Appeal 2009012027; Serial. No. 10/715,162; Tech. Center 2100
Decided: August 28, 2011
The application on appeal was a computer application for managing media file storage. The claim limitation at issue related to the program's generation of a timeline view: "a time bar that divides time into segments of unit time, each segment of unit time having a respective length along the time bar that depends upon the amount of media files associated with the respective segment of unit time." (Emphasis added.)
The Examiner rejected the claim as obvious over the media file viewers taught in Rothmuller and Parker.
Rothmuller's viewer (shown below) included a timeline 250 showing the number of photos taken as a function of time. The user selected the time period by adjusting time bands 251.
|Rothmuller, FIG. 3|
The Examiner relied on Parker's viewer for teaching a media viewer depicting media files as a histogram, with a horizontal bar "divided into segments of unit time" as recited in the claim.
|Parker, FIG. 6G (color added to better show histogram bins)|
Finally, the Examiner acknowledged that neither reference disclosed "segments of time having length along time bar (horizontally)" (emphasis added). However, the Examiner concluded that since the references showed vertical bars that represent the amount of media files associated with that time unit, it would have been obvious to "modify the vertical graphs for a specific time segment to horizontal graph along the time bar" (emphasis added). The Examiner cited In re Stevens (101 USPQ 284 (CCPA 1954)) for the proposition that adjustability involves only routine skill in the art.
The Applicant appealed, and argued in the Brief that the claimed time bar was not disclosed by the asserted combination. In particular, the Applicant argued that the combination did not teach a "length along the time bar that depends upon the amount of media files associated with the respective segment of unit time." The Applicant explained that, depending on how "segment" was interpreted, Rothmuller's timeline was either divided into segments of equal size (six months), or the segment size was independently chosen by a user through adjusting time bands 251. The Applicant further explained that Parker taught histogram bins that are uniformly spaced, where the number of bins depends on the screen resolution.
In the Answer, the Examiner responded to Applicant's arguments and further explained his position. The bar graph [in FIG. 3 of Rothmuller] "represented for certain segments of time show respective length/height, which is along the time bar for that segment of time and this respective length/height represents the amount of media/photos associated with that time segment." (Emphasis added.) Parker's FIG. 6B showed "may 6th as segment of time and may 6th as having respective length (sideways or horizontal) which depends on the number of thumbnails associated with may 6th." (Emphasis added.)
The Applicant filed a Reply Brief and argued the specific limitation "along the time bar." The Applicant noted that the "vertical bar graphs shown in Fig. 3 of Rothmuller have respective lengths transverse to the timeline that depend upon the amount of media files associated with the respective bar graph." (Emphasis added.) The Applicant made the same point with respect to FIG. 6C of Parker: "The graphs of Parker (i, e., the bars rising vertically from each of the boxes numbered 1,2,3.. .) are not "along the time bar" as asserted by the Examiner, but are transverse to the histogram timeline shown in the figure, as is typical for a conventional histogram timeline as disclosed by Parker."
The Board affirmed the obviousness rejection. However, the Board found that the combination was not necessary, since the appealed claim was obvious over either Parker or Rothmuller.
The Board's conclusion was based on claim construction of the phrase "length along the time bar that depends upon the amount." In response to the Examiner's admission that the time bars in the references were horizontal, the Applicant had argued that the histogram bins (claimed "amount") were vertical, and thus transverse to the timeline rather than "along the time bar" as claimed. The Board noted that the horizontal aspect was not "commensurate in scope with the specific language of claim 1," since claim 1 did not recite "such 'horizontally' oriented time bars." The Board interpreted the length phrase as follows:
Claim 1 does not place any limitation on what "along the time bar" mean, includes, or presents. Thus, we give this claim limitation its broadest reasonable interpretation as any relative position with respect to the time bar (whether vertically or horizontally), as specifically defined in claim 1.The Board then applied this claim interpretation to each of the references in turn:
[W]e find that the bar widths [in Rothmuller] represent the segments of unit time having a respective length (bar height) which is in a relative position with respect to the time bar since the bars are connected to the timeline and spaced along the timeline. ... We find the timeline along the horizontal axis [in Parker] to be the time bar that is divided into segments of time. We find further that the bar heights of the bars represent the "length" that "depends upon the amount of media files" since the bar height is indicative of the number of digital multimedia objects.My two cents: The Board got the claim construction on this one way wrong. The Board's interpretation of "along the time bar" completely ignored the ordinary meaning of "along." In this context, "along" implies a specific positional relationship which is not captured by the Board's interpretation of "any relative relationship."
In view of our claim interpretation above, we find that disclosure from either reference, Rothmuller or Parker, teaches and would have suggested providing "a time bar that divides time into segments of unit time, each segment of unit time having a respective length along the time bar that depends upon the amount of media files associated with the respective segment of unit time," as specifically required by claim 1.
(Internal citations omitted, emphasis added.)
The Board's discussion suggested that the Applicant relied on the time bar being horizontal, but this was misleading. The Applicant's actual argument did not rely on a specific orientation for the time bars or histogram bins, but instead (correctly) focused on the positional relationship: the histograms in the references (the claimed "amount") were transverse to, rather than along, the time bar.
The Examiner's rejection was on the right track because it didn't rely on an erroneous claim interpretation. The Examiner admitted that each of two references did not meet the "along the time bar" limitation and relied on an "obvious to adjust vertical to horizontal" rationale.
I think the Applicant wins even under this rationale. I see the switch from vertical histogram bins to a horizontal arrangements to be counterintuitive. That is, a histogram graph is expected to use different axes for the bin size and the time (or whatever grouping arrangement is used). So a POSITA would not find it obvious to flip the bins from a vertical arrangement to a horizontal arrangement. Ideally, the Applicant would argue this point, rather than leaving a hole in the attack on obviousness by not addressing the Examiner's rationale.
In my view, this decision merits a request for rehearing by the Board. Such rehearings are strictly limited to arguments already on the record, but here the Applicant has everything it needs to argue "the points believed to have been misapprehended or overlooked by the Board" (the standard for rehearing).
Postscript: After losing at the BPAI, the Applicant filed an RCE and amended the claims to recite "a time bar extending in a predefined direction ... each segment of unit time having a respective width, as measured in the predefined direction." The Examiner rejected using essentially the same rationale as before (obvious to switch vertical to horizontal), though he cited different case law (In re Einstein, mere reversal is obvious).