Title :
The Knowledge-Remixing Bottleneck
Author_Institution :
Network Inst., VU Univ., Amsterdam, Netherlands
Abstract :
The Web has enabled us to construct massive knowledge bases that enable both intelligent applications and complex analyses. The construction of these knowledge bases invariably involves combining multiple sources using a stack of human and automated decisions, which leads to an opaque process that hampers both transparency and repurposing. The intricacies of this problem are concretely formulated, along with potential research directions for addressing it.
Keywords :
Internet; Web; complex analysis; intelligent applications; knowledge base; knowledge-remixing bottleneck; Data mining; Electronic mail; Electronic publishing; Encyclopedias; Internet; Knowledge based systems; Web science; data science; intelligent systems; knowledge acquisition systems; knowledge bases; natural language processing; remix;
Journal_Title :
Intelligent Systems, IEEE
DOI :
10.1109/MIS.2013.138