Hi..
Idea of Our project about using JADE in image annotation by analyzing the text will be entered by the user to describe the choosen image in a document and extract the stem word of the text to compare it with my knowledge(we use part of knowledge not all the language) such that we use 20 class to define knowledge in each class like this:
After user write text about any image for example text of beach:
**A beach is a geological land form along the shoreline of a body of water. It usually consists of loose particles which are often composed of rock, such as sand, gravel, shingle, pebbles, or cobble. The particles of which the beach is composed can sometimes instead have biological origins, such as shell fragments or coralline algae fragments.**
1- analyzed text..
2- extracted stem word from text like this..
String[]beachroot ={"beach","geological","land","shoreline","body","water","usuall","consists","loose","particles","composed","rock","sand","gravel","shingle","shingle",
"cobble","biological","origins","shell","fragment","coralline","algae"}
3-comparing stem word with element of predict such that I put score for each element in the predict ,if stem word match with element of predict take in score of element ,if not match we using word net to get synonyms and coordinates for each element in the predict since may be deliver from it to matching with root:
private String predicates []= {"sun","sand","shell","tree","ship","sky","lake","sea","palm"};
if total scoring for example >40 we will get the keyword of this predict..
but I want to use ontology of JADE if its benefit for us since I need to define 20 object from classes to compare predict of its with stem word of text and choosen the keyword of high score..plz any one expert in this domain of ontology help us quikly since no long time remaining to deliver this project..