Building the Innovation Knowledge-base

The Innovation knowledge-base contains innovation related entities and relationships (concepts, material, people, and organizations), their descriptions, and evidences of their appearance in the data collection. We build the knowledge-base through a hybrid approach that combines automatic entities/relationship detection and collective knowledge accumulation through Social Information Processing (SIP) applications. Although the automatic computational software can easily identify possible entities and co-occurrences in large collections, it suffers several problems:

1. Automatically extracted data is highly noisy and inaccurate;
2. many existing entities and relationships may not be identified by the software;
3. the co-occurrence relationships are insufficient for researchers to understand the complex interactions among innovations.

To improve the quality of the extracted data, the STICK project addresses this issue by adopting a SIP mechanisms that supports collective knowledge accumulation and collaborative sensemaking - social cognitive processes by which people collectively gather, organize, and understand knowledge. With proper incentive mechanisms, SIP systems can build large-scale knowledge-base with high quality content in a sustainable manner.

Figure 3. Construction of the Innovation Knowledge-base