The meaning of a sentence can be represented by a graph in which the nodes represent concepts, entities or actions, and the edges represent relations between the edges, indicating who does what to whom. Nodes and edges are labelled, and every node is linked to a word or a phrase in the sentence. Meaning representation graphs are used in question answering systems and other natural language processing applications. This project implements software to analyze and visualize meaning representation graphs from a linguistic framework called Minimal Recursion Semantics. The tool includes functionality such as to (a) visualize the graphs in relation to the sentences they represent, (b) search for subgraph patterns over a set of graphs, (c) display differences and similarities between a pair of graphs, and (d) test whether a number of formal graph properties hold for a graph.
This project was built with a maven rapper (mvnw for linux, mvnw.cmd for windows). This means you can use your own maven installation, but you don't have to.
Pretty sure you will just be able to import this project into IntelliJ, VsCode, Eclipse and pretty much any other java IDE that supports maven projects.
mvnw acts as a maven rapper
If you're running from the commandline:
./mvnw compile - compiles the project
./mvnw install - compiles project and generates a jar file within the target directory.
./mvnw clean - cleans the project
Important to note, any additional dependencies/libraries we require should be placed within the project pom.xml file and maven will take care of it during compilation.
Simply compile and execute the program from your IDE of choice (or run the jar generated by ./mvnw install).
java -jar target/RepGraph.jar (must be run from main project directory)
This will launch an embedded tomcat server which you can access through http//localhost:8080 on any browser. Be patient as it might take a wee bit to load on your first access of a new build.