Information processing apparatus and non-transitory computer readable medium

1. An information processing apparatus comprising:

a precessor, configured to:

acquire information including information regarding a plurality of nodes and information regarding a plurality of links connecting the plurality of nodes and that acquires constraint information regarding node pairs included in the plurality of nodes, the constraint information including a positive constraint and a negative constraint;

calculate, for each of a plurality of clusters, a classification proportion into which the plurality of nodes are classified and that calculates a degree of importance of each of the plurality of clusters, the classification proportion representing a proportion in which each of the plurality of nodes is classified as one of the plurality of clusters; and

generate a probability model for performing probabilistic clustering on the plurality of nodes, the probability model being generated by using at least each of the information regarding the links, the constraint information, the classification proportion, and the degree of importance.

2. The information processing apparatus according to claim 1 ,

wherein the processor further acquires granularity and a weighting variable for determining weighting of the constraint information, the granularity being a variable for determining a size of each of the plurality of clusters into which the plurality of nodes are classified, and

wherein the processor generates the probability model by using the information regarding the links, the constraint information, the classification proportion, the degree of importance, the granularity, and the weighting variable.

3. The information processing apparatus according to claim 1 ,

wherein the probability model includes a constraint term expressed by using the constraint information, and

wherein the constraint term is obtained by multiplying a distance between degrees of belonging by values indicated by the constraint information, the degrees of belonging respectively corresponding to a distribution of a probability at which a first node of each node pair belongs to each cluster and a distribution of a probability at which a second node of the node pair belongs to each cluster.

4. The information processing apparatus according to claim 3 ,

wherein the constraint information is expressed as a constraint matrix with n×m for the node pairs, and

wherein in a case where C denotes the constraint matrix and c nm denotes a value of the constraint matrix C, the constraint matrix C represents c nm =c mn , the positive constraint for c nm >0, no constraint for c nm =0, and the negative constraint for c nm <0.

5. The information processing apparatus according to claim 4 , wherein the processor is further configured to:

derive the classification proportion in maximizing the probability model, the classification proportion being derived by using a predetermined non-linear programming method.

6. The information processing apparatus according to claim 3 ,

wherein the constraint information is expressed as a constraint matrix with n×m for the node pairs, and

wherein in a case where C denotes the constraint matrix and c nm denotes a value of the constraint matrix C, the constraint matrix C represents C nm =C mn , Σ n c nm =0, the positive constraint for c nm >0, no constraint for c nm =0, and the negative constraint for c nm <0.

7. The information processing apparatus according to claim 3 ,

wherein the constraint information is expressed as a constraint matrix with n×m for the node pairs, and

wherein in a case where C denotes the constraint matrix and c nm denotes a value of the constraint matrix C, the value c nm located on a diagonal of the constraint matrix C indicates whether a specific node is likely to isolate from neighboring nodes on a basis of a relationship between the specific node and the neighboring nodes.

8. The information processing apparatus according to claim 7 ,

wherein a positive value c nm located on the diagonal of the constraint matrix C indicates that the specific node is likely to isolate from the neighboring nodes, and

wherein a negative value c nm located on the diagonal of the constraint matrix C indicates that the specific node is less likely to isolate from the neighboring nodes.

9. The information processing apparatus according to claim 6 , wherein the processor is further configured to:

analytically derive the classification proportion in maximizing the probability model, the classification proportion being derived by using the value c nm located on a diagonal of the constraint matrix C.

10. The information processing apparatus according to claim 6 ,

wherein each of the node pairs includes a node with a known constraint and a node with an unknown constraint, the processor is further configured to:

predict a value of a constraint between the node with the known constraint and the node with the unknown constraint in the constraint matrix C from a known value in the constraint matrix C.

11. A non-transitory computer readable medium storing a program causing a computer to execute a process comprising:

acquiring information including information regarding a plurality of nodes and information regarding a plurality of links connecting the plurality of nodes and acquiring constraint information regarding node pairs included in the plurality of nodes, the constraint information including a positive constraint and a negative constraint;

calculating, for each of a plurality of clusters, a classification proportion into which the plurality of nodes are classified and calculating a degree of importance of each of the plurality of clusters, the classification proportion representing a proportion in which each of the plurality of nodes is classified as one of the plurality of clusters; and

generating a probability model for performing probabilistic clustering on the plurality of nodes, the probability model being generated by using at least each of the information regarding the links, the constraint information, the classification proportion, and the degree of importance.

12. An information processing apparatus comprising:

means for acquiring information including information regarding a plurality of nodes and information regarding a plurality of links connecting the plurality of nodes and acquiring constraint information regarding node pairs included in the plurality of nodes, the constraint information including a positive constraint and a negative constraint;

means for calculating, for each of a plurality of clusters, a classification proportion into which the plurality of nodes are classified and calculating a degree of importance of each of the plurality of clusters, the classification proportion representing a proportion in which each of the plurality of nodes is classified as one of the plurality of clusters; and

means for generating a probability model for performing probabilistic clustering on the plurality of nodes, the probability model being generated by using at least each of the information regarding the links, the constraint information, the classification proportion, and the degree of importance.