Description:AS FILED PDF DOCUMENTS , Claims:AS FILED PDF DOCUMENTS

We Claim:
1. A method of page management, comprising:
categorizing (302), using a first Machine Learning (ML) model (108), a set of pages into
at least one of a plurality of strands, based on a set of first parameters associated with each of the
5 set of pages;
classifying (304) each of the set of pages, using a second ML model (110), into one of a
plurality of sub-strands, based on an associated set of second parameters and the associated strand
from the plurality of strands, wherein a sub-strand from the plurality of sub-strands is linked with
a subset of the set of second parameters, and wherein each of the subset of second parameters has
10 a corresponding weight that is unique for that sub-strand;
determining (306), for each of the set of pages, a score, based on the associated sub-strand,
the weight assigned to each of the subset of second parameters of the associated sub-strand, and
values of each of the subset of second parameters of the associated sub-strand; and
performing (308) an action on at least one of the set of pages based on the determined
15 score.
2. The method as claimed in claim 1, further comprising pre-processing (802) each of the set of
pages, wherein the pre-processing of each page comprises:
extracting (804) a set of data vectors from at least one source associated with the page;
20 removing (806) a subset of data vectors that correspond to noise;
normalizing (808) the remaining set of data vectors to generate a normalized set of data
vectors;
processing (902), by the first ML model (108), the normalized set of data vectors based on
attributes associated with each of the plurality of strands; and
25 classifying (904), by the first ML model (108), the page under the strand in response to the
processing, wherein the categorization of the page into a strand of the plurality of strands is based
on the classification of the page.
3. The method as claimed in claim 1, further comprising:
30 identifying (702) a first trigger event; and
initiating (704) the categorization of the set of pages in response to identifying the first
trigger event
4. The method as claimed in claim 1, further comprising:
selecting a training dataset comprising a plurality of pages pre-categorized into the plurality
of strands;
5 training the first ML model (108) to categorize pages into the plurality of strands based on
the training dataset, wherein the first ML model (108) is trained prior to the categorization of the
set of pages;
categorizing, by the first ML model (108), each of the training dataset of pages into at least
one of the plurality of strands based on the training;
10 comparing the categorization of the training dataset of pages with corresponding precategorization;
determining a degree of accuracy of the first ML model (108) based on the comparing; and
performing reinforcement learning on the first ML model (108), based on the degree of
accuracy determined for the first ML model (108).
15
5. The method as claimed in claim 1, further comprising:
determining (402), by the second ML model (110), the number of plurality of sub-strands
to be created based on values of the set of second parameters associated with each of the set of
pages;
20 creating (404), by the second ML model (110), the plurality of sub-strands;
determining (1102), by the second ML model (110), a unique set of attributes for each of
the plurality of sub-strands; and
identifying (1104) introduction of a new page, wherein the new page is absent in the set of
pages;
25 determining (1106) the set of first parameters and the set of second parameters associated
with the new page;
categorizing (1108), by the first ML model (108), the new page into at least one of the
plurality of strands, based on the set of first parameters of the new page;
matching (1110), by the second ML model (110), the set of second parameters of the new
30 page with the unique set of attributes determined for each of the plurality of sub-strands; and classifying (1112) the new page into a sub-strand from the plurality of sub-strands, wherein
the cosine distance between the set of second parameters of the new page and the unique set of
attributes of the sub-strand is the least.
5 6. The method as claimed in claim 1, further comprising:
identifying (1002) a second trigger event associated with a page from the set of pages;
reclassifying (1004) the page into a sub-strand from the plurality of sub-strands in response
to the second trigger event; and
recomputing (1006) a score for the page, based on the associated sub-strand, weight
10 assigned to each of the subset of second parameters of the associated sub-strand, and values of
each of the subset of second parameters of the associated sub-strand; wherein the second trigger
event comprises change in value of at least one of the set of second parameters associated with the
page.
15 7. The method as claimed in claim 1, further comprising training the second ML model (110),
wherein the training comprises:
selecting a training dataset of pages pre-classified into a first predefined number of a first
set of sub-strands, wherein degree of variance amongst values of the set of second parameters for
the training dataset of pages is above a threshold;
20 processing, via the second ML mode (110), the training dataset of pages;
determining, by the second ML model (110), a second predefined number of a second set
of sub-strands to be created based on values of the set of second parameters associated with each
of the training dataset of pages; and
comparing each of:
25 the second predefined number with the first predefined number; and
a set of attributes for each of the first set of sub-strands with corresponding substrand from the second set of sub-strands;
determining a degree of accuracy of the second ML model (110) based on the comparison;
and
30 performing reinforcement learning based on the degree of accuracy determined for the
second ML model (110).

8. The method as claimed in claim 1, further comprising determining compliance of the value of
each of the subset of second parameters with a corresponding predefined threshold range defined
for the sub-strand for classification of each of the set of pages.
5 9. The method as claimed in claim 8, wherein performing the action on a page from the set of
pages comprises applying a deployment decision on the page based on the score determined for
the page, wherein the score is representative of a risk associated with the page.
10. A system (100) for page management, the system (100) comprising:
10 a processor (104); and
a memory (106) communicatively coupled to the processor (104), wherein the memory
(106) stores processor instructions, which when executed by the processor (104), cause the
processor (104) to:
categorize (302), using a first Machine Learning (ML) model (108), a set of pages
15 into at least one of a plurality of strands, based on a set of first parameters associated with
each of the set of pages;
classify (304) each of the set of pages, using a second ML model (110), into one of
a plurality of sub-strands, based on an associated set of second parameters and the
associated strand from the plurality of strands, wherein a sub-strand from the plurality of
20 sub-strands is linked with a subset of the set of second parameters, and wherein each of the
subset of second parameters has a corresponding weight that is unique for that sub-strand;
determine (306), for each of the set of pages, a score, based on the associated substrand, the weight assigned to each of the subset of second parameters of the associated
sub-strand, and values of each of the subset of second parameters of the associated sub25 strand; and
performing (308) an action on at least one of the set of pages based on the determined
score.