Conclusion

The principal purpose of this initial examination of patent citation networks is to determine if the approach has any apparent merit. Although what is presented here is in no way definitive, appearances of network analysis can provide new

Fig. 10.11. In-citations for Patent 4740470.

10 100 Citations made (log)

0.001

0.0001

1e-05

0.001

0.0001

1e-05

10 100 Citations received (log)

10 100 Citations made (log)

10 100 Citations received (log)

Fig. 10.12. Log-log plots of out-citations (L) and in-citations (R). CDF - cumulative distribution functions, (1-CDF) is therefore the complement to one of the cumulative distribution functions, in other words CDF = P [X<=x] therefore 1-CDF = P[X>x]; L - left panel and R - right panel.

1000

1000

insights, as well as help verify prior findings. Part of the additional explanatory power is derived from using indirect as well as direct citations. We believe that citation network analysis is an effective means of identifying pioneering patents. By extension, it may be possible to identify early in the process the trajectory of an individual patent, as well as specify technology cycles.

At the aggregate level, we see the potential for identifying key sectors for overall technological growth (here, tentatively, chemistry) as well as reviewing the basis for the current patent classification system. Finally, the analysis yields particular patterns, which at this stage are wholly unexplainable, but at least suggest unanticipated relationships and similarities among patents.

Was this article helpful?

0 0

Post a comment