Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
Acknowledgment is made of the information disclosure statements filed on 12/14/2022, 02/17/2023, 09/19/2023, and 12/23/2025, U.S. patents and Foreign Patents have been considered.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains.
Claims 1, 2, 3, 4, 5, 6, 7, and 23 are rejected under 35 U.S.C. 103 as being unpatentable over US20230324881A1 (Lee) in view of US20050064302A1 (Kotani) and further in view of US20230044490A1 (Zhang).
In regards to claim 1, (Lee) shows a process proximity correction method comprising:
generating a second layout by performing machine learning-based process proximity correction on the first layout based on first to n-th features of the first to m-th patterns, wherein n is a natural number greater than or equal to 2; Lee [0023]–[0024] teaches that a second layout is generated by performing an ML-based PPC method on the first layout through ML-based inference on features of the patterns. Lee [0045] teaches extracting features from each pattern including tone, direction, length, density, sublayer, and width and space of neighboring segments. Lee [0048] teaches generating a prediction model through ML based on the extracted features.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose receiving a first layout including first to m-th regions, wherein each of the first to m-th regions include first to m-th patterns and m is a natural number equal to or greater than 3; wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions, wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m;
Kotani teaches receiving a first layout including first to m-th regions, wherein each of the first to m-th regions include first to m-th patterns and m is a natural number equal to or greater than 3; Kotani [0026]–[0031] teaches selecting cell patterns from a cell library where each cell pattern is a distinct layout region composed of multiple patterns, and placing the corrected cell patterns on a chip layout in an arrangement of three or more adjacent regions each containing its own pattern set.
1. (Kotani) differs from the claimed invention in that it does not explicitly disclose wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions, wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m;
Zhang teaches wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions; wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m; Zhang [0056]–[0063] teaches partitioning a design layout into a plurality of cells where each cell corresponds to a distinct region of the layout, and assigning a distinct plurality of variables (sub-features) within each cell that are specific to the target pattern of that region. Zhang [0063] and [0068] teach that during inference, only the variables of the relevant cell region are active while variables of non-target cells are set to zero, directly teaching a region-indexed sub-feature structure within each feature.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 2, (Lee) shows the method of claim 1, wherein the generating of the second layout includes:
extracting the first to n-th features of the first to m-th patterns from the first layout; Lee [0045]–[0047] teaches extracting one or more features from each pattern in the first layout in operation S121, including characteristics of each pattern and influences from adjacent patterns during etching, and tagging the extracted features to each pattern.
generating an after-cleaning inspection (ACI) image by performing machine learning-based inference based on the first to n-th features; Lee [0048]–[0052] teaches generating a prediction model through ML based on the extracted features, and in operation S125, predicting the ACI by inputting the layout into the prediction model to perform ML-based inference on the features and produce a predicted ACI image.
In regards to claim 3, (Lee) shows the method of claim 2, wherein the generating of the ACI image includes:
performing first machine learning-based inference on the first to n-th features, wherein the first machine learning-based inference is based on linear regression; Lee [0048] teaches that “linear regression, which is a one-to-one function and strong to extrapolation, may be used” as the first ML approach for generating the prediction model from the extracted pattern features.
performing second machine learning-based inference on a result of the first machine learning-based inference, wherein the second machine learning-based inference is based on non-linear regression; Lee [0048] further teaches that “an advanced ML having strong interpolation performance such as random forest may be used” as the second ML approach applied on the result of the first, and that “the correction convergence and the performance of the prediction model may be complemented with each other.”
In regards to claim 4, (Lee) shows the method of claim 3;
Lee teaches wherein the performing of the first machine learning-based inference includes performing the first machine learning-based inference on each of first to l-th sub-features included in each of the first to k-th features; Lee [0048]–[0050] teaches performing linear regression-based inference for each per-area condition model, generating a separate prediction model for each area region (e.g., center, middle, edge), constituting per-region sub-feature inference.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose where l is a natural number smaller than or equal to m;
Zhang teaches where l is a natural number smaller than or equal to m; Zhang [0062]–[0066] teaches independently adjusting values of the region-specific variables within each cell during the iterative inference process, directly teaching inference performed individually on each region’s sub-feature set.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 5, (Lee) shows the method of claim 2, wherein the method further comprises:
correcting the first layout based on a difference between the ACI image and a target ACI image; Lee [0052]–[0054] teaches that in operation S126, the predicted ACI is compared with the ACI target, and when the difference is greater than a preset threshold the first layout is corrected again in operation S124.
performing the machine learning-based inference based on the first to n-th features of the corrected first layout for generating a corrected ACI image; Lee [0054]–[0055] teaches that operations S124 and S125 are repeated — the corrected layout is input to the prediction model to generate a new predicted ACI image — until the predicted ACI is within the allowable range of the ACI target.
In regards to claim 6, (Lee) does not show wherein the correcting of the first layout includes correcting the first pattern of the first region, wherein generating the corrected ACI image includes: generating the corrected ACI image by performing the machine learning-based inference based on the (k+1)-th to n-th features of the corrected first pattern, and a first sub-feature of each of the first to k-th features of the corrected first pattern;
Zhang teaches wherein the correcting of the first layout includes correcting the first pattern of the first region, wherein generating the corrected ACI image by performing the machine learning-based inference based on the (k+1)-th to n-th features of the corrected first pattern, and a first sub-feature of each of the first to k-th features of the corrected first pattern; Zhang [0063]–[0068] teaches that when performing inference on a pattern in a specific cell (first region), only the variables (sub-features) of that first cell region are active alongside the shared features, while variables of all other cell regions are set to zero, directly teaching inference using the shared (k+1)-th to n-th features combined with only the first-region sub-feature of the individual features when correcting a first-region pattern.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 7, (Lee) does not show wherein the correcting of the first layout includes correcting a pattern of one the first to m-th regions, wherein the generating the corrected ACI image includes: generating the corrected ACI image by performing the machine learning-based inference based on the (k+1)-th to n-th features of the corrected pattern;
Zhang teaches wherein the correcting of the first layout includes correcting a pattern of one the first to m-th regions, wherein the generating the corrected ACI image includes: generating the corrected ACI image by performing the machine learning-based inference based on the (k+1)-th to n-th features of the corrected pattern; Zhang [0074]–[0078] teaches that for patterns in shared or symmetric regions, only the shared symmetric variables (shared features) are active during inference while the region-specific individual sub-feature variables are not applied, directly teaching inference based only on the shared (k+1)-th to n-th features for a pattern in a shared region.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 23, (Lee) shows a computing device for performing process proximity correction, the device comprising:
generate a second layout by performing machine learning-based process proximity correction based on first to n-th features of the first to m-th patterns, where n is a natural number greater than or equal to 2; Lee [0023]–[0024] and [0048] teach that a processor executing an ML-based PPC module generates a second layout by performing ML-based inference on extracted features of the patterns.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose receive a first layout including first to m-th regions, wherein each of the first to m-th regions include first to m-th patterns and m is a natural number equal to or greater than 3; wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions, wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m;
Kotani teaches receive a first layout including first to m-th regions, wherein each of the first to m-th regions include first to m-th patterns and m is a natural number equal to or greater than 3; Kotani [0062] teaches an information processing apparatus including a CPU and storage section executing programs for carrying out PPC procedures on a layout that includes a plurality of cell pattern regions each containing its own pattern set.
1. (Kotani) differs from the claimed invention in that it does not explicitly disclose wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions, wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m;
Zhang teaches wherein each of the first to k-th features includes first to l-th sub-features of each of the first to l-th patterns included in each of the first to l-th regions; wherein k is a natural number smaller than or equal to n and l is a natural number smaller than or equal to m; Zhang [0092]–[0095] teaches a computer system with a processor executing instructions for partitioning a design layout into a plurality of cells and assigning region-specific sub-feature variable sets within each cell, where non-target cell variables are set to zero during inference.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over US20230324881A1 (Lee) in view of US20050064302A1 (Kotani) and US20230044490A1 (Zhang) as applied to claims 1 above, respectively, and further in view of US20070038417A1 (Huang).
In regards to claim 8, (Lee) as modified by Kotani and Zhang does not show a first sub-feature of each of the first to k-th features, of the first pattern included in the first region has a first weight, wherein the first sub-feature of each of the first to k-th features, of the second pattern included in the second region has a second weight different from the first weight;
Huang teaches a first sub-feature of each of the first to k-th features of the first pattern included in the first region has a first weight, wherein the first sub-feature of each of the first to k-th features of the second pattern included in the second region has a second weight different from the first weight; Huang [0017]–[0018] teaches a weighted merit function for OPC model regression in which a weighting factor wi is applied individually to each feature data point, where different weighting factors are assigned to different data points, directly teaching differential weighting of the same sub-feature type across different patterns/regions.
1. The motivation to combine Lee, Kotani, Zhang, and Huang is that all four references relate to model-based semiconductor pattern correction. A person of ordinary skill in the art would have incorporated Huang’s differential per-data-point weighting into the Lee-Kotani-Zhang ML-based PPC framework to differentially weight sub-features across regions with reasonable expectation of success.
Claims 9, 10, 13, 15, 18, 20, and 21 are rejected under 35 U.S.C. 103 as being unpatentable over US20230324881A1 (Lee) in view of US20050064302A1 (Kotani).
In regards to claim 9, (Lee) shows a process proximity correction method comprising:
extracting first to third features of the first to third patterns; Lee [0045]–[0047] teaches extracting one or more features from each pattern in the layout in operation S121, including tone, direction, length, density, sublayer, and width and space of neighboring segments, and tagging the extracted features to each corresponding pattern.
generating a process proximity correction model, wherein the generating of the process proximity correction model includes performing machine learning on: first-first feature data about the first feature of the first pattern included in the first region; first-second feature data about the first feature of the second pattern included in the second region; second feature data about the second feature of the first to third patterns respectively included in the first to third regions; and measure data of an after-cleaning inspection (ACI) image generated from the first layout; Lee [0045]–[0048] teaches generating a PPC prediction model through ML by performing ML on extracted pattern features including per-region feature data paired with ACI measurement data as ground truth.
predicting an ACI image of the second layout using the process proximity correction model; Lee [0052]–[0053] teaches operation S125 in which the ACI is predicted by inputting the second layout into the prediction model.
correcting the second layout based on a difference between the predicted ACI image and a target ACI image; Lee [0053]–[0055] teaches operation S126 in which the predicted ACI is compared with the ACI target and when the difference exceeds the threshold the second layout is corrected again.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose receiving a first layout including a first region including a first pattern, a second region including a second pattern, and a third region including a third pattern; and correcting the first layout to generate a second layout;
Kotani teaches receiving a first layout including a first region including a first pattern, a second region including a second pattern, and a third region including a third pattern; Kotani [0026]–[0031] teaches receiving a layout composed of cell pattern regions, including a first cell region containing a first cell pattern, a second cell region containing a second cell pattern, and a third cell region containing a third cell pattern placed on a chip.
Kotani teaches correcting the first layout to generate a second layout; Kotani [0028]–[0033] teaches performing PPC on cell patterns to generate a corrected second layout.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
In regards to claim 10, (Lee) as modified by Kotani shows the method of claim 9:
wherein each of the first to third patterns includes a plurality of sub-patterns; Kotani [0027] teaches that the cell pattern of the metal interconnection layer is composed of several individual patterns such as power line patterns and corner patterns, directly teaching that each region’s cell pattern is composed of a plurality of sub-patterns.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
In regards to claim 13, (Lee) as modified by Kotani shows the method of claim 9, wherein the generating of the process proximity correction model includes:
performing first machine learning on the first-first feature data, the first-second feature data, and the second feature data to generate a first model, wherein the first machine learning is based on linear regression; Lee [0048] teaches that “linear regression, which is a one-to-one function and strong to extrapolation, may be used” as the first ML approach for generating a first model from the feature data.
performing second machine learning on a result of the first model to generate a second model, wherein the second machine learning is based on non-linear regression; Lee [0048] further teaches that “an advanced ML having strong interpolation performance such as random forest may be used” as the second ML approach applied on the result of the first model, and that the correction convergence and model performance are complemented when the two approaches are combined.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
In regards to claim 15, (Lee) as modified by Kotani does not show wherein the first pattern overlaps a boundary line between the first region and the second region contacting each other, wherein the generating of the process proximity correction model includes performing machine learning on the measure data of the ACI image, and the first-first feature data and the first-second feature data except for the second feature data;
wherein the generating of the process proximity correction model includes performing machine learning on the measure data of the ACI image, and the first-first feature data and the first-second feature data except for the second feature data; Lee [0057]–[0061] teaches performing per-area ML analysis for patterns at boundaries between adjacent wafer areas using the feature data of both adjacent areas, directly teaching ML on the feature data of both adjacent regions for a boundary-overlapping pattern.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose wherein the first pattern overlaps a boundary line between the first region and the second region contacting each other;
Kotani teaches wherein the first pattern overlaps a boundary line between the first region and the second region contacting each other; Kotani [0032]–[0037] teaches that a pattern positioned at the boundary between neighboring cell regions receives optical proximity effect influence from both adjacent regions, and that PPC for such boundary-neighborhood patterns must account for the neighboring cell region’s patterns.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
In regards to claim 18, (Lee) as modified by Kotani shows the method of claim 9:
wherein each of the first to third features includes at least one of: a size of each of the first to third patterns; a density of the first to third patterns; a distance between adjacent ones of the first to third patterns; a size of one of the first to third patterns and a size of a pattern neighboring thereto; an angle defined between adjacent ones of the first to third patterns; or a relative position in a vertical direction of each of the first to third patterns arranged vertically; Lee [0046] explicitly enumerates extracted features including tone (angle), direction, length and width (size), density, sublayer (vertical position), and width and space of neighboring segments (distance and size of neighboring pattern), constituting all feature types recited in the claim.
In regards to claim 20, (Lee) as modified by Kotani does not show wherein the first layout includes first coordinates indicating the first region, second coordinates indicating the second region, and third coordinates indicating the third region;
Kotani teaches wherein the first layout includes first coordinates indicating the first region, second coordinates indicating the second region, and third coordinates indicating the third region; Kotani [0032]–[0036] teaches that each cell region in the layout is identified by coordinate markers indicating the bounds and position of each cell region, directly teaching first, second, and third coordinates in the layout indicating the respective first, second, and third regions.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
In regards to claim 21, (Lee) as modified by Kotani does not show wherein the first layout includes a first sub-layout of the first region, a second sub-layout of the second region, and a third sub-layout of the third region;
Kotani teaches wherein the first layout includes a first sub-layout of the first region, a second sub-layout of the second region, and a third sub-layout of the third region; Kotani [0043]–[0045] teaches that each cell pattern region in the layout is embodied as a separate GDS sub-layout file (e.g., GDS2), directly teaching a first sub-layout for the first region, a second sub-layout for the second region, and a third sub-layout for the third region.
1. The motivation to combine Lee and Kotani is that both references relate to process proximity correction for semiconductor manufacturing. A person of ordinary skill in the art would have combined Lee’s ML-based PPC model generation and iterative ACI correction loop with Kotani’s multi-region cell structural framework with reasonable expectation of success.
Claims 11, 12, 14, and 16 are rejected under 35 U.S.C. 103 as being unpatentable over US20230324881A1 (Lee) in view of US20050064302A1 (Kotani) as applied to claim 9 above, respectively, and further in view of US20230044490A1 (Zhang).
In regards to claim 11, (Lee) as modified by Kotani does not show wherein the first feature includes a plurality of first sub-features, wherein the first-first feature data includes a plurality of first-first sub-feature data about the plurality of first sub-features of the first pattern included in the first region, wherein the first-second feature data includes a plurality of second-first sub-feature data about the plurality of first sub-features of the second pattern included in the second region;
Zhang teaches wherein the first feature includes a plurality of first sub-features, wherein the first-first feature data includes a plurality of first-first sub-feature data about the plurality of first sub-features of the first pattern included in the first region, wherein the first-second feature data includes a plurality of second-first sub-feature data about the plurality of first sub-features of the second pattern included in the second region; Zhang [0058]–[0063] teaches that within each cell corresponding to a region of the layout, a plurality of variables is assigned constituting the plurality of sub-features of the feature for that region. The variables assigned to the first cell region constitute first-first sub-feature data and the variables assigned to the second cell region constitute first-second sub-feature data, directly teaching that the first feature is divided into a plurality of sub-features with region-specific sub-feature data sets for the first and second regions respectively.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 12, (Lee) as modified by Kotani does not show wherein the second feature includes a plurality of second sub-features, wherein the second feature data includes a plurality of second sub-feature data about the plurality of second sub-features of the first to third patterns;
Zhang teaches wherein the second feature includes a plurality of second sub-features, wherein the second feature data includes a plurality of second sub-feature data about the plurality of second sub-features of the first to third patterns; Zhang [0074]–[0078] teaches assigning symmetric shared variables that are applied uniformly across multiple cells and constitute the shared sub-features of the second feature. The shared variable set encompasses feature data from all cells (first to third regions) simultaneously, directly teaching that the second feature includes a plurality of second sub-features and that the second feature data includes sub-feature data for the first to third patterns.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 14, (Lee) as modified by Kotani does not show wherein the generating of the process proximity correction model includes: performing machine learning on the measure data of the ACI image, and the first-first feature data except for the first-second feature data and the second feature data; performing machine learning on the measure data of the ACI image, and the first-second feature data except for the first-first feature data and the second feature data; and, performing machine learning on the measure data of the ACI image, and the second feature data except for the first-first feature data and the first-second feature data;
Zhang teaches performing machine learning on the measure data of the ACI image, and the first-first feature data except for the first-second feature data and the second feature data; Zhang [0082]–[0088] teaches generating a result library by independently processing each selected target pattern’s cell variable set (feature set) in isolation against performance ground truth data and storing the results separately, corresponding to an isolated ML pass for the first-first feature set against the ACI measure data excluding the other feature sets.
Zhang teaches performing machine learning on the measure data of the ACI image, and the first-second feature data except for the first-first feature data and the second feature data; and, performing machine learning on the measure data of the ACI image, and the second feature data except for the first-first feature data and the first-second feature data; Zhang [0086]–[0089] teaches that the result library includes individually processed results for each pattern/cell feature set, corresponding to isolated ML passes for the first-second feature set and the second feature set against the ACI measure data, each processed independently of the other feature sets.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
In regards to claim 16, (Lee) as modified by Kotani does not show wherein the generating of the process proximity correction model further includes performing machine learning on first-third feature data about the third feature of the first pattern included in the first region, and the measure data of the ACI image;
performing machine learning on additional feature data of a pattern in a specific region against ACI measure data; Lee [0045]–[0047] teaches extracting multiple feature types from each pattern including harmonics and next/previous segment information, supporting the extraction of a third feature type from the first pattern in the first region.
1. (Lee) differs from the claimed invention in that it does not explicitly disclose performing machine learning on first-third feature data about the third feature of the first pattern included in the first region, and the measure data of the ACI image;
Zhang teaches performing machine learning on first-third feature data about the third feature of the first pattern included in the first region, and the measure data of the ACI image; Zhang [0085]–[0087] teaches that the result library is expanded by processing additional pattern/cell feature sets against performance data, where additional feature types specific to individual cell regions are incorporated against performance data, directly teaching an additional ML pass on third-feature data of the first region against ACI measure data.
1. The motivation to combine Lee, Kotani, and Zhang is that all three references relate to ML-based semiconductor pattern correction. A person of ordinary skill in the art would have combined Lee’s ML-based PPC framework with Kotani’s multi-region cell structure and Zhang’s per-region indexed variable approach to achieve a region-specific sub-feature ML-based PPC system with reasonable expectation of success.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANWER AHMED ALAWDI whose telephone number is (703)756-1018. The examiner can normally be reached Monday - Friday 8:00 am - 5:30 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Jack Chiang can be reached on (571)-272-7483. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ANWER AHMED ALAWDI/Examiner, Art Unit 2851 /JACK CHIANG/ Supervisory Patent Examiner, Art Unit 2851
Prosecution Insights