The power to unify imagination and reason is a specific capacity we humans share. It underlies all our human progression. But what if the underlying principles of our imagination and the ability to generate ideas is given to artificial intelligent systems? With the development of Yooray we dealt with the patterns, structures that lead to the split second ideas occur.
Writing a theoretical thesis -a retrospective simultaneous analysis of iconic ideas in art, science, physics, politics and philosophy- we noticed that existing search systems do not easily suggest matters of an associative kind.
- The system had to find associative information, it had to comprehend the meaning of words within a given language and handle vernacular jargon and spelling mistakes;
- The whole system should learn continuously in order to find associative information;
- The system should be able to quickly and easily exploit matters of an associative kind;
- The system should give the user the ability to operate the information on the level of association.
- If the user makes any changes within the front-end, (deletes certain parts of the query result, adds new questions, splits questions or combines new and old questions) the back-end is able to find new information on the level of meaning.
- The back-end should be able to work beyond the first data set.
- More users should be able to operate Yooray without any interference between users.
- The front- and back-end, should transmit all necessary changes within the system to one another.
Investigation of structures that lead to the moment ideas occur. We noticed that existing search systems do not easily suggest matters of an associative kind.
We were inspired by how the neural system works and used this knowledge to develop software that builds an adaptable relational network (using nodes and connections).
Yooray has been tested and showed positive results. The user can exploit connections between totally different fields or topics, just by combining, deleting or adding new (multiple) questions.
To enable this we studied how the “meaning” of words -as a most important part of human language- could be the basis of our Yooray’s structure.