Potentially useful US Patent Classification statistics

[Jeffrey Dixon]

As another analyzer feature, Petapator could generate (or somehow help a user to identify) one or more subsets of U.S. Classifications (or truncated classification terms) that captures all or  some threshold percentage of the patents returned by the user search.  

The subset(s) could be selected/determined based on the smallest number of classifications required to capture all results and/or the smallest sample of patents contained in the union of the classification groups (or largest, or a range of sizes).   The idea would be to produce one or more suggested subsets of classifications that could be useful in a refined follow-up search in which the user would take out one or more other restrictions (typically a keyword restriction).  In my own past searching, I have used intuition/trial-and-error to choose a sample of classifications for follow-up searches.  The Petapator bar chart will already help me do that more efficiently, but it could be even better with some more automated analytics.  

Possible specific implementation - with the classifications sorted by frequency, split the bar for each classifcation below the highest-frequency classification into two colors, one color representing the number of results that overlap with one or more of the higher-frequency classifications listed above, and another color representing the number of results that do not overlap.    

A simpler implementation - give frequency statistics for parent classifications and not just for the narrower subclassifications.  This would give a user an idea of how segregated or integrated the results are across subclassifcations within a given parent classification.  

I would love to hear others' thoughts on this as well as the Petapator developers' thoughts.     

[Kenneth Yip]

Hi Jeffrey,

Do you mean that I should have better group of sub classifications?  For example, I can group by the first four letters of IPC, then for each group, I can show bar charts for each subcalss.

If possible, could you post some pictures of what you want to achieve? FYI, I am working hard to develop a new version that has better charts.

Thanks

Kenneth

--
You received this message because you are subscribed to the Google Groups "User Forum Petapator / Aspator" group.
To unsubscribe from this group and stop receiving emails from it, send an email to user--petapator-a...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jeffrey Dixon]

Yes to showing the totals for the parent classification and for its child subclassifications, but also the overlaps.  My initial bar chart idea is a little hard to explain, but I think a Venn diagram would be better anyway.  

The Venn diagram wouldn't necessarily have to be for every subclassification in the search, and indeed that would probably get way too messy.  Maybe it could just be for the N most frequent classificaitons, where the user chooses N?  

--
You received this message because you are subscribed to a topic in the Google Groups "User Forum Petapator / Aspator" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/user--petapator-aspator/13xODkbgLjM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to user--petapator-a...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
  ------m--------m------
        | | (OO) | |
            ||(~)||
   
         Monkey

[Jeffrey Dixon]

Showing how more than three different groups overlap could get convoluted with a Venn diagram, but visualization could be aided by color coded shape outlines, and/or allowing the user to highlight one or more shapes (increase its lineweight, shade it?) by clicking on it.