurality of inclination distributions, ID(X,R) is an inclination distribution of a second plurality of inclination distributions, inc(X, I) is a time increment of the inclination distribution of the first plurality of inclination distributions, inc(X, R) is a time increment of the inclination distribution of the second plurality of inclination distributions, lean(X, I) is a value of the inclination distribution of the first plurality of inclination distributions, and lean(X, R) is a value of the inclination distribution of the second plurality of inclination distributions.

127 more elements

An apparatus, comprising:

a user interface configured to display a plurality of stimuli to a user and to receive from the user a plurality of responses associated with the plurality of stimuli;
a memory operably coupled to the user interface and storing:
(1) a compressed multidimensional data profile associated with the user, the compressed multidimensional data profile including a first plurality of inclination distributions;
and (2) a set of engagement rules, each engagement rule of the set of engagement rules including an associated second plurality of inclination distributions;
and
a processor operably coupled to the memory and configured to:
retrieve the compressed multidimensional data profile from the memory;
match the compressed multidimensional data profile to the set of engagement rules to produce a matched set of engagement rules based on the first plurality of inclination distributions and the second pluralities of inclination distributions;
select an engagement rule of the matched set of engagement rules;
and
send a signal to cause the user interface to display a stimulus of the plurality of stimuli to the user via the user interface based on the selected engagement rule,
the matching the compressed multidimensional data profile to the set of engagement rules being performed according to:
ID
⁡
(
X
,
I
)
⁢
(
m
)
⁢
ID
⁡
(
X
,
R
)
=
(
Σ
⁡
(
min
⁡
(
lean
⁡
(
X
,
I
)
i
,
lean
⁡
(
X
,
R
)
i
)
)
,
⁢
i
=
1
,
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
⁢
(
Σ
⁡
(
lean
⁡
(
X
,
I
)
i
)
,
i
=
1
,
⁢
10
(
min
⁡
(
inc
⁡
(
X
,
I
)
,
inc
⁡
(
X
,
R
)
)
)
)
,
where (m) is a matching operator, X is a data dimension of a plurality of data dimensions, ID(X,I) is an inclination distribution of the first plurality of inclination distributions, ID(X,R) is an inclination distribution of the second plurality of inclination distributions, inc(X, I) is a time increment of the inclination distribution of the first plurality of inclination distributions, inc(X, R) is a time increment of the inclination distribution of the second plurality of inclination distributions, lean(X, I) is a value of the inclination distribution of the first plurality of inclination distributions, and lean(X, R) is a value of the inclination distribution of the second plurality of inclination distributions.

129 more elements

Dependencies are inferred from the claim text.

Facts

Application: 15/092,351Filed 6 Apr 2016
Published: US 2017/0293855 A112 Oct 2017
Granted: 6 Aug 2019
Term ends: Not estimatedThe source records don’t support an estimate.

Classification6 codes

G06F 16/907
Details of database functions independent of the retrieved data types: retrieval characterised by using metadata, e.g. metadata not derived from…
G06F 16/2264
Indexing structures: multidimensional index structures
G06N 5/022
Knowledge representation; Symbolic representation: knowledge engineering; Knowledge acquisition
G06N 20/00
Machine learning

Family

US 2017/0293855 A1Pre-grant publication · published 12 Oct 2017
US 10,373,074 B2This patent

From the US filing history. Foreign family members aren’t listed.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key