Method and system for active ranking of browser search engine results
A computer system and method a computer system and method for processing a search query result includes identifying a plurality of result pages in response to a search query submitted from a computing device directed to a collection of pages, determining a relevancy ranking of the result pages in accordance with a multiple dimension parameter set that includes metrics relating to the search query itself and also includes metrics unique to a subscriber associated with the search query, and providing the result pages in accordance with the determined relevancy ranking. This provides an active ranking process for the search results before they are provided to a user.
1. A computer method of processing a search query result, the computer method comprising:
identifying a plurality of result pages in response to a search query submitted from a computing device that is associated with a subscriber, wherein the search query is directed to a collection of pages;
determining a relevancy ranking for each of the result pages in accordance with a parameter set that includes metrics relating to the search query itself and also includes metrics unique to the subscriber associated with the computing device, and includes metrics related to the computing device from which the search query was submitted, wherein each of the parameter set metrics, when applied to the result pages, provides a re-ordering of the result pages, wherein the parameter set metrics are applied according to a tunable priority received from an administrative console input, and the determined relevancy ranking comprises a single merged ordering of the respective re-orderings; and
providing the result pages in accordance with the determined relevancy ranking;
wherein the administrative console input specifies parameter set processing for a plurality of ranking operations, and the administrative console input is used to select or deselect each of the ranking operations and specify an order of operation for the selected ranking operations, wherein determining a relevancy ranking includes:
associating each result page with a relevancy ranking value; and
ordering the plurality of result pages in accordance with the associated relevancy ranking values of the result pages; and
wherein the relevancy ranking value is adjusted with a tunable parameter value, and wherein the relevancy ranking value AR is calculated according to
∀
u
AR
k
(
u
)
=
c
[
α
∑
i
(
tf
i
*
log
(
D
df
)
)
+
β
∑
v
∈
⋃
u
P
(
v
)
+
λ
∑
v
∈
⋃
u
DSI
(
v
)
+
γ
(
1
-
(
u
x
-
P
(
V
x
)
)
2
+
(
u
y
-
P
(
V
y
)
)
2
)
]
+
(
1
-
c
)
where the terms are defined by the following:
AR=Active Rank matrix (sorted order of vectors)
u=set of total search results
v=set of metadata attributes associated with each search result item
k=a given Active Rank row (value for a specific search result)
c=normalization coefficient≦1
α=bias to weight keyword matching effects (0≦α≦1)
β=bias to weight Personalization profiles (0≦β≦1)
λ=bias to weight Device Specificity effects (0≦λ≦1)
γ=bias to weight LBS geometric distance effects (0≦γ≦1)
tf i =term frequency (keyword counts) or number of times a term i occurs in a search result page
df i =document frequency or number of pages in the search result pages containing term i
D=number of documents in the database
P=Personalization profile vector
DSI=Device Specific Index vector
(u x , u y )=each search result's geocoded location, if any
(v x , v y )=cellular user's actual physical location (stored in vector P) provided by the cellular network.
2. A computer method as defined claim 1 , wherein the search query metrics for determining the relevancy ranking include data relating to aggregate popularity of keywords contained in the search query.
3. A computer method as defined in claim 1 , wherein the metrics related to the computing device includes rendering capabilities of the computing device.
4. A computer method as defined in claim 3 , wherein the rendering capabilities of the computing device relate to display screen resolution.
5. A computer method as defined in claim 3 , wherein the rendering capabilities of the computing device relate to audio playback features.
6. A computer method as defined in claim 3 , wherein the rendering capabilities of the computing device relate to network bandwidth.
7. A computer method as defined in claim 1 , wherein the metrics related to the computing device includes geographic location of the computing device.
8. A computer method as defined in claim 1 , wherein determining a relevancy ranking includes deleting a result page if the determined relevancy ranking of the result page is below a deletion value.
9. A computer method as defined in claim 1 , wherein associating each result page is performed for each of the multiple parameter set metrics, and ordering the plurality of result pages comprises resolving respective relevancy rankings of each of the parameter set metrics.
10. A computer method as defined in claim 9 , wherein the multiple dimension parameter set includes metrics relating to the search query itself, to the subscriber associated with the search query, to capabilities of the computing device, and to geographic location of the computing device.
11. A computer method as in claim 1 , wherein bias values are received from the administrative console and include a bias value relating to the search query itself, a bias value relating to the subscriber metrics, a bias value relating to the computing device, and a bias value relating to location of the computing device upon sending the search query, such that the received bias values adjust the corresponding parameter set metrics.
12. A computer method as in claim 1 , wherein the parameter set metrics include at least two from among the set of query keywords, popularity, subscriber, personalization, device, device-specific, location, and location-specific metrics.
13. A computer system that processes a search query result, the computer system comprising:
a network interface for communicating with a network through which access to a collection of pages is obtained,
a processor that identifies a plurality of result pages in response to a search query submitted from a computing device that is associated with a subscriber, wherein the search query is directed to the collection of pages, determines a relevancy ranking for each of the result pages in accordance with a parameter set that includes metrics relating to the search query itself and also includes metrics unique to the subscriber associated with the computing device, and includes metrics related to the computing device from which the search query was submitted, wherein each of the parameter set metrics, when applied to the result pages, provides a re-ordering of the result pages, wherein the parameter set metrics are applied according to a tunable priority received from an administrative console input, and the determined relevancy ranking comprises a single merged ordering of the respective re-orderings and provides result pages in accordance with the determined relevancy ranking;
wherein the administrative console input specifies parameter set processing for a plurality of ranking operations, and the administrative console input is used to select or deselect each of the ranking operations and specify an order of operation for the selected ranking operations, wherein determining a relevancy ranking includes:
associating each result page with a relevancy ranking value; and
ordering the plurality of result pages in accordance with the associated relevancy ranking values of the result pages; and
wherein the relevancy ranking value is adjusted with a tunable parameter value, and wherein the relevancy ranking value AR is calculated according to
∀
u
AR
k
(
u
)
=
c
[
α
∑
i
(
tf
i
*
log
(
D
df
)
)
+
β
∑
v
∈
⋃
u
P
(
v
)
+
λ
∑
v
∈
⋃
u
DSI
(
v
)
+
γ
(
1
-
(
u
x
-
P
(
V
x
)
)
2
+
(
u
y
-
P
(
V
y
)
)
2
)
]
+
(
1
-
c
)
where the terms are defined by the following:
AR=Active Rank matrix (sorted order of vectors)
u=set of total search results
v=set of metadata attributes associated with each search result item
k=a given Active Rank row (value for a specific search result)
c=normalization coefficient ≦1
α=bias to weight keyword matching effects (0 ≦α≦1)
β=bias to weight Personalization profiles (0 ≦β≦1)
λ=bias to weight Device Specificity effects (0 ≦λ≦1)
γ=bias to weight LBS geometric distance effects (0 ≦γ≦1)
tf i =term frequency (keyword counts) or number of times a term i occurs in a search result page
df i =document frequency or number of pages in the search result pages containing term i
D=number of documents in the database
P=Personalization profile vector
DSI=Device Specific Index vector
(u x , u y )=each search result's geocoded location, if any
(v x , v y )=cellular user's actual physical location (stored in vector P) provided by the cellular network.
14. A computer system as defined claim 13 , wherein the search query metrics for determining the relevancy ranking include data relating to aggregate popularity of keywords contained in the search query.
15. A computer system as defined claim 13 , wherein the metrics related to the computing device includes rendering capabilities of the computing device.
16. A computer system as defined in claim 15 , wherein the rendering capabilities of the computing device relate to display screen resolution.
17. A computer system as defined in claim 15 , wherein the rendering capabilities of the computing device relate to audio playback features.
18. A computer system as defined in claim 15 , wherein the rendering capabilities of the computing device relate to network bandwidth.
19. A computer system as defined in claim 13 , wherein the metrics related to the computing device includes geographic location of the computing device.
20. A computer system as defined in claim 13 , wherein determining a relevancy ranking includes deleting a result page if the determined relevancy ranking of the result page is below a deletion value.
21. A computer system as defined in claims 13 , wherein associating each result page is performed for each of the multiple parameter set metrics, and ordering the plurality of result pages comprises resolving respective relevancy rankings of each of the parameter set metrics.
22. A computer system as defined in claim 13 , wherein the multiple dimension parameter set includes metrics relating to the search query itself, to the subscriber associated with the search query, to capabilities of the computing device, and to geographic location of the computing device.
23. A computer system as defined claim 13 , wherein bias values are received from the administrative console and include a bias value relating to the search query itself, a bias value relating to the subscriber metrics, a bias value relating to the computing device, and a bias value relating to location of the computing device upon sending the search query, such that the received bias values adjust the corresponding parameter set metrics.
24. A computer system as defined in claim 13 , wherein the parameter set metrics include at least two from among the set of query keywords, popularity, subscriber, personalization, device, device-specific, location, and location-specific metrics.
25. A computer program product for use in a computer system that executes program steps recorded in a non-transitory computer-readable media to perform a method for processing a search query result, the program product comprising:
a non-transitory recordable media:
a program of computer-readable instructions recorded in the media and executable by the computer system to perform operations comprising:
identifying a plurality of result pages in response to a search query submitted from a computing device that is associated with a subscriber, wherein the search query is directed to a collection of pages;
determining a relevancy ranking for each of the result pages in accordance with a parameter set that includes metrics relating to the search query itself and also includes metrics unique to the subscriber associated with the computing device, and includes metrics related to the computing device from which the search query was submitted, wherein each of the parameter set metrics, when applied to the result pages, provides a re-ordering of the result pages, wherein the parameter set metrics are applied according to a tunable priority received from an administrative console input, and the determined relevancy ranking comprises a single merged ordering of the respective re-orderings; and
providing the result pages in accordance with t determined relevancy ranking;
wherein the administrative console input specifies parameter set processing for a plurality of ranking operations, and the administrative console input is used to select or deselect each of the ranking operations and specify an order of operation for the selected ranking operations, wherein determining a relevancy ranking includes:
associating each result page with a relevancy ranking value; and
ordering the plurality of result pages in accordance with the associated relevancy ranking values of the result pa yes: and
wherein the relevancy ranking value is adjusted with a tunable parameter value,
wherein the relevancy ranking value AR is calculated according to
∀
u
AR
k
(
u
)
=
c
[
α
∑
i
(
tf
i
*
log
(
D
df
)
)
+
β
∑
v
∈
⋃
u
P
(
v
)
+
λ
∑
v
∈
⋃
u
DSI
(
v
)
+
γ
(
1
-
(
u
x
-
P
(
V
x
)
)
2
+
(
u
y
-
P
(
V
y
)
)
2
)
]
+
(
1
-
c
)
where the terms are defined by the following:
AR=Active Rank matrix (sorted order of vectors)
u=set of total search results
v=set of metadata attributes associated with each search result item
k=a given Active Rank row (value for a specific search result)
c=normalization coefficient ≦1
α=bias to weight keyword matching effects (0 ≦α≦1)
β=bias to weight Personalization profiles (0 ≦β≦1)
λ=bias to weight Device Specificity effects (0 ≦λ≦1)
γ=bias to weight LBS geometric distance effects (0 ≦γ≦1)
tf i =term frequency (keyword counts) or number of times a term i occurs in a search result page
df i =document frequency or number of pages in the search result pages containing term i
D=number of documents in the database
P=Personalization profile vector
DSI=Device Specific Index vector
(u x , u y )=each search result's geocoded location, if any
(v x , v y )=cellular user's actual physical location (stored in vector P) provided by the cellular network.
26. A program product as defined claim 25 , wherein the search query metrics for determining the relevancy ranking include data relating to aggregate popularity of keywords contained in the search query.
27. A program product as defined in claim 25 , wherein the metrics related to the computing device includes rendering capabilities of the computing device.
28. A program product as defined in claim 27 , wherein the rendering capabilities of the computing device relate to display screen resolution.
29. A program product as defined in claim 27 , wherein the rendering capabilities of the computing device relate to audio playback features.
30. A program product as defined in claim 27 , wherein the rendering capabilities of the computing device relate to network bandwidth.
31. A program product as defined in claim 25 , wherein the metrics related to the computing device includes geographic location of the computing device.
32. A program product as defined in claim 25 , wherein associating each result page is performed for each of the multiple parameter set metrics, and ordering the plurality of result pages comprises resolving respective relevancy rankings of each of the parameter set metrics.
33. A computer program product as defined in claim 32 , wherein the multiple dimension parameter set includes metrics relating to the search query itself, to the subscriber associated with the search query, to capabilities of the computing device, and to geographic location of the computing device.
34. A program product as defined in claim 25 , wherein bias values are received from the administrative console and include a bias value relating to the search query itself, a bias value relating to the subscriber metrics, a bias value relating to the computing device, and a bias value relating to location of the computing device upon sending the search query, such that the received bias values adjust the corresponding parameter set metrics.
35. A program product as defined in claim 25 , wherein the parameter set metrics include at least two dimensions selected from among the set of query keywords, popularity, subscriber, personalization, device, device-specific, location, and location-specific metrics.
36. A program product as defined in claim 25 , wherein determining a relevancy ranking includes deleting a result page if the determined relevancy ranking of the result page is below a deletion value.