# US 2012/0084236 A1 · Recording medium storing decision tree generating program, decision tree generation method and decision tree generating apparatus

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2012/0084236 A1
- Document ID: US20120084236A1
- Status: Published application
- Kind: A1
- Application number: 13/200,354
- Filed: 2011-09-23
- Published: 2012-04-05
- Earliest claimed: 2009-03-30
- Assignees: Fujitsu Limited
- Inventors: Kuniaki Shimada; Yasuhide Matsumoto; Yukihiro Watanabe; Yuji Wada; Masazumi Matsubara; Kenji Morimoto; Hiroshi Otsuka
- Web page: https://patentgrep.com/patent/US20120084236A1

## CPC

- G06N 5/02: Knowledge representation; Symbolic representation
- G06N 20/00: Machine learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A constraint condition DB storing a constraint condition that stipulates a structure of a decision tree is referenced, and a decision tree is generated from a case set where values of a plurality of attributes and a conclusion are associated with one another so that the structure of the decision tree, which is stipulated by the constraint condition, is satisfied. Accordingly, for example, even if a new case is added to the case set, a basic structure of the decision tree is succeeded by the constraint condition, thereby avoiding a situation where an operator needs to significantly modify the decision tree. Therefore, operations of modifying the decision tree by the operator can be reduced.

## Claims

### Claim 1

A non-transitory computer readable medium storing decision tree generating program for causing a computer having an external storage device storing a constraint condition that stipulates a structure of a decision tree to execute a procedure, the procedure comprising:
reading a case set where values of a plurality of attributes and a conclusion are associated with one another;
reading the constraint condition from the external storage device; and
generating a decision tree from the case set read so that the structure of the decision tree, which is stipulated by the constraint condition, is satisfied.

### Claim 2 (depends on 1)

The medium according to claim 1, further comprising:
editing the decision tree generated;
generating the constraint condition from a modification history of the decision tree edited; and
storing the generated constraint condition in the external storage device.

### Claim 3 (depends on 2)

The medium according to claim 2, wherein
the constraint condition is composed of at least a first constraint condition stipulating an ignored attribute, a second constraint condition stipulating that an attribute is subordinate to other attributes, and a third constraint condition stipulating that an attribute is superordinate to other attributes.

### Claim 4 (depends on 3)

The medium according to claim 3, wherein
the generating the constraint condition comprises generating the first constraint condition that ignores an attribute if there is the attribute deleted from the decision tree, generating the second constraint condition stipulating that an attribute is subordinate to other attributes if there is the attribute moved to a position far from a root of the decision tree, or generating the third constraint condition stipulating that an attribute is superordinate to other attributes if there is the attribute added to the decision tree or if there is the attribute moved to a position close to the root of the decision tree.

### Claim 5 (depends on 3)

The medium according to claim 3, wherein
the other attributes in the second constraint condition and the third constraint condition are respectively represented with a set.

### Claim 6 (depends on 1)

The medium according to claim 1, further comprising:
extracting a difference between a latest decision tree generated and a decision tree immediately before the latest decision tree; and
displaying the difference on a display device.

### Claim 7

A decision tree generating method for causing a computer having an external storage device storing a constraint condition that stipulates a structure of a decision tree to execute, the method comprising:
reading a case set where values of a plurality of attributes and a conclusion are associated with one another;
reading the constraint condition from the external storage device; and
generating a decision tree from the read case set so that the structure of the decision tree, which is stipulated by the constraint condition, is satisfied.

### Claim 8 (depends on 7)

The decision tree generating method according to claim 7, further comprising:
editing the generated decision tree according to an instruction of an operator;
generating a constraint condition from a modification history of the decision tree edited; and
storing the generated constraint condition in the external storage device.

### Claim 9

A decision tree generating apparatus comprising:
an external storage device storing a constraint condition that stipulates a structure of a decision tree;
case inputting means for reading a case set where values of a plurality of attributes and a conclusion are associated with one another;
constraint condition reading means for reading the constraint condition from the external storage device; and
decision tree generating means for generating a decision tree from the case set read by the case inputting means so that the structure of the decision tree, which is stipulated by the constraint condition, is satisfied.

### Claim 10 (depends on 9)

The decision tree generating apparatus according to claim 9, further comprising:
decision tree editing means for editing the decision tree generated by the decision tree generating means according to an instruction of an operator;
constraint condition generating means for generating a constraint condition from a modification history of the decision tree edited by the decision tree editing means; and
constraint condition storing means for storing the generated constraint condition in the external storage device.
