|
|
Author |
Topic: do-it-ypurself patent search methods (Read 1656 times) |
|
Sue
Guest
|
 |
do-it-ypurself patent search methods
« on: Oct 19th, 2005, 10:22am » |
 Quote  Modify
 Remove
|
I was hoping for some advice on the best methods for patent searches... Here is my current method, suing the USPTO website:
1. Use different combinations of key words to search in "all fields". All results are then pasted into an excel spreadsheet.
2. review each patent found, and if relevant, document the class and subclass number. Irrelevant patents are deleted from the spreadsheet.
3. search each subclass. paste allresults into the spreadsheet. review each patent for relevance, delete the irrelevant ones.
4. review the referenced patents for each relevant patent. pase into spread sheet, review each one, and delete the irrelevant ones.
5. The spreadsheet can then be 'sorted' by patent number, so that redundant results (e.g. the same relevant patent came up with keyword search and with subclass search) can be eliminated.
The result is a spreadsheet with hypertext links to all relevant patents.
Is this a decent patent search?
Am I overlooking a technique?
Should I use a different search engine instaed of teh USPTO?
Is there a quicker way to do this?
|
|
 IP Logged |
|
|
|
JimIvey
Moderator
Senior Member
 
Posts: 2584
|
|
Re: do-it-ypurself patent search methods
« Reply #1 on: Oct 19th, 2005, 11:35am » |
Quote Modify
|
I'm no searching expert, but I'll offer a few observations.
I'm not sure why you're using a spreadsheet. Sorting can be done in any of a number of contexts, including using the *nix command, "sort" -- available to Windows users through cygwin. Most word processing programs also include a sort function -- especially for tables. However, there's nothing wrong with using a spreadsheet -- it just seemed to stick out as an anomaly.
The most significant thing that jumps out at me is the use of patents (and particularly US patents) exclusively. If you're doing a prior art search, all public documents, public uses, and offers for sale can be prior art. If you're only interested in US patents, e.g., for a quick assessment of infringement liability risk, you may be considering the entire universe of relevant documents (except for perhaps applications that may be about to issue). To the extent you're interested in all relevant prior art, your search is incomplete.
One last observation: the system you describe is somewhat mechanical (some decisions made without user discretion) and therefore may lend itself to automation. In short, you may be able to implement much of that as a simple shell script in Linux/Unix or some *nix variant -- including cygwin for Windows.
In terms of the "coverage" of your searching technique for US patents, it's probably more thorough than what I do, but then I'm not a searcher. For any assessment of relevance, looking at the actual patent (and seeing the drawings) helps tremendously. There are many ways to get quick, free downloads of patent images -- my favorite and a natural in your situation (particularly is automated by scripting) is the pat2pdf shell script for *nix. Check your favorite search engine for URL.
Regards.
|
| « Last Edit: Oct 19th, 2005, 5:28pm by JimIvey » |
IP Logged |
--
James D. Ivey
Law Offices of James D. Ivey
http://www.iveylaw.com
|
|
|
Sue
Guest
|
|
Re: do-it-yourself patent search methods
« Reply #2 on: Oct 19th, 2005, 5:15pm » |
Quote Modify
Remove
|
The inventions I am considering are all medical devices. If all prior art needs to be considered - then a medical literature search is a given. But other than that, I dont know where else to look. Am I required to search International patents as well for prior art? If so, what search engine is good?
|
|
IP Logged |
|
|
|
JimIvey
Moderator
Senior Member
Posts: 2584
|
|
Re: do-it-ypurself patent search methods
« Reply #3 on: Oct 19th, 2005, 5:28pm » |
Quote Modify
|
That raises an interesting point. You're not required to search at all. Whether you ought to search is an entirely different question.
Whether you ought to search depends on the importance to your business model of having an issued (and enforceable) patent. If it's absolutely essential that you can ultimately get your patent issued, then searching can save you much wasted effort if you can reasonably quickly find that you won't get a patent.
If you've determined that you ought to search, the next question is how much? Searching helps you reduce risk that there's some prior art out there that you don't know about. You can never completely eliminate that risk and the amount by which you reduce that risk, when mapped against searching effort and cost, follows a curve of diminishing returns. In other words, a little bit of effort gives you a lot of information. After you've search for a while, small increments in knowledge of the universe of applicable prior art are expensive in time and effort (or perhaps in money as well).
So, the short answer (that really isn't an answer) is that you should search until you're comfortable with the amount of risk that prior art unknown to you is out there and will prevent you from getting your patent.
I hope that helps.
|
|
IP Logged |
--
James D. Ivey
Law Offices of James D. Ivey
http://www.iveylaw.com
|
|
|
Sue
Guest
|
|
Re: do-it-ypurself patent search methods
« Reply #4 on: Oct 19th, 2005, 5:51pm » |
Quote Modify
Remove
|
Thank you both for your replies, they are quite helpful. Ive heard others recommend using professional searchers, but I think that the search process is something I need to learn for myself to avoid wasting time and money on fruitless ideas. I have a number of ideas for inventions, and work as an academic physician in a very technologically evolving area of medicine, so I want to learn the basic ins-and-outs of patenting medical devices. Thinking of the search in terms of comfort level and diminishing returns is especially helpful.
|
|
IP Logged |
|
|
|
|
|
Forums
Help
Search
Members
Calendar
Login
Register
Reply
Send Topic
Print