Unified Semantic Network (USN)
The Unified Semantic Network (USN) is a unification of several semantic networks and lexical resources. It has been created using graph databases. The main motivation behind it, to enable relational inference by exploiting the potential knowledge that is yielded by unification of multiple resources.
We created an algorithm called discounted shortest path to infer affective ratings and prior sentiment polarities of words. We created two lexicons for prior sentiment polarities called
Our USN code will soon be available freely for download for research purposes.
If you are interested, please visit usn.cis.fiu.edu, and contact us at Ugan Yasavur and Christine Lisetti.
We created an algorithm called discounted shortest path to infer affective ratings and prior sentiment polarities of words. We created two lexicons for prior sentiment polarities called
- SentiLex for prior sentiment polarities
- AffectWords for discrete affective ratings of words.
Our USN code will soon be available freely for download for research purposes.
If you are interested, please visit usn.cis.fiu.edu, and contact us at Ugan Yasavur and Christine Lisetti.
