To identify prior arts (patent/non-patent citations), various paid and/or unpaid databases can be used. Few of those popular databases are: Orbit, Derwent, XLSCOUT, Patsnap, Lens.org, Google patents, Google, IEEE, Google scholar, IETF standards, 3GPP standards, ETSI, Espacenet, J-platpat, KIPRIS, CNIPA etc.
Depending on the databases, there are numerous methods to search for prior arts. One method would include an automated search (for example in XLSCOUT – AI based module) where either the inventive paragraph or the drafted claim elements/features are entered, and an artificial intelligent module takes over from there to identify all the related/relevant/closest possible prior arts.
Next, we can perform manual searches by making search strings/queries/strategies using all the important/related key words/phrases of the invention. These can further help us in identifying major assignees/inventors and/or important patent classifications related to the invention which can be used in forming further search strategies.
We can also look for more related prior arts present/cited on an identified relevant prior art. Further, all these databases mostly have their own syntax/methods for making search strategies.
Moreover, if client has already some patents/products/papers published on the same technology, then those references/citations can be identified and can help in steering the search accordingly, to find the closest possible prior art.