Information processing apparatus, information processing method, and information processing computer program product
An information processing apparatus 10 includes a processing unit 20. The processing unit 20 calculates, for each pair of pieces of data defined in a graph representing relational data 16 A indicating a relationship between a plurality of pieces of data in a graph structure, a degree of relevance between a target category to be analyzed and another category other than the target category on the basis of a score calculated according to a degree of relationship between a pair of pieces of data constituting the pair and an evaluation index for evaluating the degree of relevance between the categories.
1 . An information processing apparatus comprising:
a hardware processor configured to:
acquire relational data indicating relationships among a plurality of pieces of data, each piece of data being associated with a category;
generate, from the relational data, a graph represented in a graph structure and defining pairs of pieces of data based on the relationships, the generating comprising, for each pair of a target category to be analyzed and an other category different from the target category, generating a corresponding graph that represents pieces of data belonging to the target category and the other category and relationships therebetween;
calculate, for each pair of pieces of data defined in the corresponding graph, a score according to a degree of relationship between the pieces of data of the pair;
register the score in a score database in association with the pair of pieces of data;
calculate, based on (i) the score registered for pairs of pieces of data and (ii) an evaluation index for evaluating a degree of relevance between categories, a degree of relevance between the target category to be analyzed and the other category;
generate a target graph by specifying, using the calculated degree of relevance, at least one specified other category having a degree of relevance satisfying a criterion, and linking data belonging to the specified other category and data belonging to the target category according to the degree of relevance to form the target graph represented in a graph structure;
calculate, using the target graph, a representation vector for each of a plurality of pieces of data included in the target graph;
extract, based on features represented by the representation vectors, a tendency pattern for the plurality of pieces of data; and
control output to output at least one of (i) information indicating the degree of relevance, (ii) the target graph, (iii) the representation vectors, or (iv) information indicating the tendency pattern.
2 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
generate the graph from the relational data for each pair of the target category and the other category on a one-to-one basis.
3 . The information processing apparatus according to claim 1 , wherein the graph is represented by the graph structure linked according to the relationship in which a plurality of pieces of data included in the relational data is represented by the relational data.
4 . The information processing apparatus according to claim 3 , wherein the hardware processor is configured to
generate the graph with an adjusted weight of a linkage between pieces of data.
5 . The information processing apparatus according to claim 3 , wherein the hardware processor is configured to
generate the graph with a weight of a linkage between pieces of data adjusted to a specified value.
6 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate, as the score, a number of paths for the each pair, the number of paths being a number of routes reaching from one data to another data, both constituting the pair with a predetermined number of hops or less.
7 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate, as the score, a weighted average value of a number of paths for the each pair, the number of paths being a number of routes reaching from one data to another data, both constituting the pair with a predetermined number of hops or less.
8 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate, as the score, a value represented by an inner product of a first matrix and a transposed matrix of the first matrix, the first matrix being obtained by multiplying an adjacent matrix of the graph by a predetermined matrix of dimensions according to a number of dimensions of data of the graph.
9 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate, as the degree of relevance, a value obtained by combining a plurality of the evaluation indexes.
10 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate, as the degree of relevance, a value of the evaluation index satisfying a preset constraint.
11 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate the degree of relevance on a basis of the evaluation index and the score calculated for each of some pairs sampled from a plurality of pairs.
12 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
calculate the degree of relevance on a basis of the evaluation index for which an input is received.
13 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
output the degree of relevance between the target category and the another category.
14 . The information processing apparatus according to claim 1 , wherein the hardware processor is configured to
specify the other category similar to the target category using the degree of relevance, and generate a target graph in which data belonging to the specified other category and data belonging to the target category are linked according to the degree of relevance and which is represented in a graph structure.
15 . The information processing apparatus according to claim 14 , wherein the hardware processor is configured to
calculate a representation vector of each of a plurality of pieces of data included in the target graph using the target graph.
16 . The information processing apparatus according to claim 15 , wherein the hardware processor is configured to
calculate a representation vector of each piece of data included in the target graph so as to hold a graph structure of the target graph.
17 . The information processing apparatus according to claim 15 , wherein the hardware processor is configured to
extract a tendency pattern of a plurality of pieces of data represented by the representation vector on a basis of a feature represented by the representation vector.
18 . The information processing apparatus according to claim 15 , wherein the hardware processor is configured to
extract, as recommendation data for each of the plurality of pieces of data, another data group having a degree of similarity equal to or greater than a predetermined value with respect to each of a plurality of pieces of data represented by the representation vector on a basis of a degree of similarity of the plurality of pieces of data represented by the representation vector.
19 . An information processing method executed by an information processing apparatus, the method comprising:
acquiring relational data indicating relationships among a plurality of pieces of data, each piece of data being associated with a category;
generating, from the relational data, a graph represented in a graph structure and defining pairs of pieces of data based on the relationships, the generating comprising, for each pair of a target category to be analyzed and an other category different from the target category, generating a corresponding graph that represents pieces of data belonging to the target category and the other category and relationships therebetween;
calculating, for each pair of pieces of data defined in the corresponding graph, a score according to a degree of relationship between the pieces of data of the pair;
registering the score in a score database in association with the pair of pieces of data;
calculating, based on (i) the score registered for pairs of pieces of data and (ii) an evaluation index for evaluating a degree of relevance between categories, a degree of relevance between the target category to be analyzed and the other category;
generating a target graph by specifying, using the calculated degree of relevance, at least one specified other category having a degree of relevance satisfying a criterion, and linking data belonging to the specified other category and data belonging to the target category according to the degree of relevance to form the target graph represented in a graph structure;
calculating, using the target graph, a representation vector for each of a plurality of pieces of data included in the target graph;
extracting, based on features represented by the representation vectors, a tendency pattern for the plurality of pieces of data; and
controlling output to output at least one of (i) information indicating the degree of relevance, (ii) the target graph, (iii) the representation vectors, or (iv) information indicating the tendency pattern.
20 . An information processing computer program product having a non-transitory computer readable medium including programmed instructions stored thereon, wherein the instructions, when executed by a computer, cause the computer to execute:
acquiring relational data indicating relationships among a plurality of pieces of data, each piece of data being associated with a category;
generating, from the relational data, a graph represented in a graph structure and defining pairs of pieces of data based on the relationships, the generating comprising, for each pair of a target category to be analyzed and an other category different from the target category, generating a corresponding graph that represents pieces of data belonging to the target category and the other category and relationships therebetween;
calculating, for each pair of pieces of data defined in the corresponding graph, a score according to a degree of relationship between the pieces of data of the pair;
registering the score in a score database in association with the pair of pieces of data;
calculating, based on (i) the score registered for pairs of pieces of data and (ii) an evaluation index for evaluating a degree of relevance between categories, a degree of relevance between the target category to be analyzed and the other category;
generating a target graph by specifying, using the calculated degree of relevance, at least one specified other category having a degree of relevance satisfying a criterion, and linking data belonging to the specified other category and data belonging to the target category according to the degree of relevance to form the target graph represented in a graph structure;
calculating, using the target graph, a representation vector for each of a plurality of pieces of data included in the target graph;
extracting, based on features represented by the representation vectors, a tendency pattern for the plurality of pieces of data; and
controlling output to output at least one of (i) information indicating the degree of relevance, (ii) the target graph, (iii) the representation vectors, or (iv) information indicating the tendency pattern.