Subgraph Matching by Postponing Cartesian Products using NEO4J Graph DB and Protiens dataset.
Implemented the algorithm in paper Efficient Subgraph Matching by Postponing Cartesian Products by Fei Biy, Lijun Changy, Xuemin Liny, Lu Qinz, Wenjie Zhangy.
Designed and implemented query decomposition described in Section 3
Designed and implemented Compact Path-Index (CPI) using Neo4j along with root node selection described in Appendix A.6.
Using the ground truth in Proteins, decomposed the different query graphs, computed CPIs for the decompositions, and provided a Java program that automatically checks if all the expected solutions are contained in the computed CPIs
This is a JAVA Gradle project.
The source code is under the following path
./src/main/java
The configuration file to connect to database is located in ./src/main/resources folder
The default configuration is as follows
| Key | Value |
|---|---|
| neo4jdbpath | /pathtographdb/graph.db |
| proteinsfolderpath | /pathtotarget/target |
| proteinsquerypath | /pathtoquery/query |
| proteingroundtruthpath | /pathtogroundtruth/ground_truth |
The ./src/test/java contains test cases that uses JUNIT
The basic framework is in the file ./src/main/java/CPI
The Query decompositions are in the ./src/main/java/QueryGraph class.
The main Program.java has all the options to run the program.