WO2016048132A1 - Intelligent collaborative learning system - Google Patents

Intelligent collaborative learning system Download PDF

Info

Publication number
WO2016048132A1
WO2016048132A1 PCT/MY2015/050105 MY2015050105W WO2016048132A1 WO 2016048132 A1 WO2016048132 A1 WO 2016048132A1 MY 2015050105 W MY2015050105 W MY 2015050105W WO 2016048132 A1 WO2016048132 A1 WO 2016048132A1
Authority
WO
WIPO (PCT)
Prior art keywords
nodes
learners
subnetwork
profile
learner
Prior art date
Application number
PCT/MY2015/050105
Other languages
French (fr)
Inventor
Khalil BOUZEKRI
Jasbeer Singh A/l ATMA SINGH
Farouq Hatem HAMAD
Dickson Lukose
Original Assignee
Mimos Berhad
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mimos Berhad filed Critical Mimos Berhad
Publication of WO2016048132A1 publication Critical patent/WO2016048132A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G06Q10/101—Collaborative creation, e.g. joint development of products or services
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B5/00—Electrically-operated educational appliances
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers

Definitions

  • the invention relates to an intelligent collaborative learning system, in particular on how the learning groups are composed.
  • An aim of the invention is to provide a collaborative learning system which is dynamic in nature in order to overcome the above problem.
  • a collaborative learning system for a network of learners comprising:
  • each node having a profile and representing a learner or a group of learners selected from the network, forming a subnetwork;
  • edges between the nodes representing relationships between the nodes
  • pairs of nodes being selectable such that if the respective profiles and/or relationships are similar enough the nodes can be merged to form one node with a single profile
  • the selected learners are thus grouped into sets in a dynamic manner, such that if the profiles change over time, or a member leaves, a suitable replacement can be automatically found while substantially maintaining the learning attributes of the group.
  • the system allows a good collaborative environment to be formed by the learners of the group with the same tuition, and/or have personalised tuition according to difficulty, and/or have the same tuition taught using different materials according to the group profile.
  • the profile includes attributes including any or any combination of proficiency, interaction history, cognitive profile, and/or the like.
  • the relationships include any or any combination of validity of proficiencies, cognitive similarity, collaborative willingness, and/or the like. Typically the relationships are weighted.
  • nodes are merged and/or replacements are selected on the basis that the profiles of two nodes and the collaborative willingness therebetween substantially match. It will be appreciated that match in this context means above a predefined threshold of similarity, but does not necessarily mean identity.
  • the learners that wish to join the subnetwork but are not selected therefor are tracked as potential replacements.
  • the suitable replacement learners are disregarded from being selected if they are not available.
  • the nodes are merged if the similarities between the attributes and/or relationships are within predefined threshold ranges.
  • an iterative strategy is used to merge nodes until the set has enough learners to meet a predefined threshold.
  • the nodes with highest similarity are grouped first, and then successive strategies are used to add further nodes in order of decreasing similarity.
  • FIG. 1 illustrates a schematic overview of the invention.
  • Figure 2 illustrates the process flow for managing the network.
  • Figure 3 illustrates a network built from the example in Table 1.
  • FIG. 4 illustrates the process flow of the Learners Grouping component
  • Figure 5 illustrates the process flow of the Personalized Pairing of Nodes component
  • Figure 6 illustrates the process flow of the Select Personalized Pair of Nodes component
  • FIG. 7 illustrates the process flow of the Index Selected Nodes component
  • FIG 8 illustrates the process flow of the Learner Replacement component
  • FIG. 1 there is illustrated a schematic overview 2 of the invention in which learners are selected 4 from a network 6, grouped 8 into a set according to their profiles as defined by the strategies 10, wherein if a learner leaves the set, a suitable replacement learner 12 is automatically selected from the network.
  • Table 1 below provides an example of several profiles, with the following attributes:
  • Proficiency the competency of a learner with regard to a learning goal (e.g.
  • Cognitive profile includes learning styles (e.g. visual, logical) or multiple intelligence (e.g. interpersonal, intrapersonal) each sub-attribute having a defined range (e.g. from 0 to 100, where 0 means the learner is not of this attribute, 100 the learner is maximally of this attribute).
  • learning styles e.g. visual, logical
  • multiple intelligence e.g. interpersonal, intrapersonal
  • John has the following attributes:
  • Figure 2 depicts the process flow 100 for managing the network, in this case adding a new learner into the network.
  • Process starts by creating 102 a new node in the network which is linked to the learner L (already present in Learner Model 104),
  • learner L is part of the network and linked to other learners by Collaborative Willingness (CW) and Cognitive Profile Similarity (CPS) edges.
  • CW Collaborative Willingness
  • CPS Cognitive Profile Similarity
  • Figure 3 depicts the network 122 built from the example in Table 1.
  • Figure 4 depicts the process flow 200 of the Learners Grouping component, which allows learners to be grouped in a personalized way (including profiling of groups) for a specific learner goal, number of learners necessary to start and groups size:
  • System takes as input a number of learners n, a group size s and a learning goal.
  • Process starts by sending a call 202 to the network for anyone interested to join this collaborate e-learning session.
  • Step 3 After a certain time 204 waiting for people, if there are not enough learners, the process goes back to Step 2, otherwise 206 it chooses 208 n learners out of all who accepted to join the collaborative e-learning session (e.g. randomly chosen or based on any parameters such as privileges etc.)
  • the other learners not chosen at Step 3 are saved 210 in Potential Replacements 212 for future use by the replacement process. 5.
  • the sub-network composed of only the nodes corresponding to the learners chosen is then extracted 214 from the full network of learners 106.
  • Figure 5 depicts the process flow of the Personalized Pairing of Nodes component 222, which allows pairing of two nodes in the network by selecting the most personalized pair of nodes 302, profiling the two selected nodes and merging the two selected nodes in the sub-network. The process follows as below:
  • the Select Personalized Pair of Nodes component receives the network of learners and selects the best pair in the network according to Relationships Importance Strategies 224 (see Figure 6).
  • the two selected nodes are then profiled/indexed 304 using a series of indexing strategies 226 (see Figure 7). 3. Finally the two selected nodes are grouped/merged 306 in the sub-network 215 and indexed using the index of Step 2. The two nodes are merged together, adding a link to the pair's attributes previously set in Step 2, then collaborative willingness, cognitive profile similarity and validity edges are added between the node resulting from the merging and the other nodes in the network.
  • Figure 6 depicts the process flow of the Select Personalized Pair of Nodes component 302, which allows selecting the most personalized pair of nodes based on a series of strategies. The process follows as below:
  • the Select Next Strategy component 402 receives the series of ordered strategies and selects the first strategy which has not already been selected. Examples of strategies include:
  • Step 1 If a strategy has been selected 404 (i.e. all the strategies have not been tested yet) then the selected strategy is applied 406 to the network 215. If there are any results 408, one is randomly chosen 410 (i.e, there can be several results after applying one strategy) and the process ends, otherwise repeat Step 1.
  • Figure 7 depicts the process flow of the Index Selected Nodes component 304, which allows indexing/profiling of two nodes selected in the network based on a series of grouping strategies.
  • the process follows as below: 1.
  • the cognitive profile 506 of the given pair of nodes is set according to Group Cognitive Profile Strategies226c [e.g. ⁇ Strategyl : average ⁇ , and the nodes are Mary and Jack, then the cognitive profile of the pair (Mary, Jack) is (65,40,60)] .
  • Figure 8 depicts the process flow of the Learner Replacement component 600, which allows dynamically replacing a learner in a group. The process follows as below:
  • the system takes as input the learner leaving the group, say 1. 2.
  • the process then updates 602 the group of 1 say g, by removing 1 from g and re- indexing the group's attributes (re-use previous process in Figure 5 - Index Select Groups - by just inversing the outcomes, because a member is deleted rather than being added)
  • the process then checks 604 if the potential replacements 212 previously saved are available, and filters out non-available replacements (e.g. not online at the moment, or in another collaborative e-learning session, etc)
  • the process then ranks 606 the available potential replacements based on the 'closeness' of the potential replacement with the group g by using a series of strategies (re-use previous process in Figure 6 - apply strategies to network - by just applying to the sub-sub-network composed of the group g and the potential replacement)
  • Index/Profile 622 the pair ⁇ and g (same as process in Figure 5 - Index Selected Nodes)
  • Group 624 the pair ⁇ and g (same than process in Figure 5 - Group Nodes)
  • collaborative learning groups formed are dynamic in that if a learner leaves a group, they are replaced with the most appropriate replacement learner identified thereby substantially maintaining the learning attributes of the group.

Abstract

A collaborative learning system (2) for a network of learners (6) where a subnetwork (215) of representative nodes is created, each node having a learning profile, and edges representing relationships between the nodes wherein learners not selected for 5 the subnetwork are tracked (210) such that if a learner leaves the subnetwork, a suitable replacement (212) is selected (620) based on profile and/or relationship similarity between the replacement learner and the subnetwork nodes.

Description

INTELLIGENT COLLABORATIVE LEARNING SYSTEM
Field of Invention
The invention relates to an intelligent collaborative learning system, in particular on how the learning groups are composed.
Background
In learning systems, it is known to group students according to their respective proficiencies in the subject being taught. This allows the subject to be taught at an appropriate pace to suit each group, which enhances the learning experience by making the learning more effective.
However, this grouping method is static, in that the selection of students is made at the start of the course.
In collaborative learning systems, where two or more people learn something together, students may leave and/or join at any time. Therefore the existing grouping method is inappropriate as the students change over time and as a result the group becomes more heterogeneous with a wider range of proficiencies that have to be catered for, thereby reducing the learning efficiency. An aim of the invention is to provide a collaborative learning system which is dynamic in nature in order to overcome the above problem.
Summary of Invention
In an aspect of the invention, there is provided a collaborative learning system for a network of learners comprising:
a plurality of nodes, each node having a profile and representing a learner or a group of learners selected from the network, forming a subnetwork;
edges between the nodes representing relationships between the nodes;
pairs of nodes being selectable such that if the respective profiles and/or relationships are similar enough the nodes can be merged to form one node with a single profile
characterised in that learners not selected from the network are tracked such that if a learner leaves the subnetwork, a suitable replacement is selected based on profile and/or relationship similarity between the replacement learner and the subnetwork nodes.
Advantageously, the selected learners are thus grouped into sets in a dynamic manner, such that if the profiles change over time, or a member leaves, a suitable replacement can be automatically found while substantially maintaining the learning attributes of the group.
In addition, the system allows a good collaborative environment to be formed by the learners of the group with the same tuition, and/or have personalised tuition according to difficulty, and/or have the same tuition taught using different materials according to the group profile.
In one embodiment the profile includes attributes including any or any combination of proficiency, interaction history, cognitive profile, and/or the like.
In one embodiment the relationships include any or any combination of validity of proficiencies, cognitive similarity, collaborative willingness, and/or the like. Typically the relationships are weighted.
In one embodiment nodes are merged and/or replacements are selected on the basis that the profiles of two nodes and the collaborative willingness therebetween substantially match. It will be appreciated that match in this context means above a predefined threshold of similarity, but does not necessarily mean identity.
In one embodiment the learners that wish to join the subnetwork but are not selected therefor are tracked as potential replacements.
In one embodiment the suitable replacement learners are disregarded from being selected if they are not available.
In one embodiment when a suitable replacement is selected, the profile and/or relationship of the corresponding subnetwork node to which it is added is recalculated. In a further aspect of the invention there is provided a method of forming a collaborative learning group comprising the steps of:
sending a call to join the learning group to a network of learners;
selecting a plurality of learners from the learners that respond and creating a subnetwork of nodes therefrom, each node having a profile defined by learning attributes; adding edges between pairs of nodes defined by learning relationships merging nodes with similar attributes and/or relationships therebetween to form sets of learners;
characterised in that the details of the non-selected users that respond are saved as potential replacements such that if a learner leaves the subnetwork, a replacement learner is selected from the potential replacements based on profile and/or relationship similarity between the replacement learner and the subnetwork nodes.
In one embodiment the nodes are merged if the similarities between the attributes and/or relationships are within predefined threshold ranges.
In one embodiment an iterative strategy is used to merge nodes until the set has enough learners to meet a predefined threshold. Thus the nodes with highest similarity are grouped first, and then successive strategies are used to add further nodes in order of decreasing similarity.
Brief Description of Drawings
It will be convenient to further describe the present invention with respect to the accompanying drawings that illustrate possible arrangements of the invention. Other arrangements of the invention are possible, and consequently the particularity of the accompanying drawings is not to be understood as superseding the generality of the preceding description of the invention. Figure 1 illustrates a schematic overview of the invention.
Figure 2 illustrates the process flow for managing the network.
Figure 3 illustrates a network built from the example in Table 1.
Figure 4 illustrates the process flow of the Learners Grouping component
Figure 5 illustrates the process flow of the Personalized Pairing of Nodes component
Figure 6 illustrates the process flow of the Select Personalized Pair of Nodes component
Figure 7 illustrates the process flow of the Index Selected Nodes component
Figure 8 illustrates the process flow of the Learner Replacement component
Detailed Description
With reference to Figure 1 , there is illustrated a schematic overview 2 of the invention in which learners are selected 4 from a network 6, grouped 8 into a set according to their profiles as defined by the strategies 10, wherein if a learner leaves the set, a suitable replacement learner 12 is automatically selected from the network.
Table 1 below provides an example of several profiles, with the following attributes:
· Proficiency: the competency of a learner with regard to a learning goal (e.g.
Novice, Qualified);
• Interaction history: past interactions of a learner with other learners, used to compute the 'Collaborative Willingness' of a learner to work with other learners, defined by a certain range (e.g. from 0 to 10, where 0 means no willingness and 10 means maximum willingness);
• Cognitive profile: includes learning styles (e.g. visual, logical) or multiple intelligence (e.g. interpersonal, intrapersonal) each sub-attribute having a defined range (e.g. from 0 to 100, where 0 means the learner is not of this attribute, 100 the learner is maximally of this attribute).
Figure imgf000007_0001
In the example, John has the following attributes:
Proficiency: proficient with regard to the input learning goal,
Interaction history: his previous interactions were not good with Jack, but good with Mary, and Cognitive profile: he is visual (80%), not very logical (20%), and Interpersonally neutral (50%).
Figure 2 depicts the process flow 100 for managing the network, in this case adding a new learner into the network.
1. System takes as input of learner L, and has access to the current network of learners
106.
2. Process starts by creating 102 a new node in the network which is linked to the learner L (already present in Learner Model 104),
3. Then extract 108 all the other nodes of the network, say Nl, ... ,Nk.
4. For each node 110 Ni, i = 1...k,
a. Compute 112 collaborative willingness of L and Ni, say CW(L,Ni), for example by:
1. by summing if applicable (i.e. both learners already interacted together) the part of their interaction history related to each other [e.g. {L = John} and {Ni = Jack}, their collaborative willingness is CW(L,Ni) = 2 + 5 = 7]
2. if 114 the collaborative willingness is below a threshold t, then subtract a value v [e.g. t = 10, thus CW(L,Ni) = CW(L,Ni) - 10 = -3 ] b. Add 116 a collaborative willingness edge between L and Ni with weightage calculated in previous step
c. Compute 118 cognitive profile similarity of L and Ni, say CPS(L,Ni), for example by:
1. Using the following formula: CP - CP ,}ι'
CPSij, Γ) = 100
[e.g. {L = John}, {Ni = Mary}, CPS(L,Ni) 83] d. Add 120 a cognitive profile similarity edge between L and Ni with weightage calculated in previous step
5. In the end the learner L is part of the network and linked to other learners by Collaborative Willingness (CW) and Cognitive Profile Similarity (CPS) edges.
Figure 3 depicts the network 122 built from the example in Table 1. Figure 4 depicts the process flow 200 of the Learners Grouping component, which allows learners to be grouped in a personalized way (including profiling of groups) for a specific learner goal, number of learners necessary to start and groups size:
1. System takes as input a number of learners n, a group size s and a learning goal.
2. Process starts by sending a call 202 to the network for anyone interested to join this collaborate e-learning session.
3. After a certain time 204 waiting for people, if there are not enough learners, the process goes back to Step 2, otherwise 206 it chooses 208 n learners out of all who accepted to join the collaborative e-learning session (e.g. randomly chosen or based on any parameters such as privileges etc.)
4. The other learners not chosen at Step 3 are saved 210 in Potential Replacements 212 for future use by the replacement process. 5. The sub-network composed of only the nodes corresponding to the learners chosen is then extracted 214 from the full network of learners 106.
6. Validity edges between nodes in the sub-network are then added 216 as follows:
For each pair of nodes (N1,N2), check the pair proficiency validity by using a predefined set of pair proficiency schemas 218, and add a validity edge between Nl and N2 if the pair proficiency is valid, [e.g. {pair proficiency schemas = {(Proficient, Beginner), (Expert, Beginner), (Expert, Qualified)} .
John and Jack are OK (Proficient with Beginner) but John and Mary are not OK (no schema for expert and proficient)]
7. If all groups are formed 220 (i.e. each node is composed of s learners - s being the input size for a group), then the process ends. Otherwise a personalized pairing 222 of nodes is done (see Figure 6)
Figure 5 depicts the process flow of the Personalized Pairing of Nodes component 222, which allows pairing of two nodes in the network by selecting the most personalized pair of nodes 302, profiling the two selected nodes and merging the two selected nodes in the sub-network. The process follows as below:
1. The Select Personalized Pair of Nodes component receives the network of learners and selects the best pair in the network according to Relationships Importance Strategies 224 (see Figure 6).
2. The two selected nodes are then profiled/indexed 304 using a series of indexing strategies 226 (see Figure 7). 3. Finally the two selected nodes are grouped/merged 306 in the sub-network 215 and indexed using the index of Step 2. The two nodes are merged together, adding a link to the pair's attributes previously set in Step 2, then collaborative willingness, cognitive profile similarity and validity edges are added between the node resulting from the merging and the other nodes in the network.
Figure 6 depicts the process flow of the Select Personalized Pair of Nodes component 302, which allows selecting the most personalized pair of nodes based on a series of strategies. The process follows as below:
1. The Select Next Strategy component 402 receives the series of ordered strategies and selects the first strategy which has not already been selected. Examples of strategies include:
{ Strategy 1: Must have a Validity edge and a positive CW edge, Then choose the pair with the highest positive CW edge}
{Strategy2: Must have a Validity edge and not have a negative CW edge, then choose the pair with the highest CPS edge}
{Strategy3: Must have positive CW edge, then choose highest CW edge} {Strategy4: Choose highest CPS edge}
2. If a strategy has been selected 404 (i.e. all the strategies have not been tested yet) then the selected strategy is applied 406 to the network 215. If there are any results 408, one is randomly chosen 410 (i.e, there can be several results after applying one strategy) and the process ends, otherwise repeat Step 1.
Otherwise 412 (i.e. no results from all the strategies) choose one of the nodes in the network not representing a formed group (i.e. not of the desired size yet) and containing the biggest number of learners (i.e. representing one of the biggest not formed group) and one of the nodes containing the smallest number of learners, and the process ends,
[e.g. with sub-network in Figure 3, Strategyl does not apply, but Strategy2 applies and gives as a result the pair (Mary, Jack)]
Figure 7 depicts the process flow of the Index Selected Nodes component 304, which allows indexing/profiling of two nodes selected in the network based on a series of grouping strategies. The process follows as below: 1. The proficiency 502 of the given pair of nodes is set according to Group Proficiency Strategies226a [e.g. {Strategyl : Expert+Novice = Proficient}, and the nodes are Mary and Jack, then the proficiency of the pair (Mary, Jack) is proficient] .
2. The interaction history (i.e. collaborative willingness 504) of the given pair of nodes is set according to Group Interaction History Strategies 226b [e.g. {Strategyl : average}, and the nodes are Mary and Jack. Then the interaction history of the pair (Mary, Jack) is John=6] .
3. The cognitive profile 506 of the given pair of nodes is set according to Group Cognitive Profile Strategies226c [e.g. {Strategyl : average}, and the nodes are Mary and Jack, then the cognitive profile of the pair (Mary, Jack) is (65,40,60)] .
Figure 8 depicts the process flow of the Learner Replacement component 600, which allows dynamically replacing a learner in a group. The process follows as below:
1. The system takes as input the learner leaving the group, say 1. 2. The process then updates 602 the group of 1 say g, by removing 1 from g and re- indexing the group's attributes (re-use previous process in Figure 5 - Index Select Groups - by just inversing the outcomes, because a member is deleted rather than being added)
3. The process then checks 604 if the potential replacements 212 previously saved are available, and filters out non-available replacements (e.g. not online at the moment, or in another collaborative e-learning session, etc)
4. The process then ranks 606 the available potential replacements based on the 'closeness' of the potential replacement with the group g by using a series of strategies (re-use previous process in Figure 6 - apply strategies to network - by just applying to the sub-sub-network composed of the group g and the potential replacement)
5. If no replacement found 608 yet:
a. If still potential replacements 610 not explored, pick one of the highest ranked potential replacements, say Γ
i. Call Γ for replacement 612
ii. After a certain time 614, if Γ accepted then mark replacement has found.
iii. Go to Step 5.
b. If no more potential replacements, call 616 all network 106 for replacement until a replacement is found 618
6. When replacement is found, say Γ
a. Add 620 Γ into the sub-network 215
b. Index/Profile 622 the pair Γ and g (same as process in Figure 5 - Index Selected Nodes) c. Group 624 the pair Γ and g (same than process in Figure 5 - Group Nodes)
Thus the collaborative learning groups formed are dynamic in that if a learner leaves a group, they are replaced with the most appropriate replacement learner identified thereby substantially maintaining the learning attributes of the group.
It will be appreciated by persons skilled in the art that the present invention may also include further additional modifications made to the system which does not affect the overall functioning of the system.

Claims

A collaborative learning system (2) for a network of learners (6) comprising: a plurality of nodes, each node having a profile and representing a learner or a group of learners selected from the network, forming a subnetwork;
edges between the nodes representing relationships between the nodes;
pairs of nodes being selectable such that if the respective profiles or relationships are similar enough the nodes can be merged to form one node with a single profile;
characterised in that learners not selected from the network are tracked such that if a learner leaves the subnetwork, a suitable replacement (12) is selected based on profile and relationship similarity between the replacement learner and the subnetwork nodes.
A collaborative learning system according to claim 1 wherein the profile includes attributes including one of more of proficiency, interaction history, and cognitive profile.
A collaborative learning system according to claim 1 wherein the relationships include one or more of validity of proficiencies, cognitive similarity, and collaborative willingness.
4. A collaborative learning system according claim 1 wherein the relationships are weighted. A collaborative learning system according to claim 1 wherein the learners that wish to join the subnetwork but are not selected therefor are tracked (210) as potential replacements.
A collaborative learning system according to claim 1 wherein when a suitable replacement is selected, the profile or relationship of the corresponding subnetwork node to which it is added is recalculated (622).
A method of forming a collaborative learning group comprising the steps of:
sending a call (202) to join the learning group to a network of learners;
selecting (208) a plurality of learners from the learners that respond and creating (214) a subnetwork of nodes therefrom, each node having a profile defined by learning attributes;
adding (216) edges between pairs of nodes defined by learning relationships merging nodes with similar attributes or relationships therebetween to form sets of learners;
characterised in that the details of the non-selected users that respond are saved (210) as potential replacements (212) such that if a learner leaves the subnetwork, a replacement learner is selected (620) from the potential replacements based on profile or relationship similarity between the replacement learner and the subnetwork nodes.
8. A method according to claim 7 wherein the nodes are merged (306) if the similarities between the attributes or relationships are within predefined threshold ranges. 9. A method according to claim 7 wherein an iterative strategy (302) is used to merge nodes until the set has enough learners to meet a predefined threshold.
PCT/MY2015/050105 2014-09-22 2015-09-18 Intelligent collaborative learning system WO2016048132A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
MYPI2014002707 2014-09-22
MYPI2014002707 2014-09-22

Publications (1)

Publication Number Publication Date
WO2016048132A1 true WO2016048132A1 (en) 2016-03-31

Family

ID=54844017

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/MY2015/050105 WO2016048132A1 (en) 2014-09-22 2015-09-18 Intelligent collaborative learning system

Country Status (1)

Country Link
WO (1) WO2016048132A1 (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050164154A1 (en) * 2004-01-23 2005-07-28 Geodesic Dynamics Demand initiated customized e-learning system
WO2013049907A1 (en) * 2010-10-07 2013-04-11 Clevru Corporation Method, system and computer program for providing an intelligent collaborative content infrastructure

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050164154A1 (en) * 2004-01-23 2005-07-28 Geodesic Dynamics Demand initiated customized e-learning system
WO2013049907A1 (en) * 2010-10-07 2013-04-11 Clevru Corporation Method, system and computer program for providing an intelligent collaborative content infrastructure

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"Mi-iLMS: Interactive Learning and Management", 22 July 2011 (2011-07-22), XP054975580, Retrieved from the Internet <URL:http://www.youtube.com/watch?v=l0UKZsgFvIg> [retrieved on 20141029] *

Similar Documents

Publication Publication Date Title
Naicker et al. Teachers’ reflections on distributive leadership in public primary schools in Soweto
CN108614865B (en) Personalized learning recommendation method based on deep reinforcement learning
Murawski Addressing diverse needs through
Jensen Poor students, rich teaching: Mindsets for change
CN107133894B (en) Online learning grouping method based on complex network theory
Wieber et al. Collective action control by goals and plans: applying a self-regulation perspective to group performance
Ma et al. Spreading teaching innovations in social networks: The bridging role of mentors
CN105427694B (en) Computer aided test method for intelligent teaching
Slim et al. An automated framework to recommend a suitable academic program, course and instructor
CN110246065A (en) A kind of grouping-hierarchy curricula-variable cource arrangement method
Horn et al. Recontextualizing practices: situative methods for studying the development of motivation, identity, and learning in and through multiple contexts over time
WO2016048132A1 (en) Intelligent collaborative learning system
Supianto et al. Designing scaffolding system in a problem-posing learning environment
CN110782961A (en) Intelligent hearing speech rehabilitation method and device, electronic equipment and medium
Little et al. Learning and Working: elements of the Diploma Disease thesis examined in England and Malaysia
Kober Intercultural professionalisation in adult education
Rush The myth of the eccentric genius: Some thoughts about political correctness and art education
CN113868515B (en) Self-adaptive learning path recommendation method based on ant colony algorithm
Bielefeldt Characteristics of engineering faculty engaged in the scholarship of teaching and learning
NWODOH Entrepreneurial and Functional Vocational Education as a Means of Achieving the National Objectives of Vision 20: 2020 in Nigeria
CN105070122A (en) Teaching object management method and teaching terminal
CN105869469A (en) Auxiliary system for computer-assisted instruction
Okubo et al. A friendship network model on Heider's balance theory considering Assertiveness and Empathy
Astutik Social and cultural agency in the practice of anti-intolerance and radicalism education in religion-based high schools
Russett et al. An Innovative Approach to College Student Drinking

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15808039

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15808039

Country of ref document: EP

Kind code of ref document: A1