Analysis

The research status of the NSE field can be assessed using three kind of analysis:

1) Bibliographic analysis
      Statistics for patents and NSF grants are presented in tables and graphs on country, assignee, technology field, etc. The system also enables the user to identify the key countries, assignees, and technology fields in different time periods by using the number of cites indicator.

2) Citation analysis
      The citation networks are presented for: countries, institutions and technology fields. The top 100 links of each network (according to the number of citations between the nodes) are visualized using an open source graph drawing software, Graphviz, provided by AT&T Labs (Gansner and North, 2000). The NanoMapper system enables users to study the citation networks over time and study knowledge diffusion between the analytical units.
      In the networks, the direction of the links represents the direction of the knowledge flow. For example, a link from “Country A” to “Country B” means that country A’s patents had cited country B’s patents; the number beside the link is the total number of these citations.

3) Topic analysis
      The content map tool developed by the Arizona Artificial Intelligence Lab was used to visualize the topics in nanotechnology patents/grants.
      In the content map toolkit, technology topics in the documents are extracted using a Natural Language Processing tool, the Arizona Noun Phraser. The technology topics are organized by the multi-level self-organization map algorithm (Chen et al., 1996; Ong et al., 2005), which calculates the topic similarities according to the co-occurrence patterns of key phrases in document titles and abstracts.
      On the content map interface, the topics are listed in a folder tree, and positioned geographically on a graph according to their similarity. Each node in the tree, corresponding to a region in the map, is a topic (keyword) identified from the document. Conceptually closer technology topics will be positioned closer geographically.
      Each time interval's content map was compared with the previous time interval's content map to visualize changes of topic areas. For each topic area, a growth rate was computed as the ratio between the number of documents in the current time period and that of the previous time period. A baseline growth rate was computed as the ratio between the total number of documents in the current time period and that of the previous time period. A topic region with similar growth rate to the base growth rate was assigned a green color. The topic region with higher (lower) growth rate was assigned a warmer (colder) color. If the topic was brand new, a red color was assigned to the region.