NZ785400A - Method and system for determining collaboration between employees using artificial intelligence (ai) - Google Patents
Method and system for determining collaboration between employees using artificial intelligence (ai)Info
- Publication number
- NZ785400A NZ785400A NZ785400A NZ78540022A NZ785400A NZ 785400 A NZ785400 A NZ 785400A NZ 785400 A NZ785400 A NZ 785400A NZ 78540022 A NZ78540022 A NZ 78540022A NZ 785400 A NZ785400 A NZ 785400A
- Authority
- NZ
- New Zealand
- Prior art keywords
- employees
- collaboration
- employee
- nodes
- embeddings
- Prior art date
Links
- 238000010801 machine learning Methods 0.000 claims abstract 13
- 230000000875 corresponding Effects 0.000 claims 4
- 230000004044 response Effects 0.000 claims 4
- 229940035295 Ting Drugs 0.000 claims 2
- 230000002776 aggregation Effects 0.000 claims 2
- 238000004220 aggregation Methods 0.000 claims 2
- 230000000977 initiatory Effects 0.000 claims 2
- 230000002787 reinforcement Effects 0.000 claims 2
Abstract
method and system for determining collaboration between employees is disclosed. In some embodiments, the method includes receiving a plurality of collaboration parameters associated with a set of employees. The method further includes creating a plurality of employee nodes associated with the set of employees in a hierarchical tree, based on the plurality of collaboration parameters and a first pre-trained machine learning model. The method further includes generating a plurality of vector embeddings associated with the plurality of employee nodes, based on the first pre-trained machine learning model. The method further includes determining a degree of collaboration between at least two employees from the set of employees based on one or more vector embeddings from the generated plurality of embeddings. of employees in a hierarchical tree, based on the plurality of collaboration parameters and a first pre-trained machine learning model. The method further includes generating a plurality of vector embeddings associated with the plurality of employee nodes, based on the first pre-trained machine learning model. The method further includes determining a degree of collaboration between at least two employees from the set of employees based on one or more vector embeddings from the generated plurality of embeddings.
Description
PATENT ATION
METHOD AND SYSTEM FOR DETERMINING COLLABORATION BETWEEN
EMPLOYEES USING ARTIFICIAL INTELLIGENCE (AI)
NAVIN SABHARWAL
AMIT AGRAWAL
DESCRIPTION
cal Field
Generally, the disclosure relates to Artificial igence (AI). More specifically,
the disclosure relates to a method and system for determining collaboration between
employees using AI.
Background
Generally, every organization may be an integration of numerous departments
or teams collaborating together to ensure that organization goals are met smoothly. With
collaboration, employees of same teams or different teams may work at their full potential.
Therefore, ment and efficient use of resources may be crucial for smooth functioning
of every organization. However, tracking of different aspects related to the employees for
the oration may be a tedious task, especially when the employees have concurrent
tasks or when the employees work on multiple projects simultaneously. As a result,
managing the resources for collaboration amongst employees may be time consuming and
an inefficient process, especially for large organizations. Inefficient ation of the
resources at hand may lead to compromise in profits of the zation. In n
scenarios, managing the resources manually may lead to the risk of mistakes.
Accordingly, there is a need for a robust, streamlined and efficient method and
system to determine the collaboration between the employees of the organization.
SUMMARY OF INVENTION
In one embodiment, a method of determining collaboration between
employees is disclosed. The method may include receiving a plurality of collaboration
parameters associated with a set of employees. The method may further include creating a
ity of employee nodes ated with the set of employees in a chical tree,
based on the plurality of collaboration ters and a first pre-trained machine learning
model. It should be noted that, the hierarchical tree further comprises a plurality of edges
and each of the plurality of edges interconnects at least two of the set of employee nodes.
The method may further include generating a plurality of vector embeddings associated with
the plurality of employee nodes, based on the first pre-trained machine learning model. The
method may further include determining a degree of collaboration between at least two
employees from the set of ees based on one or more vector embeddings from the
generated plurality of embeddings. It should be noted that, the degree of collaboration
corresponds to association between the at least two ees.
In r embodiment, a system for determining collaboration n
employees is disclosed. The system includes a processor and a memory communicatively
coupled to the processor. The memory may store processor-executable instructions, which,
on execution, may causes the processor to receive a plurality of collaboration parameters
associated with a set of employees. The processor-executable ctions, on execution,
may further cause the processor to create a plurality of employee nodes associated with the
set of employees in a hierarchical tree, based on the plurality of collaboration parameters
and a first pre-trained machine learning model. It should be noted that, the hierarchical tree
further comprises a plurality of edges and each of the plurality of edges interconnects at
least two of the set of employee nodes. The processor-executable ctions, on execution,
may further cause the processor to generate a ity of vector embeddings associated
with the ity of ee nodes, based on the first pre-trained machine learning model.
The processor-executable instructions, on execution, may further cause the processor to
determine a degree of collaboration between at least two employees from the set of
employees based on one or more vector embeddings from the generated ity of
embeddings. It should be noted that, the degree of collaboration corresponds to association
between the at least two employees.
In yet another embodiment, a ansitory computer-readable medium
storing computer-executable instruction for determining collaboration between employees is
disclosed. The stored instructions, when executed by a processor, may cause the processor
to perform operations including receiving a plurality of collaboration parameters associated
with a set of employees. The operations may further include ng a ity of employee
nodes associated with the set of employees in a hierarchical tree, based on the plurality of
collaboration parameters and a first pre-trained machine learning model. It should be noted
that, the hierarchical tree further comprises a plurality of edges and each of the plurality of
edges interconnects at least two of the set of employee nodes. The operations may further
include generating a ity of vector embeddings ated with the plurality of employee
nodes, based on the first ained machine learning model. The operations may further
include determining a degree of collaboration n at least two employees from the set
of employees based on one or more vector embeddings from the generated plurality of
embeddings. It should be noted that, the degree of collaboration corresponds to association
between the at least two employees.
It is to be tood that both the foregoing general ption and the
following detailed description are exemplary and explanatory only and are not restrictive of
the invention, as claimed.
BRIEF DESCRIPTION OF THE DRAWINGS
The present disclosure can be best understood by reference to the following
description taken in conjunction with the anying drawing figures, in which like parts
may be referred to by like numerals
illustrates a functional block diagram of an Artificial Intelligence (AI)
based collaboration system for determining collaboration between employees, in
accordance with an embodiment.
illustrates a functional block diagram of various modules within a
memory of an AI based collaboration system for determining collaboration between
employees, in accordance with an ment.
illustrates a flowchart of a method for determining collaboration between
employees, in ance with an embodiment.
rates a flowchart of a method for generating a plurality of vector
embeddings, in accordance with an embodiment.
rates a flowchart of a method for determining a degree of
collaboration between at least two ees, in accordance with an embodiment.
illustrates a flowchart of a method evaluating performance of each of
the set of employees, in accordance with an embodiment.
FIGS. 7A – 7B depicts a pictorial representation of a hierarchical tree and an
employee relationship graph for each of a set of employees, in accordance with an
exemplary embodiment.
s a plurality of vector embedding generated for a set of
employees, in accordance with an exemplary embodiment.
depicts a scenario of collaborating a new employee with at least one of
the set of employees, in accordance with an exemplary embodiment.
FIGS. 10A – 10C illustrates a tabular representation for input data
corresponding to collaboration parameters ated with employees, in accordance with
an ary embodiment.
illustrates an AI based collaboration system trained on a reinforcement
learning approach, in accordance with an ary embodiment.
illustrates a collaboration system that uses inverse reinforcement
learning to perform hyperparameter tuning, in accordance with an exemplary embodiment.
illustrates a transfer learning approach to create a new environment
for an AI based collaboration system, in accordance with an exemplary embodiment.
DETAILED DESCRIPTION OF THE DRAWINGS
The following description is presented to enable a person of ordinary skill in
the art to make and use the disclosure and is provided in the context of ular applications
and their requirements. Various modifications to the embodiments will be readily nt
to those skilled in the art, and the generic principles defined herein may be applied to other
embodiments and applications t departing from the spirit and scope of the disclosure.
Moreover, in the following description, numerous details are set forth for the purpose of
explanation. r, one of ordinary skill in the art will realize that the disclosure might be
ced without the use of these ic details. In other instances, well-known structures
and devices are shown in block diagram form in order not to obscure the description of the
sure with unnecessary detail. Thus, the disclosure is not intended to be d to the
embodiments shown, but is to be accorded the widest scope consistent with the principles
and features disclosed herein.
While the disclosure is described in terms of particular examples and
rative figures, those of ordinary skill in the art will recognize that the sure is not
limited to the examples or figures described. Those skilled in the art will recognize that the
operations of the various ments may be implemented using hardware, software,
re, or ations thereof, as appropriate. For example, some processes can be
d out using processors or other digital circuitry under the control of software, firmware,
or hard-wired logic. (The term “logic” herein refers to fixed hardware, programmable logic
and/or an appropriate combination thereof, as would be ized by one skilled in the art
to carry out the recited functions). Software and firmware can be stored on computerreadable
storage media. Some other processes can be ented using analog circuitry,
as is well known to one of ordinary skill in the art. Additionally, memory or other storage, as
well as communication components, may be employed in embodiments of the invention.
The present disclosure tackles limitations of existing systems to facilitate
determination of collaboration between employees working in a same organization. As will
be appreciated, the employees may be working in a same team or different teams of an
organization. In order to facilitate determination of collaborations between employees, the
present disclosure introduces an AI based collaboration . In order to determine the
collaboration between employees, the t disclosure may determine a degree of
collaboration between at least two employees from a set of employees. The degree of
collaboration may be determined based on one or more vector embeddings from a
generated ity of embeddings. In an embodiment, the degree of oration may
correspond to association between the at least two employees. In addition, the plurality of
vector embeddings may include extracting one or more set of edges initiating from a root
node of a plurality of the employee nodes and ating at an ated leaf node of the
plurality of the employee nodes in a hierarchical tree.
Moreover, the present disclosure may determine the degree of collaboration
between at least two employees based on a plurality of oration ters ated
with each of the set of employees. The plurality of collaboration parameter may include, but
is not limited to, at least one or more of employee skillset, employee role, employee rating,
collaboration complexity, and collaboration satisfaction. Further, the present disclosure may
facilitate computation of score for each of the set of employees in order to facilitate ranking
of each of the set of employees working in the organization.
In an ment, the present disclosure may train the AI based collaboration
system by exposing to a new environment during initial training process. The AI based
collaboration system may utilize an active learning algorithm. Based on the active learning
algorithm, the AI based system may determine collaboration satisfaction among employees
in order to determine a degree of collaboration n at least two employees from the set
of employees. For this, the AI based collaboration system may identify one or more first
employees and one or more second employees from the set of employees. In an
embodiment, the one or more first employees may correspond to employees may
employees providing ance to one or more of the set of ees. In addition, the one
or more second employees may correspond to employees ing assistance from one or
more of the set of employees. Further, based on identifications of the one or more first
employees and the one or more second employees a feedback may be generated.
Thereafter, based on the generated feedback the AI based collaboration system may
evaluate performance of each of the set of employees. This has been explained in detail in
conjunction to to .
Referring now to a functional block diagram for a network 100 of an AI
based collaboration system for determining collaboration between employees is illustrated,
in accordance with an embodiment. With reference to there is shown an AI based
collaboration system 102 that includes a memory 104, a processor(s) 106, I/O devices 108
and a machine learning (ML) model 112. The I/O devices 108 of the AI based collaboration
system 102 may further include an I/O interface 110. Further, in the network environment
100, there is shown a server 114, a database 116, external devices 118 and a
communication network 120 (hereinafter referred as k 120).
The AI based collaboration system 102 may be communicatively coupled to
the server 114, and the external devices 118, via the network 120. Further, the AI based
collaboration system 102 may be communicatively coupled to the database 116 of the server
114, via the network 120. A user or an strator (not shown in the may interact
with the AI based collaboration system 102 via the user interface 110 of the I/O device 108.
The AI based collaboration system 102 may include suitable logic, circuitry,
interfaces, and/or code that may be configured to ine collaboration between
employees of an organization, based on a plurality of collaboration parameters associated
with the employees. Such employees may be from a same team or a different team in the
organization and working at different levels of a hierarchy in the organization. The plurality
of collaboration parameter may include, but is not limited to, at least one or more of employee
skillset, employee role, employee rating, oration complexity, and collaboration
satisfaction. The AI based collaboration system 102 may correspond to a tree based
hierarchical collaboration system.
AI based collaboration framework associated with the AI based oration
system 102 may be implemented on but are not limited to, a server, a desktop, a laptop, a
notebook, a tablet, a smartphone, a mobile phone, an application server, or the like. By way
of an example, the AI based collaboration system 102 may be implemented as a plurality of
distributed cloud-based resources by use of several technologies that are well known to
those skilled in the art. Other examples of implementation of the AI based collaboration
system 102 may include, but are not limited to, a oud server and a media server.
The I/O s 108 may be configured to provide inputs to the AI based
collaboration system 102 and render output on user equipment. In an embodiment, the user
equipment, may pond to the external devices 118. By way of an example, the user
may provide , i.e., the plurality of collaboration parameters via the I/O s 108 by
using the user interface 110. In on, the I/O devices 108 may be configured to render
information associated with ranks of the employees computed for each of the set of
employees by the AI based collaboration system 102.
Further, the I/O device 108 may be configured to y results (i.e., a degree
of collaboration between at least two employees from the set of employees) generated by
the AI based collaboration system 102, to the user. By way of another example, the user
interface 110 may be configured by the user to provide inputs to the AI based collaboration
system 102. Thus, for example, in some embodiment, the AI based collaboration system
102 may ingest the plurality of collaboration parameters via the user interface 110. Further,
for example, in some ments, the AI based collaboration system 102 may render
ediate results (e.g., a score computed for each of the set of employees, and a
ck generated for each of the set of employees) or final s (e.g., the degree of
collaboration between at least two employees, and results of evaluation performed for each
of the set of employees) to the user via the user ace 110.
The memory 104 may store ctions that, when executed by the processor
106, may cause the processor 106 to determine collaboration between employees. The
sor 106 may determine the collaboration between each of the set of employees based
on the plurality of collaboration parameters associated with each of the set of employees, in
accordance with some ments. As will be described in greater detail in conjunction
with to , in order to determine collaboration between each of the set of
employees, the processor 106 in conjunction with the memory 104 may perform various
functions including on of a plurality of employee nodes associated with the set of
employees, generation of a plurality of vector embeddings associated with the plurality of
employee nodes, and computation of score associated with each of the set of employees,
and identification of one or more first employees and one or more second employees for
each of the set of employees.
The memory 104 also store various data (e.g., a plurality of collaboration
parameters, the plurality of vector embeddings, the degree of collaboration, the computed
score, and ranks associated with each of the set of employees) that may be captured,
sed, and/or required by the AI based collaboration system 102. The memory 104 may
be a non-volatile memory (e.g., flash memory, Read Only Memory (ROM), Programmable
ROM , Erasable PROM (EPROM), ically EPROM (EEPROM) , etc.)
or a volatile memory (e.g., Dynamic Random-Access Memory (DRAM), Static Random-
Access memory (SRAM), etc.).
In accordance with an embodiment, the AI based collaboration system 102
may be configured to deploy the ML model 112 to use output of the ML model 112 to
generate real or near-real time inferences, take decisions, or output prediction results. The
ML model 112 may be deployed on the AI based collaboration system 102, once the ML
model 112 is trained on the AI based collaboration system 102 for determining the degree
of oration between at least two employees from the set of employees.
In accordance with one embodiment, the ML model 112 may correspond to a
first pre-trained machine learning model. In accordance with an embodiment, the first pretrained
machine learning model may correspond to a graph neural network model that may
be used by the AI based collaboration system 102 to determine the degree of collaboration
between at least two employees from the set of employees. Examples of the graph neural
network model includes, but not limited to, Long Short-Term Memory (LSTM), LSTM – GRU
(Long Short-Term Memory – Gated Recurrent Units) of Neural Network.
The ML model 112 may be configured to create the plurality of employee
nodes. The ML model 112 may create the plurality of employee nodes in order to assist the
AI based collaboration system 102 to generate a plurality of vector embeddings. In
accordance with another embodiment, the ML model 112 may correspond to a second
machine ng model (such as, a Rank-Net model). The ML model 112 may be trained to
determine oration satisfaction between each of the set of employees. In an
embodiment, the collaboration satisfaction may correspond to one of a successful
collaboration and an unsuccessful oration between at least two employees from the
set of employees.
Further, the AI based collaboration system 102 may interact with the server
114 or the external device 118 over the k 120 for sending and receiving various types
of data. The external device 118 may include, but not be limited to a desktop, a laptop, a
notebook, a netbook, a tablet, a smartphone, a remote server, a mobile phone, or another
computing system/device.
The network 120, for example, may be any wired or wireless communication
k and the examples may include, but may be not limited to, the Internet, Wireless
Local Area Network (WLAN), Wi-Fi, Long Term Evolution (LTE), Worldwide Interoperability
for Microwave Access (WiMAX), and l Packet Radio Service (GPRS).
In some embodiments, the AI based collaboration system 102 may fetch
information associated with each of the set of employees from the server 114, via the
network 120. The database 116 may store information associated with existing logies
or the new logy in demand.
In ion, the AI based collaboration system 102 may be configured to
e the plurality of collaboration ters ated with the set of employees. The
AI based collaboration system 102 may be r configured to create the plurality of
employee nodes associated with the set of employees in the hierarchical tree. In an
embodiment, the hierarchical tree may include a plurality of edges. In addition, each of the
ity of edges may interconnect at least two of the set of employee nodes. Further, the
AI based collaboration system 102 may generate the plurality of vector embeddings
associated with the plurality of employee nodes. The AI based collaboration system 102
may then determine the degree of collaboration between at least two employees from the
set of employees. In order to determine the degree of collaboration, the AI based
collaboration system 102 may e the score for each of the set of employees based on
the identified collaboration satisfaction. Thereafter, the AI based collaboration system 102
may rank each of the set of employees based on the associated computed score. In addition,
the AI based collaboration system 102 may generate a feedback for each of the set of
employees. Based on the ted feedback, the AI based collaboration system 102 may
evaluate performance of each of the set of employees. This is further explained in detail in
conjunction with to .
Referring now to a functional block diagram of various modules within
a memory of an AI based collaboration system for determining oration between
employees is illustrated, in ance with an embodiment of the present sure.
is explained in conjunction with
With reference to there is shown input data 202, a database 204
coupled with the memory 104. The memory 104 may include a training module 206, a
reception module 208, a creation module 210, a generation module 212, a determination
module 214, and an evaluation module 220. The ination module 214 may further
include a computing module 216, and a ranking module 216. The modules 206-220 may
include routines, programs, objects, ents, data structures, etc., which perform
particular tasks or implement particular abstract data types. The modules 206-220 described
herein may be implemented as software modules that may be executed in a cloud-based
computing environment of the AI based collaboration system 102.
In accordance with an embodiment, the memory 104 may be configured to
receive the input data 202. The input data 202 may correspond to data ated with a
plurality of collaboration parameters associated with a set of ees. In an embodiment,
the plurality of collaboration parameters may include, but is not limited to, at least one or
more of employee skillset, employee role, employee rating, collaboration complexity, and
collaboration satisfaction. The memory 104 may be ured to receive the input data 202
in the database 204 from the external device 118. Additionally, the input data 202 may
include information associated with the set of employees.
The database 204 may serve as a repository for storing data processed,
received, and generated by the modules 206-220. The data generated as a result of the
execution of the modules 0 may be stored in the database 204
In conjunction to the training module 204 may be configured to train
the ML model 112. It should be noted that, the ML model 112 may correspond to each of a
first pre-trained e learning model, a second ained ML model, and a third pretrained
ML model. Based on training, the training module 204 may assist the creation
module 210 in creation of a plurality of employee nodes associated with each of the set of
employees. Once the plurality of employee nodes is created, the training module 204 may
assist the generation module 212 to generate the ity of vector embedding associated
with the plurality of employee nodes.
In an embodiment, the plurality of employee nodes and the plurality of vector
embedding may be generated based on the first pre-trained machine learning model. In
addition, the training module 204 may assist the generation module 212 to generate one or
more second vector embeddings of the plurality of vector embeddings. In an embodiment,
one or more second vector embeddings may be generated based on the third ained
machine learning model. Further, the ng module 204 may be configured to work with
the computation module 216 for computing a score corresponding to each of the set of
employees. In an embodiment, the score may be computed based on the second pre-trained
machine learning model.
During operation, the reception module 208 may be configured to e the
plurality of collaboration parameters associated with each of the set of employees as the
input data 202. In one embodiment, the reception module 208 may directly receive each of
the plurality of oration parameters from the external devices 118. In another
embodiment, the reception module 208 may fetch the plurality of collaboration parameters
from the database 204. The plurality of collaboration parameters associated with the set of
ees may include, but is not d to at least one or more of employee et,
employee role, employee rating, collaboration complexity, and collaboration satisfaction.
In an embodiment, collaboration complexity may represent level of
oration or assistance provided by an employee to another employee of an
organization. Further, the collaboration complexity may take any values from Low, Medium,
and High. In addition, the oration complexity may take any other ordinal values that
shows order from lowest to highest. er, the employee rating associated with each of
the set of employees may represent rating of an employee providing an assistance to
another employee of same or different team. In an embodiment, employee rating associated
with the employee may be positively impacted when the employee provides an assistance
to another employee. rly, the employee rating associated with the employee may be
negatively impacted when the ee receives an assistance from another employee of
the organization.
The plurality of collaboration parameters may correspond to tabular data
shown in A. Such tabular data corresponding to the plurality of collaboration
parameters may reflect ordinal values of the plurality of oration parameters. The
reception module 206 may be configured to pre-process the input data 202 associated with
the plurality of collaboration parameters having ordinal values into numerical values. For
example, the collaboration parameter corresponding to collaboration complexity associated
with each of the set of employees may be high, medium and low value that is converted to
1, 2 and 3 respectively by the reception module 206, as represented in Fig. 10B. Further, in
r example, the employee rating associated with each of the set of employees may
take any values between 1 to 5. In accordance with an embodiment, the value 1 may
represent lowest rating and 5 may represent highest rating. The rating may be provided
based on the assistance received or provided by at least one employee from the set of
employees. Moreover, in some embodiments, definition of lowest and t may be
different for employee rating.
In accordance with an embodiment, the input data 202 of the plurality of
collaboration parameters with ordinal values may be converted into a passable format for a
first pre-trained machine learning model by using one hot representation for the input data
202. The one hot representation (also known as one hot embedding) may map input data
202 which may be a categorical value data into a Neural k passable format. Such
format may allow to train an embedding layer of the first ained machine learning model
for each of the plurality of performance parameters. The input data 202 with one hot
entation may be fed to hidden layers of the first pre-trained machine learning model
to handle a much smaller size of preprocessed input data as compared to the input data with
ordinal values.
Further, the on module 210 may be configured to create a plurality of
ee nodes associated with the set of employees in a hierarchical tree. The hierarchical
tree may correspond to a graph that comprises two components, namely, the plurality of
employee nodes and plurality of edges. The hierarchical tree may include the plurality of
edges that connects the ity of employee nodes. In accordance with an ment,
the ity of edges may be directed to show directional dependencies between the
ity of employee nodes. In accordance with r embodiment, the plurality of edges
may be undirected. In accordance with an embodiment, the first pre-trained machine
learning model may directly operate on structure of the chical tree. Further, each of
the plurality of edges may interconnect at least two of the set of employee nodes. In
accordance with an embodiment, the creation module 210 may create each of the plurality of
employee nodes based on the plurality of collaboration parameters and the first pre-trained
machine learning model. As will be appreciated, the first pre-trained machine learning model
may correspond to any deep neural network model (for example, an attention based deep
neural network model and a Convolution Neural k (CNN) . Further, the
creation module 210 may be configured to transmit the plurality of employee nodes to the
generation module 212.
Upon receiving the plurality of employee nodes, the generation module 212
may be ured to generate a plurality of vector embeddings associated with the plurality
of ee nodes. In an embodiment, the tion module 212 may generate each of
the plurality of vector embeddings based on the first pre-trained machine model. In
accordance with an embodiment, the generation module 212 may be configured to t
one or more set of edges initiating from a root node of the plurality of the employee nodes
and culminating at an associated leaf node of the plurality of the employee nodes in the
hierarchical tree.
In accordance with an embodiment, to find a required team (say, a team
specializing in e learning domain) for collaboration, the generation module 212 may
generate one or more first vector embeddings of the plurality of vector embeddings
corresponding to one or more first nodes. In addition, the generation module 212 may
generate one or more second vector embeddings of the plurality of vector embeddings
corresponding to one or more second nodes. The one or more second vector ings
may be generated by the tion module 212, based on aggregation of the generated
one or more first vector embeddings by using a third pre-trained machine learning model.
Further, based on the one or more first nodes and the one or more second nodes generated,
the generation module 212 may identify a set of linked nodes that correspond to the required
team for collaboration.
The AI based collaboration system 102 may provide feedback to each of the
set of employees based on collaboration n employees, (such as, at least two
employees from the set of employees). The feedback may then be utilized to evaluate
performance of each of the set of employees. Therefore, in accordance with an embodiment,
the generation module 212 may be ured to generate the feedback for each of the set
of employees. In an embodiment, a positive feedback may be generated for one or more
first ees. The one or more first employee may correspond to employees providing
assistance to one or more of the set of employees. In accordance with an embodiment, a
negative feedback may be generated for one or more second employees. The one or more
second employees may correspond to employees receiving assistance from one or more of
the set of employees.
The plurality of vectors embedding created and the feedback generated may
be stored in database 204 for further computation. It may be noted that the process of storing
the vector ings in the database 204 may continue, until the vector embeddings
associated with each of the plurality of employee nodes is generated and stored. The
plurality of vector embeddings and the generated feedback stored in the database 204 may
further be utilized by the ination module 214.
In one embodiment, the determination module 21 may fetch the plurality of
vector embeddings from the database 204. In another embodiment, the determination
module 214 may receive each of the ity of vector embeddings from the generation
module 212. Upon ing the plurality of vector embeddings, the ination module
214 may be configured to determine a degree of collaboration between at least two
employees from the set of employees.
In order to determine the degree of collaboration, the ation module 216
of the determination module 214 may be configured to compute a score for each of the set
of employees based on collaboration satisfaction. The collaboration satisfaction may
correspond to one of a successful collaboration and an unsuccessful collaboration. In an
embodiment, the computation module 216 may compute the score using the second ined
e learning model. Example of the second pre-trained machine learning model
may include a Rank-Net Neural Network model. Further, the computation module 216 may
be configured to transmit the computed score to the ranking module 218. Upon receiving
the score computed for each of the set of employees, the ranking module 218 may generate
a rank for each of the set of employees based on the computed score associated with each
of the set of employees. In an embodiment, the rank may be ted in order to determine
the degree of collaboration between the at least two employees from the set of employees.
In one ment, the evaluation module 220 may fetch the feedback
generated for each of the set of employees from the database 204. In another ment,
the evaluation module 220 may be configured to receive the feedback generated from the
generation module 212. Upon ing the feedback generated for each of the set of
employees, the evaluation module 220 may be configured to evaluate the performance of
each of the set of employees.
In particular, as will be appreciated by those of ordinary skill in the art, various
modules 206-220 for performing the techniques and steps described herein may be
implemented in the AI based collaboration system 102, either by hardware, software, or
ations of hardware and software. For example, suitable code may be accessed and
executed by the one or more processors on the AI based oration system 102 to
perform some or all of the techniques described herein. Similarly, application specific
integrated circuits (ASICs) configured to perform some or all of the processes described
herein may be included in the one or more processors on the host computing system. Even
though FIGs 1-2 describe about the AI based collaboration system 102, the functionality of
the components of the AI based oration system 102 may be implemented in any
ing devices.
Referring to a flowchart of a method for determining collaboration
between employees is illustrated, in accordance with an embodiment. is explained in
ction with and
With reference to the oration between employees may be
determined based on various es of k environment 100 (for example, the AI based
collaboration system 102, and the server 114). Moreover, various modules depicted within
the memory 104 of the AI based collaboration system 102 in may be configured to
perform each of the steps mentioned in the present
At step 302, a plurality of collaboration parameters may be received. Each of
the plurality of collaboration ters received may be associated with the set of
employees. In an embodiment, each employee from the set of employees may working in a
same team or a different team of an organization. r, each of the plurality of
oration parameters may include, but is not limited, to at least one or more of employee
skillset, employee role, employee rating, collaboration complexity, and collaboration
satisfaction.
Once the plurality of collaboration parameters is received, at step 304, a
plurality of employee nodes may be created ated with each of the set of employees
in a hierarchical tree. The hierarchical tree may include a plurality of edges and each of the
plurality of edges interconnects at least two of the set of employee nodes. In an embodiment,
the plurality of employee nodes may be created based on the collaboration parameters
ed and a first pre-trained machine learning model. Further, the plurality of employee
nodes may pond to employee data associated with the plurality of collaboration
parameters. Moreover, each of the plurality of edges interconnecting the at least two of the
set of employee nodes in the hierarchical tree may correspond to relationship between two
or more employees of the set of employees of s designations.
With reference to the proposed AI based collaboration system 102 may
be order agnostic and doesn’t depend on ic order of values being implemented.
Further, at step 306, a plurality of vector embeddings may be generated. The
plurality of vectors embeddings may be generated based on each of the plurality of
employee nodes. In an embodiment, each of the plurality of vector embeddings is generated
based on the first pre-trained machine learning model. The first pre-trained machine learning
model may be trained as part of transfer learning for determining the degree of collaboration
between the at least two employees.
The one or more set of edges may initiate from a root node of the plurality of
the employee nodes and culminating at an associated leaf node of the ity of the
employee nodes in the chical tree. In an embodiment, the plurality of vector
embeddings may be generated based on a graph neural network. Moreover, each of the
plurality of vectors ing generated may be ed for a plurality of tasks. The plurality
of tasks may e identification of an employee from the set of employees with whom
collaboration can be done, and as an input for evaluating the performance of each of the set
of employees. The process of ting the plurality of vector embeddings has been
ned in greater detail in conjunction to
Thereafter, at step 308, a degree of collaboration may be determined between
two employees from the set of employees. The degree of collaboration may be determined
based on one or more vector embeddings from the generated ity of vector embeddings.
In an embodiment, the degree of collaboration may correspond to association between the
at least two employees. The process of determining the degree of collaboration has been
explained in r detail in ction to
Referring now to a flowchart of a method for generating a plurality of
vector embeddings is illustrated, in accordance with an embodiment. is explained in
conjunction with to
With reference to in order to generate the plurality of vector
embeddings as mentioned in step 306 of a first pre-trained machine learning model
may be used. The first pre-trained machine learning model may correspond to the ML model
112. In an embodiment, the first pre-trained machine learning model may be trained as part
of transfer learning to determine the degree of collaboration between the at least two
employees. er, the first pre-trained machine learning model may be ured to
compute a Q value of each of the set of employees using a reinforcement ng algorithm.
In an embodiment, the computed Q-value may correspond to probability of one employee
from the set of employees that is preferred over other employees from the set of employees
for the collaboration.
In order to generate the plurality of vector embeddings corresponding to each
of the plurality of employee nodes, at step 402, one or more set of edges may be extracted
from the hierarchical tree. The one or more set of edges extracted may initiate from the root
node of the plurality of the employee nodes and ate at an associated leaf node of the
plurality of the employee nodes in the hierarchical tree.
Further, at step 404, one or more first vector embeddings of the plurality of
vector embeddings may be generated. The one or more first vector embeddings may be
ted corresponding to one or more first nodes of the hierarchical tree.
Once the one or more first vector embeddings are generated, at step 406, one
or more second vector embeddings of the plurality of vector embeddings may be generated.
The one or more second vector embeddings may be generated corresponding to one or
more second nodes of the hierarchical tree. In an embodiment, the one or more second
vector embeddings may be generated based on aggregation of the generated one or more
first vector embeddings and a third pre-trained machine learning model. The third pre-trained
machine learning model may correspond to the ML model 112. It should be noted that, the
one or more first nodes may be lower in hierarchy than the one or more second nodes in the
hierarchical tree.
fter, at step 408, a set of linked nodes may be identified from the one
or more first nodes and the one or more second nodes. In an ment, each of the set
of linked nodes may correspond to a required team for collaboration. By way of an example,
the set of linked nodes may depict collaboration between two employees from the set of
employees for a previously ped product in a certain . Further, based on the
set of linked nodes identified, the team required to collaborate on future projects may be
identified. The process of generating the plurality of vector embeddings has been explained
via example in ction to FIGS. 7A-7B and
Referring now to a flowchart of a method for determining a degree of
collaboration between at least two employees is rated, in accordance with an
ment. is explained in conjunction with to
With reference to in order to determine the degree of collaboration
between at least two employees from the set of employees as mentioned in step 308 of a second pre-trained machine learning model may be used. The second pre-trained
machine learning model may correspond to machine learning model 112. In order to
determine the degree of collaboration, at step 502, a score corresponding to each of the set
of employees may be computed. In an embodiment, the score may be computed based on
the collaboration satisfaction among each of two employees from the set of employees using
the second pre-trained machine learning model. The collaboration satisfaction may
correspond to one of a successful collaboration and an unsuccessful oration.
By way of an example, when one employee provides assistance to another
employee of a team in completion of a task and the task gets completed sfully based
on the assistance provided, then the collaboration between those two employees may
correspond to the successful collaboration. Example of the task may correspond to a module
or a product developed by an ee. By way of another example, when one employee
es assistance to another employee of a team in completion of the task and the task
couldn’t get completed based on the assistance provided, then the collaboration between
those two employees may correspond to the unsuccessful collaboration. In an embodiment,
the successful collaboration may be scored higher than the unsuccessful collaboration.
Once the score corresponding to each of the set of employees is ed,
at step 504, a rank may be generated for each of the set of employees. The rank may be
generated based on the computed score. Further, based on the rank generated for each of
the set of employees, the degree of collaboration between at least two employees may be
determined.
Referring now to a flowchart of a method evaluating performance of
each of the set of employees is illustrated, in accordance with an embodiment. is
explained in conjunction with to
In order to evaluate mance of each of the set of employees, at step 602,
one or more first employees and one or more second employees may be identified for each
of the set of employees, based on the determined degree of collaboration between at least
two employees from the set of employees. In an ment, the one or more first
employees may correspond to employees providing assistance to one or more of the set of
employees. In addition, the one or more second employees may correspond to employees
receive assistance from one or more of the set of employees.
Based on identification of the one or more first employees and the one or more
second ees, at step 604, a feedback may be generated for each of the set of
employees. In an embodiment, a positive feedback may be generated for each of the one
or more first ees fied, i.e., the employees ing assistance to one or more
set of employees. A negative feedback may be generated for each of the one or more
second employees, i.e., the employees receiving ance from one or more set of
employees. Further, based on the feedback generated, at step 606, performance of each
of the set of employees may be evaluated. By way of an example, the one or more first
employees providing assistance may be evaluated better than the one or more second
employees receiving assistance.
Referring now to -7B, a pictorial representation of a hierarchical tree
700A and an employee relationship graph 700B for each of a set of employees is depicted,
in accordance with an exemplary ment. – 7B is explained in conjunction with
to
In , the hierarchical tree 700A may consist of a root node 702a, a set
of internal nodes 704a, and a set of leaf nodes 706a. By way of an example, the root node
702a may represent but not limited to, a head of project (also referred as a project head), a
head of department or, a head of an organization. The set of al nodes 704a may
represent but not d to, team leader. Each of the plurality of internal node 704a may be
lower in hierarchy to the root node 702a. Further, each of the leaf nodes 706a may represent
an individual employee in the organization. As will be appreciated, number of leaf nodes
under each of the set of internal nodes may represent number of employees working under
a particular team lead or under a particular ss group. In an embodiment, each of the
leaf nodes 706a may correspond to one or more first nodes. In on, the set of internal
nodes 704a may pond to the one or more second nodes of the hierarchical tree.
In , the hierarchical tree 700A may correspond to a graph that
comprises two components, namely, a plurality of employee nodes and a plurality of edges.
The plurality of employee nodes may include the root node 702a, the set of internal nodes
704a, and the set of leaf nodes 706a. the chical tree 700A may depict some of the
plurality of employees working under a respective team lead from a set of team leads of the
organization.
Each of the set of team leads may be working under project head ‘H1’. Further,
the set of team leads working in the organization may be depicted as ‘Team Lead T1’, ‘Team
Lead T2’ up to ‘Team Lead N’. By way of an example, each of the plurality of employees
working under ‘Team Lead T1’ may be depicted as yee A’ to ‘Employee N1’. Similarly,
the plurality of employees working under ‘Team Lead T2’ may be depicted as ‘Employee B’
to ‘Employee N2’. Further, the plurality of employees working under ‘Team Lead N’ may be
depicted as ‘Employee C’ to ‘Employee N3’. It should be noted that, in the zation there
may be multiple employees working under multiple different team s.
In an embodiment, the at least two of the plurality of employees may
orate with one or more plurality of employees for a particular task, such as, but not
d to, ping a module for a product. For example, the employee A and the
employee N1 may collaborate on a certain project. Since the employee A and the ee
N1 work under the leadership of the team lead T1, the performance evaluation of the team
lead T1 may be positively impacted for encouraging the collaboration between the employee
A and the employeeN1. In another example, the employee B and the ee C may
collaborate for an assigned task. Since the employee B and the employee C work under the
leadership of the team lead T2 and the team lead TN respectively, the performance
evaluation of both the team lead T2 and the team lead TN may be positively impacted for
encouraging the oration among team s (i.e., the employee B and the
employee C).
r, based on the collaboration between the team lead T2 and the team
lead TN, the performance evaluation for the project head ‘H1’ may be positively impacted.
In addition, the collaboration determined amongst the employees may facilitate fication
of ees from the plurality of employees with same skillset associated with each of the
set of employees. This is further explained with reference to .
In , there is shown an employee relationship graph 700B constructed
for four employees of an organization. The employee relationship graph 700B may represent
a plurality of nodes and a set of edges. In the employee onship graph 700B, each of
the ity of nodes, namely, ‘E1’, ‘E2’, ‘E3’, and ‘E4’, may represent the four employees
of the organization. Further, ‘f1’, ‘f2’, ‘f3’ upto ‘fn’ may represent a plurality of collaboration
parameters associated with each of the four employees. The plurality of collaboration
ter may e, but is not limited to, at least one or more of employee skillset,
employee role, employee rating, collaboration complexity, and collaboration satisfaction.
Further, the edges may be directed to show directional dependencies between the nodes
(‘E1’, ‘E2’, ‘E3’, and ‘E4’) based on the assistance provided or received by one of the
employees from the four employees.
By way of an example, in the employee relationship graph 700B, an edge
connecting two employees ‘E1’ and ‘E2’ may depict that the employee ‘E1’ provided the
ance to the employee ‘E2’ in a certain domain (such as, a machine learning domain).
Hence, the employee ‘E1’ may receive a ve feedback and the employee ‘E2’ who is
receiving the assistance may receive a negative feedback. Similarly, an edge connecting
two employees ‘E2’ and ‘E4’ may depict that the employee ‘E2’ provided the assistance to
the employee ‘E4’ on the machine learning domain. In this scenario, the employee ‘E2’ may
be the positively impacted, while the employee ‘E4’ may be negatively impacted. Moreover,
the employee ‘E1’ may also be positively impacted, based on the assistance provided by
the employee ‘E1’ to the employee ‘E2’ in the machine learning domain that further helped
the employee ‘E2’ to provide assistance to the employee ‘E4’. Therefore, the employee ‘E2’
and the ee ‘E4’ may have received the assistance from the ee ‘E1’, where the
employee ‘E2’ may have received the assistance directly from the employee ‘E1’, while the
employee ‘E4’ may have received the assistance indirectly from the employee ‘E1’.
In addition, each of the plurality of vector embeddings corresponding to each
of the plurality of employee nodes may be generated based on corresponding neighboring
nodes in the ee relationship graph 700B in a way similar to word or sentence
embedding in Natural Language Processing (NLP) problems.
Referring now to a plurality of vector embeddings generated for a set
of employees is depicted, in accordance with an embodiment. is explained in
conjunction to – 7B.
In an embodiment, a sub-graph 800 of the hierarchical tree 700A is
represented. The sub-graph 800 represented may depict the plurality of employees, i.e.,
employee ‘A1’, employee ‘A2’ up to employee ‘N1’, working under a team lead ‘T1’ 802 for
a n team. The aph 800 may correspond to a subtree with nodes whose internal
node 802 represents the team lead T1 (or a manager/a reviewer) and the leaf nodes 804
may represent juniors (A1 to N1) of the team lead T1. For example, team represented by
the sub graph 800 may correspond to, but not limited to, a big-data team, a project
management team, and a sales team.
In an ment, the leaf nodes 804 corresponding to A1 to N1 in the ph
800 may be referred as first nodes. The internal node 802, i.e., the team lead ‘T1’ may
be referred as a second node. The second node may be higher in hierarchy than the first
nodes in the subgraph 800. In addition, a plurality of vector embeddings corresponding to
the ees A1 to N1 may be generated. The ity of vector embeddings may be
generated based on each of a ity of collaboration parameters ated with
employees A1 to N1. The plurality of vector embeddings for each of the employees A1 to
N1 may be indicative of the degree of collaboration among counterparts.
In accordance with an embodiment, edges between the nodes may implement
Neural Network models, such as, but not limited to, orward NN, and recurrent NN to
populate information for nodes corresponding to superiors (such as the team lead T1). In
accordance with an embodiment, the vector embeddings for the team lead ‘T1’ may be
generated based on employees A1 to N1 working under him.
In accordance with an embodiment, the vector embeddings for the team lead
‘T1’ for second node may be based on the plurality of vector embeddings of the first nodes
representing employees A1 to N1 and initial vector embeddings of the team lead T1 that is
based on each of the collaboration parameters of the team lead T1 to generate final vector
embeddings for the team lead T1. Such vector embeddings may be generated based on any
neural network implementation. Examples of neural network may include, but is not limited
to, Long Short-Term Memory (LSTM) – Gated Recurrent Units (GRU), LSTM, and GRU.
In accordance with an embodiment, the AI based collaboration system 102
may identify employees with same skillset from the plurality of employees based on the
plurality of vector representations for each node ponding to each of the set of
ees in the organization. By way of an example, employees those have knowledge or
have worked in past in a specific technological domain may collaborate in future for
development of a product of that specific technological domain. Examples of technological
domain may include, but is not limited to, Information Technology (IT), Machine Learning
(ML), Java, Python, Project Management, and Business Development.
ing now to a scenario 900 of collaborating a new ee with
at least one of the set of employees is depicted, in accordance with an exemplary
embodiment. In the scenario 900, a set of two teams i.e., a first team 902 and a second
team 904 in an organization is ed. Both, the first team 902 and the second team 904
may include a set of four employees each. The set of four ees in the first team 902
may include employee ‘E1’, employee ‘E2’, employee ‘E3’ and employee ‘E4’ and hence
represented by nodes E1, E2, E3 AND E4 respectively. Similarly, the set of four employees
in the second team 904 may include employee ‘E6’, employee ‘E7’, employee ‘E8’, and
employee ‘E9’ and hence represented by nodes E6, E7, E8 AND E9 respectively. In
ance with an exemplary embodiment, a new employee ‘E5’ joins the organization. In
accordance with an ment, the AI based collaboration system 102 may identify at least
one ee from the first team 902 and the second team 904 for collaboration with the
new employee ‘E5’ for performing a task in future or for training purpose.
In order to fy the at least one employee for collaboration with the new
employee ‘E5’, the AI based collaboration system may evaluate the plurality of collaboration
parameters ated with the new employee ‘E5’ with the plurality of collaboration
parameter associated with each of the set of four employees of the first team 902 and the
second team 904. As ed in the scenario 900, in one embodiment, based on evaluation
of the plurality of collaboration parameters, the new employee ‘E5’ may collaborate with the
employee ‘E2’ of the first team 902 in order to receive assistance (example: to receive
assistance for training). In an embodiment, in order to identify oration for new
employees, the AI based collaboration system 102 may use a graph based neural networks.
Referring now to A – 10C, tabular representations of input data
corresponding to the ity of collaboration parameters is illustrated, in accordance with
some exemplary ments of the present disclosure. A – 10C is explained in
conjunction with to
With reference to A, the tabular representation 1000A of a dataset (the
input data) corresponding to the plurality of collaboration parameters for a set of employees
is shown. The t may depict the plurality of collaboration parameters captured as the
input data by the AI based collaboration system 102 for each of the set of employees in
order to determine the degree of collaboration among at least two of the set of employees.
In the tabular representation 1000A, a column 1002a represents a serial
number. A column 1004a represents an employee ID for each of a first set of employees
from the set of employees. A column 1006a represents other employee ID associated with
a second set of employees from the set of employees. In an embodiment, the first set of
employees depicted via column 1004a may correspond to the one or more first employees
providing assistance to one or more set. In addition, the second set of ees depicted
via column 1006a may correspond to the one or more second employees receiving
assistance from one or more of the set of employees.
A column 1008a represents an employee’s skillset associated with each of the
first set of employees. Examples for the employee’s skillset may include, but is not limited
to, Python, Dynamic Programming, NPL, Microsoft- Structure Query Language L)
database, Java, and ML. In some embodiments, the tabular entation 1000A may
include any ation of the employee’s skillset depending upon expertise of each of the
first set of employees. A column 1010a may represent a role of each of the first set of
employees working in an organization either in same team or ent teams. As depicted
in the table 1000A, the role of employee ‘E1’ may be of a senior developer in the
organization. Similarly, the role of employee ‘E2’ may be of a developer in the zation.
In addition, the role of employee ‘E3’ may be of a data scientist in the organization. A column
1012a represents a complexity of collaboration. A column 1014a represents employees
rating (also referred as ranking) associated with each of the first set of employees. A column
1016a represents a score provided based on collaboration satisfaction of the set of
In an embodiment, the collaboration satisfaction may correspond to one of a
successful collaboration and an essful oration. Moreover, the successful
collaboration may be scored higher than the unsuccessful oration. In an embodiment,
based on assistance provided by each of the first set of ees to at least one of the
second set of employees, corresponding values for the column 1012a (complexity of
oration), the column 1014a (employees rating), and the column 1016a (collaboration
satisfaction) may be predicted. By way of an example, an ordinal value associated with the
complexity of collaboration may correspond to low, medium, and high. Similarly, the ordinal
value associated with the employees rating may correspond to a good performer, an
average performer, and an excellent performer.
In addition, the ordinal values associated with the collaboration satisfaction
may range from a value ‘1’ to a value ‘5’. In an embodiment, the l value ‘1’ for the
collaboration satisfaction may depict the unsuccessful collaboration. However, the value ‘5’
for the oration satisfaction may depict the successful oration. The data populated
in the table 1000A may not be suitable as a passable format for a graph based neural
network, such as the first pre-trained machine learning model. Hence, the data populated in
the table 1000A may be pre-processed by the AI based collaboration system 102 as shown
in B.
The tabular representation 1000B may represent numerical values of the
plurality of collaboration parameters captured for each of the set of employees. The AI based
collaboration system 102 may be configured to convert input data with ordinal values as
shown in A into numerical values. There is shown, Employee ID 1004b, other
member ID 1006b, complexity of collaboration 1012b, employee rating 1014b, and
collaboration satisfaction 1016b. As an example, column 1014 b with name “employee’s
rating” have values such as 1, 2 and 3 where “1” may replace “Low” and “3” may replace
. In some other embodiments, one-hot entation (also referred as one hot
embeddings) of ordinal values may be generated by the AI based collaboration system 102
where new features / columns may be introduced equal to number of unique values in
original column of the tabular representation 1000A. For example, columns of collaboration
parameters with multiple values (such as, Column: employee’s skillset) may be ted
to unique numeric values. In order to represent values numerically for the collaboration
ters, the AI based collaboration system 102 may be configured to convert such
values into one-hot representation.
r, in t representation, embedding layer of the first trained
machine learning model may have vector representation for a number of dimensions equal
to number of unique values (T1 to T5 of 1000B) in certain column. Column ‘T1’, ‘T2’, ‘T3’,
‘T4’, and ‘T5’ may represent unique numerical values based on a type of technologyor
language in which each of the first set of employees may be d in). By way of an
example, the collaboration parameter “employee’s skillset” may be ented numerically
in T1 to T5 of 1000B, such as, Python: [1 0 0 0 0 0], Java: [0 1 0 0 0 0], Machine Learning:
[0 0 1 0 0 0], Natural ge Processing: [0 0 0 1 0 0], MS SQL database: [0 0 0 0 1 0]
and Dynamic Programming: [0 0 0 0 0 1].
Thereafter, a graph may be constructed based on the plurality of collaboration
parameters associated with the first set of employees and the assistance provided by each
of the first set of employees to at least one of the second set of ees. Once the graph
is constructed, a set of linked nodes may be identified from the graph. In an embodiment,
each of the set of linked nodes may be based on any graph algorithm (for example: random
walk). In addition, the AI based collaboration system 102 may use the graph thm to
generate each of the ity of vector embeddings. In some embodiment, the plurality of
vector embeddings may be ted based on BERT (Bi-directional Encoder
Representations from Transformer) embedding (n case of any textual e). In on,
the BERT embeddings may be used to generate low dimensional vector representation of
the nodes representing each of the set of ees in the graph.
A tabular representation 1000C represents a number of successful
collaborations corresponding to each of the first set of employees. A column 1002c may
represent a serial number. A column 1004c may represent employees rating. The
employees rating may be based on assistance provided by each of the first set of employees
to at least one of the second set of employees. A column 1006c may ent employee
ID. A column 1008c may represent the number of successful collaborations of each of the
first set of employees with at least one of the second set of employees. Further, based on
the number of collaborations determined associated with each of the first set of employees,
an employee with highest collaboration from the first set of employees may be identified.
With reference to tabular representation 1000c, it may be depicted that the
employee E2 may not be an excellent performer like employees E1 and E3, however, the
employee E2 may be a better collaborator as compared to the employees E1 and E3. In
accordance with an embodiment, the AI based collaboration system 102 may be configured
to rank employee E2 higher as compared to the employees E1 and E3, based on a number
of successful collaborations (1008c) to determine a degree of oration between the at
least two employees of an organization.
Referring now to , a trained AI based collaboration system based on a
reinforcement learning is rated, in accordance with an exemplary embodiment.
is explained in ction with to C.
There is shown a model 1102, training data 1104 with a set of employees’ data
1106 and Q-learning algorithm 1108, apply model 1110, a test set of employees’ data 1112,
and employee’s collaboration satisfaction s 1114. In accordance with an embodiment,
the model 1102 may correspond to a trained oration system, such as the AI based
collaboration system 102. In accordance with an embodiment, the model 1102 may be
exposed to new training data 1104 when the model 1102 has never been through earlier
training process. The model 1102 may leverage any similarity measures such as cosine
similarity to find out employees with similar skillset. For example, when two employees have
same years of experience and has worked on almost same level of expertise in a particular
skillset then vector embeddings representation of those two employees may be near to each
other in vector space.
As a result, the model 1102 may learn to identify optimal reward function that
will ze reward for end goal of identifying employee with maximum successful
collaboration. In accordance with an embodiment, the set of employees’ data 1106 may
correspond to information associated with each of the set of employees. The information
may include number of successful collaborations associated with each of the set of
employees, computed score generated based on the collaboration satisfaction, the ity
of collaboration parameters, etc. Further, the Q-value algorithm 1108 may be used to
calculate a Q-value corresponding to each of the set of employees. The Q-value may be
calculated based on the reinforcement learning approach. In on, the feedback
associated with each of the set of employees may be predicted based on reinforcement
ng approach.
In an embodiment, the Q-value represents preference of a ular employee
over other employees from the set of employees across all values of the collaboration
satisfaction or employee’s rating. In other words, the Q-value may represent probability of
one employee being preferred over the other employees across different values of the
oration satisfaction or employee’s rating. Based on the calculated Q-value, the model
1102 may penalize the team leader or the ce head for giving incorrect score for the
collaboration satisfaction or employee’s rating to one employee over the other employees
from the set of employees. Moreover, the Q-values each of the set of employees along with
the associated collaboration satisfaction or employee’s rating may be used to maximize
reward.
Based on the ng data 1104 received, the model 1110 may be ted
for a test set of employees’ data 1112. The test set of employees’ data may correspond to
information associated with a new set of employees. The score provided for the oration
satisfaction to each of the test set of employees’ data may be depicted as collaboration
satisfaction 1114. With reference to the the first pre-trained machine learning model
corresponds to a Q network. The Q network may be configured to receive as input an input
observation corresponding to set of employees’ data and an input action and to generate an
estimated future reward (or penalty) from the input in accordance with each of the plurality
of collaboration parameters associated with the set of employees.
Referring now to , a trained collaboration system that uses inverse
reinforcement learning is illustrated, in accordance with an exemplary embodiment.
is explained in conjunction with to . There is shown an nmental model
1202, an e reinforcement learning model 1204, historical data 1206, policy 1208,
relevant algorithm combinations 1210, and algorithm set satisfying historical data 1212.
The reinforcement learning based trained collaboration system may
correspond to the environment model 1202. The nment model 1202 may correspond
to the apply model 1210. The environment model 1202 may employ the e
reinforcement ng model 1204. The inverse reinforcement learning model 1204 may be
configured to e the historical records 1206 to penalize and boost chances of an
employee or team of an organization to be ered for future collaboration. The historical
records 1206 may use various policies, such as the policy 1208 to penalize and boost
chances of an employee or team to be considered for future collaboration.
In an embodiment, the historical records 1206 may include detailed
information about each of the set of employees from various teams in an organization along
with the assistance provided by one employee to another employee in same team or
different team. Thereafter, the inverse reinforcement ng model 1204 may identify
combination or set of algorithms and function that will define ecture of deep learning
based recurrent neural network variations and define hyperparameter for ent layers of
a neural network. The combination or set of algorithms and function may be represented as
relevant algorithm ation 1210. In an embodiment, the e rcement learning
model 1204 may recommend more than one combination of set of algorithms and functions.
Further, the recommended combination of set of algorithms and functions may
be evaluated based on the reinforcement learning approach in order to accept one
combination of set of algorithms and ons. Moreover, one combination of set of
algorithms and functions may be accepted when it satisfies evaluation of historical records
represented as algorithm set satisfying historical records 1212. Once the one combination
of set of algorithms and functions is accepted, a new environment may be created for the
environment model 1202. In addition, the inverse reinforcement learning model 1204 may
recommend optimal values of hyperparameters corresponding to each combination of set of
algorithms and functions. Further, the optimal values of hyperparameters may be validated
against historical data received from an existing environment of the environment model
1202. This process is known as model hyperparameter tuning.
Referring now to , a transfer learning approach to create a new
environment for an AI based collaboration system is depicted, in accordance with an
exemplary embodiment. is explained in conjunction with to . There is
shown a ained model 1302, a set of collaborated employees 1304 associated with the
pre-trained model 1302, a new model 1306, and a set of collaborated ees 1308
associated with the new model 1306.
In an embodiment, the transfer learning approach may be used to leverage
training of an AI based collaboration system (such as, the AI based collaboration system
102) from previous implementation to new implementation. The new model 1306 may
correspond to the new environment ted for the environment model 1202 based on
acceptance of one combination of the set of algorithms and functions. The new model 1306
may receive the optimal values of hyperparameters ented as extracted pre-trained
hyperparameters from the pre-trained model 1302.
fter, the new model 1306 may identify a degree of collaboration
between at least two employees from the set of employees based on the optimal values of
hyperparameters received from the pre-trained model 1302. In an embodiment, the transfer
learning approach may enable ing of knowledge from an ng environment or
implementation of the AI based collaboration system 102. The knowledge corresponds to
optimal values (i.e., the plurality of vector ings) of the plurality of oration
parameters and hyperparameter required for the implementation of the AI based
collaboration system 102. Further, the optimal values of the plurality of collaboration
parameters and hyperparameter may be utilized to develop the new environment for the AI
based collaboration system 102. This may require less training time as compared to starting
from scratch or from vanilla model. The vanilla model may correspond to a standard, usual,
and unfeatured version of the AI based collaboration system 102.
In accordance with an embodiment, the AI based collaboration system 102
may be configured to modify the first pre-trained machine ng model (such as, the
environment model 1202) with erable knowledge for a target system to be evaluated.
The transferable knowledge may pond to optimal values associated with the ity
of vector embeddings ponding to each of the plurality of collaboration parameters.
In accordance with an embodiment, the AI based oration system 102
may be configured to tune the first pre-trained machine ng model (such as, the pretrained
model 1302) using specific characteristics of the target system to create a target
model (such as, the new model 1306). In accordance with an embodiment, the AI based
collaboration system 102 may be configured to evaluate the target system mance
using the target model (such as, the new model 1306) to predict system mance of the
target system for determining the degree of collaboration among at least two employees
from the set of employees working in an organization.
Further, the AI based collaboration system 102 may enable a plurality of
employees working in an organization to leverage its usage. In certain other scenario, an
employee from a set of employees who needs help or assistance from other employees may
leverage the AI based collaboration system 102. By way of an example, the employee may
ask query like “Can you help me to find out developer who is an expert in machine learning?”
via a user interface 110 of the AI based collaboration system 102. As a response, the AI
based collaboration system 102 may connect the employee via REST API sentational
State Transfer Application Programming Interface) to get details of employee and
render/display response using the I/O s 108.
Various embodiments provide a method and system for determining
collaboration between employees. The disclosed method and system may receive a ity
of collaboration parameters associated with a set of employees. The system and method
may then create a plurality of employee nodes ated with the set of employees in a
hierarchical tree. The plurality of employee nodes may be created based on the plurality of
collaboration parameters and a first pre-trained machine learning model. Further, the system
and the method may generate a plurality of vector embeddings associated with the plurality
of ee nodes, based on the first pre-trained machine learning model. Thereafter, the
system and the method may determine a degree of collaboration n at least two
employees from the set of employees based on one or more vector embeddings from the
generated plurality of ings.
The system and method provide some advantages like the disclosed system
and the method may provide an AI based collaboration system for determining collaboration
between employees of any roles in an zation. The AI based collaboration system may
enable organization for efficient utilization of resources. Further, the sed AI based
collaboration system may utilize the collaboration, or the assistance ation associated
with employees to provide rating for performance evaluation to each employee working in
the organization. This may boost resource management and improve the overall
performance and productivity of different teams working in the organization. In addition, the
AI based collaboration system may facilitate working of the zation much more
efficiently because of the collaboration. Also, resources ated with handling of
collaboration task manually, may be available for other tasks. Moreover, the disclosed AI
based collaboration system may enable team leader or practice head to find employee of
particular skillset due to unplanned absence of a particular employee.
It will be appreciated that, for clarity purposes, the above description has
described embodiments of the disclosure with reference to ent functional units and
processors. However, it will be apparent that any suitable distribution of onality
between ent onal units, processors or s may be used without detracting
from the disclosure. For example, functionality illustrated to be performed by separate
processors or controllers may be performed by the same processor or controller. Hence,
references to specific functional units are only to be seen as references to suitable means
for providing the described functionality, rather than indicative of a strict logical or physical
structure or organization.
Although the present disclosure has been described in connection with some
embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the
scope of the present disclosure is limited only by the claims. Additionally, although a feature
may appear to be described in tion with particular embodiments, one skilled in the
art would recognize that s es of the described embodiments may be combined
in accordance with the disclosure. Furthermore, although individually listed, a plurality of
means, elements or process steps may be ented by, for example, a single unit or
processor. Additionally, although individual features may be included in ent claims,
these may possibly be advantageously combined, and the inclusion in different claims does
not imply that a combination of features is not feasible and/or advantageous. Also, the
inclusion of a e in one category of claims does not imply a limitation to this category,
but rather the feature may be equally applicable to other claim categories, as appropriate.
Claims (19)
1. A method determining collaboration between employees, the method comprising: receiving, by an AI based collaboration system, a ity of collaboration parameters associated with a set of ees; creating, by the AI based collaboration system, a plurality of employee nodes associated with the set of employees in a hierarchical tree, based on the plurality of oration parameters and a first pre-trained e learning model, wherein the hierarchical tree r comprises a plurality of edges and each of the plurality of edges interconnects at least two of the set of employee nodes; generating, by the AI based collaboration system, a plurality of vector embeddings associated with the plurality of employee nodes, based on the first pre-trained machine learning model; and determining, by the AI based oration system, a degree of collaboration n at least two employees from the set of employees based on one or more vector embeddings from the generated plurality of embeddings, wherein the degree of collaboration corresponds to association between the at least two employees.
2. The method of claim 1, wherein the plurality of collaboration parameters associated with the set of employees comprises at least one or more of employee skillset, employee role, employee rating, collaboration complexity, and collaboration satisfaction.
3. The method of claim 2, wherein determining the degree of collaboration between the at least two employees from the set of employees comprises: computing a score for each of the set of employees based on the collaboration satisfaction using a second ained machine learning model, wherein the collaboration action corresponds to one of a successful collaboration and an unsuccessful collaboration, and n the successful collaboration is scored higher than the unsuccessful collaboration; and ranking each of the set of employees based on the associated computed score to determine the degree of oration between the at least two employees.
4. The method of claim 1, n tingo the plurality of vector embeddings comprises extracting one or more set of edges initiating from a root node of the plurality of the employee nodes and culminating at an associated leaf node of the plurality of the employee nodes in the hierarchical tree.
5. The method of claim 1, further comprising: identifying one or more first employees and one or more second employees for each of the set of employees, wherein the one or more first employees provide assistance to one or more of the set of employees and the one or more second employees receive assistance from one or more of the set of employees; generating a feedback for each of the set of employees, wherein a positive feedback is generated in response to identifying the one or more first employees and a negative ck is generated in response to identifying the one or more second employees; and evaluating the performance of each of the set of employees, based on the feedback.
6. The method of claim 1, wherein ting the ity of vector embeddings comprises: generating one or more first vector embeddings of the plurality of vector embeddings corresponding to one or more first nodes; ting one or more second vector embeddings of the plurality of vector embeddings corresponding to one or more second nodes, based on aggregation of the generated one or more first vector embeddings and a third pre-trained machine learning model, and wherein the one or more first nodes are lower in hierarchy than the one or more second nodes in the hierarchical tree; and identifying a set of linked nodes from the one or more first nodes and the one or more second nodes, wherein the set of linked nodes correspond to a ed team for collaboration.
7. The method of claim 1, wherein each node of the plurality of employee nodes corresponds to employee data associated with the plurality of collaboration parameters, and wherein each of the plurality of edges interconnecting the at least two of the set of ee nodes corresponds to relationship between two or more employees of the set of employees of various designations.
8. The method of claim 1, wherein the first pre-trained machine learning model is trained as part of transfer learning to determine the degree of collaboration between the at least two employees.
9. The method of claim 1, wherein the first ained machine learning model is configured to compute a Q value of each of the set of employees using a reinforcement learning algorithm, and wherein the Q value corresponds to probability of one employee from the set of employees being preferred over other employees from the set of employees for the collaboration.
10. A system for determining collaboration n employees, the system sing: a processor; and a memory communicatively coupled to the processor, wherein the memory stores processor executable instructions, which, on execution, causes the processor to: receive a plurality of collaboration parameters associated with a set of employees; create a plurality of employee nodes associated with the set of employees in a hierarchical tree, based on the plurality of collaboration parameters and a first pretrained machine ng model, wherein the hierarchical tree further comprises a plurality of edges and each of the plurality of edges interconnects at least two of the set of employee nodes; te a plurality of vector embeddings associated with the plurality of employee nodes, based on the first ained machine learning model; and determine a degree of collaboration between at least two employees from the set of employees based on one or more vector embeddings from the generated ity of embeddings, wherein the degree of oration corresponds to association between the at least two employees.
11. The system of claim 10, n the plurality of collaboration parameters associated with the set of employees comprises at least one or more of employee skillset, employee role, employee rating, collaboration complexity, and collaboration action.
12. The system of claim 11, wherein the processor executable instructions cause the processor to determine the degree of oration between the at least two employees from the set of employees by: computing a score for each of the set of employees based on the collaboration action using a second pre-trained machine learning model, wherein the collaboration satisfaction corresponds to one of a successful collaboration and an unsuccessful collaboration, and wherein the successful collaboration is scored higher than the unsuccessful collaboration; and ranking each of the set of employees based on the associated computed score to determine the degree of collaboration between the at least two employees.
13. The system of claim 10, wherein the sor executable instructions cause the processor to te the plurality of vector embeddings by extracting one or more set of edges initiating from the root node of the plurality of the employee nodes and culminating at an associated leaf node of the plurality of the employee nodes in the hierarchical tree.
14. The system of claim 10, wherein the processor executable instructions cause the processor to: identify one or more first employees and one or more second employees for each of the set of employees, wherein the one or more first employees provide assistance to one or more of the set of employees and the one or more second employees receive ance from one or more of the set of employees; generate a feedback for each of the set of employees, wherein a positive feedback is generated in response to identifying the one or more first employees and a negative feedback is generated in response to identifying the one or more second employees; and evaluate the mance of each of the set of employees, based on the feedback.
15. The system of claim 10, wherein the processor executable instructions cause the processor to te the plurality of vector embeddings by: generating one or more first vector embeddings of the plurality of vector embeddings corresponding to one or more first nodes; generating one or more second vector embeddings of the plurality of vector ings corresponding to one or more second nodes, based on aggregation of the generated one or more first vector ings and a third pre-trained machine ng model, and n the one or more first nodes are lower in hierarchy than the one or more second nodes in the hierarchical tree; and identifying a set of linked nodes from the one or more first nodes and the one or more second nodes, wherein the set of linked nodes correspond to a ed team for collaboration.
16. The system of claim 10, wherein each node of the plurality of employee nodes corresponds to employee data associated with the plurality of collaboration parameters, and wherein each of the plurality of edges onnecting the at least two of the set of ee nodes corresponds to relationship between two or more employees of the set of employees of various designations.
17. The system of claim 10, wherein the first pre-trained machine ng model is trained as part of transfer learning to determine the degree of collaboration between the at least two employees.
18. The system of claim 10, wherein the first pre-trained machine learning model is configured to compute a Q value of each of the set of employees using a reinforcement learning algorithm, and wherein the Q value ponds to probability of one employee from the set of employees being red over other employees from the set of employees for the collaboration.
19. A non-transitory computer-readable medium storing computer-executable instructions for determining collaboration between ees, the stored instructions, when ed by a processor, cause the processor to perform operations comprising: receiving a plurality of collaboration parameters associated with a set of employees; creating a plurality of employee nodes associated with the set of ees in a hierarchical tree, based on the plurality of collaboration parameters and a first pre-trained machine learning model, wherein the hierarchical tree further comprises a plurality of edges and each of the plurality of edges interconnects at least two of the set of employee nodes; generating a plurality of vector embeddings associated with the plurality of employee nodes, based on the first pre-trained machine ng model; and ining a degree of collaboration between at least two employees from the set of employees based on one or more vector embeddings from the generated plurality of embeddings, n the degree of collaboration corresponds to association between the at least two employees.
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US17/199,829 | 2021-03-12 |
Publications (1)
Publication Number | Publication Date |
---|---|
NZ785400A true NZ785400A (en) | 2022-02-25 |
Family
ID=
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11144831B2 (en) | Regularized neural network architecture search | |
Grady et al. | Agile big data analytics: AnalyticsOps for data science | |
CN109766454A (en) | A kind of investor's classification method, device, equipment and medium | |
AU2024200940A1 (en) | Method and system for determining collaboration between employees using artificial intelligence (ai) | |
CN109784654A (en) | Task creating method, device, computer equipment and storage medium | |
US20190228297A1 (en) | Artificial Intelligence Modelling Engine | |
CN110263136B (en) | Method and device for pushing object to user based on reinforcement learning model | |
AU2024200941A1 (en) | Method and system for evaluating performance of developers using artificial intelligence (ai) | |
Jackson et al. | From natural language to simulations: Applying gpt-3 codex to automate simulation modeling of logistics systems | |
EP3961507A1 (en) | Optimal policy learning and recommendation for distribution task using deep reinforcement learning model | |
US20240012694A1 (en) | System and method for recommending an optimal virtual machine (vm) instance | |
US20240013131A1 (en) | Systems and methods for managing decision scenarios | |
CN110969549B (en) | Judicial data processing method and system | |
NZ785400A (en) | Method and system for determining collaboration between employees using artificial intelligence (ai) | |
CN116089595A (en) | Data processing pushing method, device and medium based on scientific and technological achievements | |
Yuan et al. | Practice summary: Cainiao optimizes the fulfillment routes of parcels | |
US20220292393A1 (en) | Utilizing machine learning models to generate initiative plans | |
US20200294167A1 (en) | Systems and methods for aiding higher education administration using machine learning models | |
US11941500B2 (en) | System for engagement of human agents for decision-making in a dynamically changing environment | |
US20230297850A1 (en) | GENERATING CONTEXTUAL ADVISORY FOR XaaS BY CAPTURING USER-INCLINATION AND NAVIGATING USER THROUGH COMPLEX INTERDEPENDENT DECISIONS | |
NZ785402A (en) | Method and system for evaluating performance of developers using artificial intelligence (ai) | |
CN116976640B (en) | Automatic service generation method, device, computer equipment and storage medium | |
EP4246378A1 (en) | Meta-learning model training based on causal transportability between datasets | |
US20240127912A1 (en) | Systems and methods for generating customizable digital data structures | |
US20230368130A1 (en) | Systems and methods for prioritizing orders |