High level automaton library for information extraction. Usage:
- Named Entity Recognition(NER) patterns
- Optimal match finding (for ambiguous grammars)
- Relation and fact extraction
- Structure parsing (like document structure, sententce parsing)
- Search problem solving (like Suduku)
- all standard reg exp operators: +, {n,n}, {n,}, ....
- much more: java predicates on groups, reluctant, greedy, cut operator, inner match...
- structure of match -it is possible to build syntax tree based on match
- weighted regexps allow you to encode preference of different graph path with your scoring functions, heuristic search will handle the rest.
- easy extendable framework model. You can easily write new matchers and use them with existed.
- dynamic structure of graph (allow you to use gexp to solve search problems like suduko, 8-qeen problems)
- fast - it works faster then Jape transducer (gate.ac.uk) closest project to this one
- scopes for variables: all you predicates can set/get variables from current scope/context to do their job
- easy embeddable - few line of java code and you can use power of graph-expression in you project.
Read more on wiki https://github.com/yurkor/graph-expression/blob/wiki/ProjectHome.md