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 .
DETAILED ACTION
Claims 1 – 21 are pending.
Any references to applicant’s specification are made by way of applicant’s U.S. pre-grant printed patent publication.
This action is in response to the communication filed on 3/23/26.
All objections and rejections not set forth below have been withdrawn.
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 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) 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. Patentability shall not be negated by the manner in which the invention was made.
Claims 1 – 16 and 18 – 21 are rejected under 35 U.S.C. 103 as being unpatentable over Wojnowicz et al. (Wojnowicz), “SUSPEND: Determining software suspiciousness by non-stationary time series modeling of entropy signals”, in view of Shafiq et al. (Shafiq), “Embedded Malware Detection Using Markov n-Grams”, in view of Singh et al. (Singh), US 2019/0294792 A1.
Regarding claim 1, Wojnowicz discloses:
A method of detecting malicious activity with respect to a file, the method comprising (e.g. Wojnowicz, Abstract).
While Wojnowicz demonstrates a system for detecting entropy within a file by conducting a level of entropy detection (e.g. Wojnowicz, fig. 6), Wojnowicz does not appear to explicitly teach conducting first and second levels of entropy detection.
However, Wojnowicz does suggest to combine the teachings of Shafiq, who also teaches a system detecting entropy within a file by conducting a level of entropy detection (e.g. Shafiq, Abstract), within a single system including the teachings of Wojnowicz (e.g. Wojnowicz, sect. 8.2: step 2). Thus it would have been obvious to one of ordinary skill in the art combine the entropy detection level of Shafiq with the system of Wojnowicz based upon the explicit suggestion within the art to combine simpler entropy detection methods with the more complex structural entropy detection method of Wojnowicz (e.g. Wojnowicz, sect. 6.3, par. 1; sect. 8.2: step 2).
Thus, the combination enables:
conducting a first level of ransomware detection to the file (e.g. Shafiq, Abstract; pg. 89, par. 2), wherein the first level of ransomware detection comprises:
identifying features of the file that include a measure of randomness in the file at a first granularity of the file (e.g. Shafiq, Abstract; sect. 5.2 – entropy within a file block size of 1000 bytes – i.e. “a first granularity of the file”);
inputting the features to a machine learning model that outputs a determination of whether the file has been attacked (e.g. Shafiq, sect. 6.4; sect. 7; table 3 – training and detection using a classification model which outputs a malware classification).
and determining … to conduct a second level of ransomware detection … (e.g. Wojnowicz, sect. 8.2: step 2; sect. 4.2; fig. 6 – structural entropy detection using a large scale model), wherein the second level of ransomware detection makes a second determination of whether the file has been attacked (e.g. Wojnowicz, fig. 6 – suspiciousness score) based on additional features of the file that include a second measure of randomness in the file at a finer granularity than the first granularity (e.g. Wojnowicz, sect. 4.1 – entropy detection within file blocks of 256 bytes, i.e. “a finer granularity”).
Wojnowicz teaches that the “second level” of ransomware detection uses a large model and is computationally heavy and expensive (e.g. Wojnowicz, sect. 5.3.2; sect. 6.3) and is considered too demanding for certain devices (e.g. Wojnowicz, sect. 8.1, par. 2). However, Wojnowicz does not appear to explicitly teach that the determination to conduct the detection using the more computationally heavy and expensive large scale model is “based” upon a determination made by a detection using a lighter or less complex entropy detection model.
However, like Wojnowicz, Singh teaches the detection of malware using first and second models (e.g. Singh, Abstract; fig. 1; par. 11). Furthermore, like Wojnowicz, Singh also teaches that Iwanir teaches that some models are computationally heavy and expensive (e.g. Singh, par. 12), and that a determination to use a second level of malware detection using a computationally heavy model should be performed based upon a determination of a first computationally lighter model (e.g. Singh, par. 14).
It would have been obvious to one of ordinary skill in the art to apply the teachings of Singh within the system of Wojnowicz because one of ordinary skill in the art would have been motivated by the advantages of being able to quickly identify malware using a ligther model and then verifying detection results using a heavier model when necessary (e.g. Singh, par. 14, 18; fig. 2:220, 230).
Thus, the combination enables:
… and determining whether to conduct a second level of ransomware detection based on the determination … (e.g. Wojnowicz, sect. 8.2: step 2; sect. 4.2; fig. 6; Singh, par. 14, 18; fig. 2:220, 230).
Regarding claim 2, the combination enables:
wherein the second level of ransomware detection comprises:
identifying the additional features of the file (e.g. Wojnowicz, sect. 4.1 – structural entropy detection within non-overlapping chunks);
and inputting at least the additional features to a second machine learning model that outputs the second determination of whether the file has been attacked (e.g. e.g. Wojnowicz, fig. 6 – structural entropy measurements and suspiciousness score).
Regarding claim 3, the combination enables:
wherein the determination comprises a score comprising a value in a range of possible values (e.g. Singh, par. 18 – e.g. a score ranging from 0-100%), and wherein determining whether to conduct the second level of ransomware detection based on the determination comprises determining whether to conduct the second level of ransomware detection based on a location of the value in the range of possible values (e.g. Singh, par. 18 – e.g. a second level of machine learning classification can be determined if the score was located within the range of 65 - 100%).
Regarding claim 4, the combination enables:
in response to determining not to conduct a second level of ransomware detection (e.g. Singh, fig. 2:230; par. 19), initiating an action in association with the file when the determination indicates the file has been attacked (e.g. Singh, par. 18, 19 – when the system determines that the file has probably been attacked (e.g. at 64% probability) the system may continue to execute the file instructions).
Regarding claim 5, Wojnowicz does not appear to explicitly teach that the computationally heavy model is performed within the cloud. However, Singh teaches that the computationally heavy model should be performed within the cloud so as to reduce the burden placed upon individual devices (e.g. Singh, par. 12). It would have been obvious to one of ordinary skill in the art to employ the cloud teachings of Singh within the combination of Wojnowicz because one of ordinary skill in the art would have been motivated by the teachings that the cloud is a better location for performing computationally heavy models due to greater computational resourses (e.g. Singh, par. 12, 15) and would still allow actions to be taken respecting any detected malware file Singh, par. 15, 19, 21).
Thus, the combination enables:
in response to determining that a second level of ransomware detection should be conducted, communicate the additional features of the file to a cloud environment (e.g. Singh, par. 21; fig. 1:112; Wojnowicz, sect. 6.3, par. 1; sect. 8.2, step 2; Shafiq, sect. 6.4; sect. 7; table 3 – training and detection using a classification model which outputs a malware classification).
Regarding claim 6, the combination enables:
receiving a threat indication for the file from the cloud environment (e.g. Singh, par. 19, 21 – the deep model notifies the system that the suspicious file is verified or confirmed to be malware; Wojnowicz, fig. 6);
and initiating an action in association with the file based on the threat indication (e.g. Singh, par. 15, 19, 21 – a final classification (i.e. “action”) of the malware file of the endpoint client is made – wherein the system can also make a classification (i.e. “action”) of the security risk of the malware file).
The examiner notes, that while the claims do not explicitly limit the “threat indication …from the cloud environment” as being received by the client endpoint device so that the client device can itself initiate an action associated with the file, the examiner points out that this feature, also, would have been obvious in view of the combination.
Specifically, Singh teaches that the client endpoint device requires ransomware detection to initiate actions upon files that have been identified as ransomware (e.g. Singh, 3, 4), and furthermore that the client endpoint device requests the cloud service to make a final determination as to whether a file on the client endpoint is verified to be ransomware (e.g. Singh, par. 15). Thus, it would have been obvious to one of ordinary skill in the art to recognize that the client endpoint device should receive the indication of the final ransomware determination from the cloud service and be enabled to perform an appropriate action to the file in response, because one of ordinary skill in the art would have been motivated by the teaching that the client endpoint device relies upon relies upon the cloud service to effectuate an appropriate security solution at the location of the endpoint device (e.g. Singh, par. 15).
Regarding claim 7, the combination enables:
wherein the second features comprise at least a portion of the features (e.g. e.g. Shafiq, Abstract; sect. 5.2 – first features comprising portions of the file; e.g. Wojnowicz, sect. 4.1 – second features comprising portions of the file; Singh, par. 21).
Regarding claim 8, the combination enables:
wherein the measure of randomness comprises a measure of entropy for one or more chunks of the file (e.g. e.g. Shafiq, Abstract; sect. 5.2 – entropy measurment).
Regarding claims 9 – 16, they are apparatus claims essentially corresponding to the method above, and they are rejected, at least, for the same reasons.
Furthermore, regarding claim 9, Wojnowicz does not appear to explicitly teach a
specific hardware architecture for implementing the system. However, Singh teaches a computing architecture that permits the computation of light and heavy models for malware detection (e.g. Singh, fig. 1; fig. 3), and it would have been obvious to one of ordinary skill in the art to recognize the processor, storage, and instructions teachings of Singh within the system of Wojnowicz because one of ordinary skill in the art would have been motivated by the need to implement the system in practice.
Thus, the combination enables:
A computing apparatus comprising: a storage system; a processing system operatively coupled to the storage system; program instructions stored on the storage system to detect malicious activity with respect to a file that, when executed by the processing system, direct the computing apparatus to: … (e.g. Singh, fig. 1; fig. 3), as well as “different first sized portions” and “different second-sized portions” of the file (e.g. Shafiq, sect. 5.2; Wojnowicz, sect. 4.1 – entropy detection within file blocks of 256 bytes vs 1000 bytes)
Regarding claims 18 – 21, they are medium and method claims essentially corresponding to the method and apparatus claims above, and they are rejected, at least, for the same reasons.
Furthermore, because the combination enables dividing or “chunking” the file into
first and second sized “chunks” or portions, wherein the second size is smaller than the first (e.g. Shafiq, sect. 5.2; Wojnowicz, sect. 4.1 – entropy detection within file blocks of 256 bytes vs 1000 bytes)
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Wojnowicz et al. (Wojnowicz), “SUSPEND: Determining software suspiciousness by non-stationary time series modeling of entropy signals”, in view of Shafiq et al. (Shafiq), “Embedded Malware Detection Using Markov n-Grams”, in view of Singh et al. (Singh), US 2019/0294792 A1, in view of Tian et al. (Tian), "An automated classification system based on the strings of trojan and virus families”.
Regarding claim 17, Wojnowicz does not appear to explicitly teach that identified file features should comprise a file extension of the file. However, Tian does disclose identified features as comprising a file extension (e.g. Tian, sect. 3.2.2; fig. 2 – e.g. string comprising “.dll” extension).
It would have been obvious to one of ordinary skill in the art to incorporate the strings based (i.e. file extension) feature identification of Tian within the system combination of Wojnowicz. This would have been obvious because one of ordinary skill in the art would have been motivated by the teachings that entropy detection is augmented by string detection, such as taught by Tian (e.g. Wojnowicz, sect. 6; sect. 8.2, step 2), thus protecting against malware authors who manipulate file extensions to avoid detection (e.g. Wojnowicz, sect. 7, par. 2; sect. 7.5, par. 1).
Thus, the combination enables:
wherein the features further comprise a file extension of the file (e.g. Tian, sect. 3.2.2; fig. 2 – e.g. string comprising “.dll” extension).
Response to Arguments
Applicant’s arguments with respect to the pending claims have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
See Notice of References Cited.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JEFFERY L WILLIAMS whose telephone number is (571)272-7965. The examiner can normally be reached on 7:30 am - 4:00 pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Farid Homayounmehr can be reached on 571-272-3739. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/JEFFERY L WILLIAMS/Primary Examiner, Art Unit 2495
Prosecution Insights