Mechanism for monitoring collaboration
Methods, systems, and computer program products for monitoring collaboration between members of a group and providing feedback are described. A count of group members m of a group G, a count of group items i, and a characteristic of an action a performed by a member m on an item i, where a∈A, m∈M, and i∈I, are obtained. A member-item score is computed for a time quantum T q , where the time quantum T q is a fixed-length amount of time with an arbitrary start point. A collaboration score Q G for the group G is computed, where Q G =Σ ∀m∈m Σ ∀i∈I Q m,i .
1. A computerized method for administering social software, the method comprising:
accessing, using one or more hardware processors, a database including a plurality of groups of members, each group of members comprising members that are users of the social software, each group of members having a plurality of data items associated with the group, each data item corresponding to content posted by a corresponding member of the group via the social software, wherein a count of members of the group and a count of data items of the group is obtained for a time quantum;
for each interaction between a specified member of the group and a specified data item associated with the group, updating the database to include a member-item count for the corresponding member of the group;
deriving, using one or more hardware processors, a member collaboration score for each member in the group, the member collaboration score indicating a level of interaction through the social software between the member and the plurality of data items associated with the group, the member collaboration score being determined from the member-item counts corresponding to the member;
deriving, using one or more hardware processors, a group collaboration score for each group, the group collaboration score for a specified group indicating a level of interaction through the social software for members of the specified group, the group collaboration score being determined from the member collaboration score for each member in the group; and
serving, using one or more hardware processors, a user interface to an administrator user, the user interface comprising an indication of a group collaboration score for a first group of the plurality of groups and a plurality of member scores for members of the first group.
2. The computerized method of claim 1 , wherein the member-item count is defined by Q m,i =V({right arrow over (A)} m,i )·{right arrow over (C)} G for a time quantum T q , where the time quantum T q is a fixed-length amount of time with an arbitrary start point, an action vector {right arrow over (A)} m,i is defined as a vector comprising an action component for each type of action a that a member m can perform utilizing a data item i, V is a valuation function which values each action component in the action vector {right arrow over (A)} m,i and generates a valued action vector V({right arrow over (A)} m,i ).
3. The computerized method of claim 2 , wherein the action vector {right arrow over (A)} m,i and the valued action vector V({right arrow over (A)} m,i ) have a different number of dimensions.
4. The computerized method of claim 1 , wherein the member collaboration score denotes how much a member m interacts with all data items I in the group, and is defined by Q m =Σ ∀i∈I Q m,i .
5. The computerized method of claim 2 , wherein the valuation function satisfies the following conditions:
V
a
(
0
)
=
0
lim
n
→
∞
V
a
(
n
)
=
K
va
where V a (n) represents a value of the member performing an action a a total of n times during the time quantum T q , and K va is a constant signifying a maximum valuation for the action a during time quantum T q .
6. The computerized method of claim 2 , wherein each component of the coefficient vector {right arrow over (C)} G is defined as
C
a
=
K
ca
M
×
I
NZ
,
where |M| is the count of members m in the group, |I NZ | is a total count of data items i in the group which have a non-zero action vector, and I NZ is a subset of data items I of the group where at least one member m has performed at least one action a on the data item i during the time period T q .
7. The computerized method of claim 1 , further comprising:
computing a frequency distribution function, the frequency distribution function being a histogram of a count of times a defined type of action is performed by one or more members of one or more groups;
converting the frequency distribution function into a probability distribution function by dividing each entry in the histogram by a total number of occurrences of the defined type of action;
converting the probability distribution function into a cumulative distribution function; and
using the cumulative distribution function as a valuation function for the defined type of action.
8. The computerized method of claim 6 , further comprising determining the valuation function and the coefficient vector using a machine learning technique.
9. The computerized method of claim 4 , wherein the group collaboration score of the group is defined by Q G =Σ ∀m∈M Σ ∀i∈I Q m,i , where M represents the set of members in the group, I represents the set of data items in the group, and Q m,i represents the member-item count for member m and data item i.
10. The computerized method of claim 1 , wherein a member is any entity that is capable of performing the content posting.
11. The computerized method of claim 1 , further comprising:
for each interaction between the specified member of the group and the specified data item associated with the group, updating the database to include an member-item count for the corresponding data item associated with the group; and
deriving, using one or more hardware processors, an item collaboration score for each data item in the group, the item collaboration score being determined from the member-item counts of the corresponding data item, the group collaboration score for each group also determined from the item collaboration score.
12. The computerized method of claim 11 , wherein the item collaboration score denotes how much an item i is being acted upon by all members M in the group, and is defined by Q i =Σ ∀m∈M Q m,i .
13. The computerized method of claim 1 , wherein the member-item count is a count of interactions between the specified member and the specified data item.
14. An apparatus for administering social software, the apparatus comprising:
one or more hardware processors;
memory to store instructions that, when executed by the one or more hardware processors perform operations comprising:
accessing, using one or more hardware processors, a database including a plurality of groups of members, each group of members comprising members that are users of the social software, each group of members having a plurality of data items associated with the group, each data item corresponding to content posted by a corresponding member of the group via the social software wherein a count of members of the group and a count of data items of the group is obtained for a time quantum;
for each interaction between a specified member of the group and a specified data item associated with the group, updating the database to include a member-item count for the corresponding member of the group;
deriving, using one or more hardware processors, a member collaboration score for each member in the group, the member collaboration score indicating a level of interaction through the social software between the member and the plurality of data items associated with the group, the member collaboration score being determined from the member-item counts corresponding to the member;
deriving, using one or more hardware processors, a group collaboration score for each group, the group collaboration score for a specified group indicating a level of interaction through the social software for members of the specified group, the group collaboration score being determined from the member collaboration score for each member in the group; and
serving, using one or more hardware processors, a user interface to an administrator user, the user interface comprising an indication of a group collaboration score for a first group of the plurality of groups and a plurality of member scores for members of the first group.
15. The apparatus of claim 14 , wherein the member-item count is defined by Q m,i =V({right arrow over (A)} m,i )·{right arrow over (C)} G for a time quantum T q , where the time quantum T q is a fixed-length amount of time with an arbitrary start point an action vector {right arrow over (A)} m,i is defined as a vector comprising an action component for each type of action a that a member m can perform utilizing a data item i, V is a valuation function which values each action component in the action vector {right arrow over (A)} m,i and generates a valued action vector V({right arrow over (A)} m,i ).
16. The apparatus of claim 14 , wherein the member collaboration score denotes how much a member m interacts with all data items I in the group, and is defined by Q m =Σ ∀i∈I Q m,i .
17. The apparatus of claim 14 , the operations further comprising:
for each interaction between the specified member of the group and the specified data item associated with the group, updating the database to include an member-item count for the corresponding data item associated with the group; and
deriving, using one or more hardware processors, an item collaboration score for each data item in the group, the item collaboration score being determined from the member-item counts of the corresponding data item, the group collaboration score for each group also determined from the item collaboration score.
18. The apparatus of claim 15 , wherein each component of the coefficient vector {right arrow over (C)} G is defined as
C
a
=
K
ca
M
×
I
NZ
,
where |M| is the count of members m in the group, ├I NZ | is a total count of data items i in the group which have a non-zero action vector, and I NZ is a subset of data items I of the group where at least one member m has performed at least one action a on the data item i during the time period T q .
19. The apparatus of claim 17 , wherein the item collaboration score denotes how much an item i is being acted upon by all members M in the group, and is defined by Q i =Σ ∀m∈M Q m,i .