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 CORRESPONDENCE
Status of Claims
Claims 1, 7, 13 have been amended.
Claims 2, 3, 8, 9, 14, 15 have been cancelled.
No claims have been added.
Priority
Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 4 – 7, 10 – 13, 16 – 18 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite:
obtaining (i) a first set of aerial images of a first set of properties, and (ii) an indication of a year built of each of the first set of properties
extracting feature values for features of the first set of properties, wherein at least one of the feature values is for a feature which is extracted from the first set of aerial images;
deriving predictor variables from the feature values which are more likely to be predictive of year built by generating a visualization of the feature values to identify relationships between the feature values and the year built;
determine a year built of a property using the predictor variables of the first set of properties and the year built of each of the first set of properties different from the first set of properties;
identifying a subject property;
receiving one or more aerial images of the subject property;
extracting feature values of the subject property using the same features used to [determine a square footage of a property using the features values of the first set of properties and the square footage of each of the first set of properties], wherein at least one of the feature values is for the feature which is extracted from the one or more aerial images of the subject property;
applying the predictor variables from the feature values of the subject property to determine a year built of the subject property; and
applying the predictor variables from the feature values of the subject property to determine a confidence level for the determination of the year built of the subject property;
generating a profile of the subject property with the determined year built and confidence level for the determination of the year built; and
providing the year built of the subject property with the confidence level for the determination of the year built for display
The invention is directed towards the abstract idea of real estate evaluation and assessment, which corresponds to ““Mental Processes”, “Certain Methods of Organizing Human Activities”, and “Mathematical Concepts” as it is directed towards steps that can be performed by humans, a human in their mind, and/or through the aid of pen and paper, e.g., having a human collect and look at a plurality of aerial images for a plurality of real estate properties, collect information that include an indication of the year built of those plurality of properties, look/observe the images to extract some information about the properties, identify and collect aerial images of a property of interest, look/observe the image of the property of interest, compare the information between the plurality of properties and property of interest to determine, guess, estimate, or the like the year built of the property of interest, creating a record/profile for the property, and displaying the results of the analysis along with its confidence level.
The limitations of:
obtaining (i) a first set of aerial images of a first set of properties, and (ii) an indication of a year built of each of the first set of properties
extracting feature values for features of the first set of properties, wherein at least one of the feature values is for a feature which is extracted from the first set of aerial images;
deriving predictor variables from the feature values which are more likely to be predictive of year built by generating a visualization of the feature values to identify relationships between the feature values and the year built;
determine a year built of a property using the predictor variables of the first set of properties and the year built of each of the first set of properties different from the first set of properties;
identifying a subject property;
receiving one or more aerial images of the subject property;
extracting feature values of the subject property using the same features used to [determine a square footage of a property using the features values of the first set of properties and the square footage of each of the first set of properties], wherein at least one of the feature values is for the feature which is extracted from the one or more aerial images of the subject property;
applying the predictor variables from the feature values of the subject property to determine a year built of the subject property; and
applying the predictor variables from the feature values of the subject property to determine a confidence level for the determination of the year built of the subject property;
generating a profile of the subject property with the determined year built and confidence level for the determination of the year built; and
providing the year built of the subject property with the confidence level for the determination of the year built for display
are processes that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of a generic processor executing computer code stored on a computer medium and generic machine learning algorithm. That is, other than reciting a generic processor executing computer code stored on a computer medium and generic machine learning algorithm nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the generic processor executing computer code stored on a computer medium and generic machine learning algorithm in the context of this claim encompasses a can collect and look at aerial images of real estate properties and indications of their year built and compare them against aerial images of a property of interest and, based on the comparison, determine, guess, estimate, or the like the year built of the property of interest, as well as creating a record/profile of the property and displaying the results of the analysis along with its confidence level. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of a generic processor executing computer code stored on a computer medium and generic machine learning algorithm, then it falls within the “Mental Processes” and “Certain Methods of Organizing Human Activities” groupings of abstract ideas. Accordingly, the claims recite an abstract idea.
This judicial exception is not integrated into a practical application. In particular, the claim only recites additional elements – a generic processor executing computer code stored on a computer medium and generic machine learning algorithm to communicate information, as well as performing operations that a human can perform in their mind or using pen and paper, i.e. comparing collected information of a plurality of real estate properties against a property of interest to determine, guess, estimate the year built of a property of interest. The generic processor executing computer code stored on a computer medium and generic machine learning algorithm in the steps are recited at a high-level of generality (i.e., as a generic processor executing computer code stored on a computer medium can perform the insignificant extra solution steps of communicating information (See MPEP 2106.05(g) while also reciting that the a generic processor executing computer code stored on a computer medium are merely being applied to perform the steps that can be performed in the human mind or using pen and paper; "[use] of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more.” Therefore, according to the MPEP, this is not solely limited to computers but includes other technology that, recited in an equivalent to “apply it,” is a mere instruction to perform the abstract idea on that technology (See MPEP 2106.05(f)) such that it amounts no more than mere instructions to apply the exception using a generic processor executing computer code stored on a computer medium and generic machine learning algorithm.
Although the claim recites “train a machine learning algorithm,” the claims and specification fail to provide sufficient disclosure regarding an improvement to how a machine learning algorithm can be trained, but simply recites a high-level generic recitation that a machine learning algorithm is being trained. There is insufficient evidence from the specification to indicate that the use of the machine learning algorithm involves anything other than the generic application of a known technique in its normal, routine, and ordinary capacity or that the claimed invention purports to improve the functioning of the computer itself or the machine learning algorithm. None of the limitations reflects an improvement in the functioning of a computer, or an improvement to other technology or technical field, applies or uses a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, implements a judicial exception with, or uses a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, effects a transformation or reduction of a particular article to a different state or thing, or applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception.
Even training and applying a machine learning algorithm model is simply application of a computer model, itself an abstract idea manifestation. Further, such training and applying of a model is no more than putting data into a black box machine learning operation. The nomination as being a machine learning algorithm is a functional label, devoid of technological implementation and application details. The specification does not contend it invented any of these activities, or the creation and use of such machine learning models. In short, each step does no more than require a generic computer to perform generic computer functions. As to the data operated upon, "even if a process of collecting and analyzing information is 'limited to particular content' or a particular 'source,' that limitation does not make the collection and analysis other than abstract." SAP America, Inc. v. InvestPic LLC, 898 F.3d 1161, 1168 (Fed. Cir. 2018).
The Examiner asserts that the scope of the disclosed invention, as presented in the originally filed specification, is not directed towards the improvement of machine learning, but directed towards real estate property evaluation and the data associated with real estate properties to determine, guess, or estimate the year built of a property of interest. The specification’s disclosure on machine learning is nothing more than a high general explanation of generic technology and applying it to the abstract idea. Referring to MPEP § 2106.05(f), the training and re-training are merely being used to facilitate the tasks of the abstract idea, which provides nothing more than a results-oriented solution that lacks detail of the mechanism for accomplishing the result and is equivalent to the words “apply it,” per MPEP § 2106.05(f). The Examiner asserts that in light of the 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence, the claimed invention is analogous to Example 47, Claim 2.
Further, the combination of these elements is nothing more than a generic computing system with machine learning model(s). Because the additional elements are merely instructions to apply the abstract idea to a computer, as described in MPEP § 2106.05(f), they do not integrate the abstract idea into a practical application.
Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a generic processor executing computer code stored on a computer medium and generic machine learning algorithm to perform the steps of:
obtaining (i) a first set of aerial images of a first set of properties, and (ii) an indication of a year built of each of the first set of properties
extracting feature values for features of the first set of properties, wherein at least one of the feature values is for a feature which is extracted from the first set of aerial images;
deriving predictor variables from the feature values which are more likely to be predictive of year built by generating a visualization of the feature values to identify relationships between the feature values and the year built;
determine a year built of a property using the predictor variables of the first set of properties and the year built of each of the first set of properties different from the first set of properties;
identifying a subject property;
receiving one or more aerial images of the subject property;
extracting feature values of the subject property using the same features used to [determine a square footage of a property using the features values of the first set of properties and the square footage of each of the first set of properties], wherein at least one of the feature values is for the feature which is extracted from the one or more aerial images of the subject property;
applying the predictor variables from the feature values of the subject property to determine a year built of the subject property; and
applying the predictor variables from the feature values of the subject property to determine a confidence level for the determination of the year built of the subject property;
generating a profile of the subject property with the determined year built and confidence level for the determination of the year built; and
providing the year built of the subject property with the confidence level for the determination of the year built for display
amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
Additionally:
Claim 4 is directed towards the recitation of generic technology and applying it to the abstract idea to, somehow, determine the garage size of the property of interest.
Claims 5, 6 are directed the insignificant extra solution activity of gathering and receiving information and describing the information.
The remaining claims are similar in scope to what has been discussed above.
In summary, the dependent claims are simply directed towards providing additional descriptive factors that are considered for determining, guessing, or estimating the year built of a real estate property. Accordingly, the claims are not patent eligible.
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 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, 4 – 7, 10 – 13, 16 – 18 are rejected under 35 U.S.C. 103 as being unpatentable over Rose et al. (US PGPub 2019/0333175 A1) in view of Lookabaugh (WO 2019/113397 A1) in further view of CaseGuard (How to Understand Confidence Level for Better Machine Learning).
In regards to claims 1, 7, 13, Rose discloses (Claim 1) a computer-implemented method for use in determining a year in which a subject property was built, the method comprising; (Claim 7) a computer system configured for use in determining a year built a subject property was built, the computer system comprising one or more processors configured to; (Claim 13) a non-transitory computer-readable memory storing instructions thereon, that when executed by one or more processors, cause the one or more processors to:
obtaining, by one or more processors, (i) a first set of aerial images of a first set of properties, and (ii) an indication of a year built of each of the first set of properties (¶ 12, 16, 75, 77, 98, 121, 122, 215, 227 wherein aerial images and an indication of the year built of each property are received; ¶ 16, 215 wherein the system applies data from the aerial images to determine the street address of the property and to further determine the year it was built);
extracting, by the one or more processors, feature values for features of the first set of properties, wherein at least one of the feature values is for a feature which is extracted from the first set of aerial images (¶ 12, 16, 75, 77, 98, 121, 122, 215, 227 wherein a plurality of different features can be extracted from the aerial images, such as, but not limited to, square footage, renovations, additions that add square footage, and etc., which can further be obtained from other sources of information, such as, but not limited to, city records, MLS listings, social listings, and etc. The “values” for the features can be as simple as yes/no, exist/does not exist, has/does not have, numerical, a description, or any combination thereof.);
In regards to:
deriving, by the one or more processors, predictor variables from the feature values which are most likely to be predictive of year built […];
training, by the one or more processors, a machine learning algorithm to determine a year built of a property using the predictor variables of the first set of properties and the year built […] different from the first set of properties
(¶ 12, 13, 14, 15, 17, 18, 103, 104, 106, 121, 162, 182, 184, 215 wherein machine learning is trained to determine a property parameter, such as, but not limited to, the square footage of a property using data mining of external sources, such as, but not limited to, property records, listings, photographic images (e.g., satellite/aerial, street view, videos), characteristics/features of the home (determining that a garage has been converted to a “Granny Flat” by determining that the home should have a garage and comparing this information to, for example, images to determine that the garage is no longer there, thereby increasing the livable square footage of the property), and etc. Rose further discloses that subsets of the feature values can be correlated with square footage by associating these two pieces of information with one another.;
¶ 12, 13, 14, 103, 104, 106, 119, 121, 154 wherein machine learning is trained to determine the square footage and qualitative build grade of a property using data mining of external sources, such as, but not limited to, property records, listings, photographic images (e.g., satellite/aerial, street view, videos), and etc.; ¶ 16, 215 wherein the system applies data from the aerial images to determine the street address of the property and to further determine the year it was built;
¶ 5, 16, 77, 88 wherein a property of interest is identified, for example, via its street address; ¶ 16, 116, 179 wherein a property is identified based on, at least, street address and is different from the set of properties that this property will be compared against, i.e. the system can be trained using semi-supervised labels that can be determined using a clustering technique to find properties similar to those flagged by previous human annotated labels and previous semi-supervised labels.; ¶ 124 wherein a machine learning model is used to determine if actual data (e.g., satellite images) match with records to determine if a renovation has been made of record, i.e. identifying an unknown or inaccurate property parameter;
NOTE: For the purposes of compact prosecution, the Examiner has provided an alternate interpretation in view of Lookabaugh in the event that the applicant does not agree with this analysis to teach that a machine learning model can be trained using data from other properties and applied to a subject property to compare specific common characteristics and allow the machine learning model to determine, estimate, calculate, or the like information not found in the subject property.);
identifying, by one or more processors, a subject property (¶ 16 wherein a property is identified based on, at least, street address; ¶ 5, 16, 77, 88 wherein a property of interest is identified, for example, via its street address; ¶ 16, 116, 179 wherein a property is identified based on, at least, street address and is different from the set of properties that this property will be compared against, i.e. the system can be trained using semi-supervised labels that can be determined using a clustering technique to find properties similar to those flagged by previous human annotated labels and previous semi-supervised labels.; ¶ 124 wherein a machine learning model is used to determine if actual data (e.g., satellite images) match with records to determine if a renovation has been made of record, i.e. identifying an unknown or inaccurate property parameter.
NOTE: For the purposes of compact prosecution, the Examiner has provided an alternate interpretation in view of Lookabaugh in the event that the applicant does not agree with this analysis to teach that a machine learning model can be trained using data from other properties and applied to a subject property to compare specific common characteristics and allow the machine learning model to determine, estimate, calculate, or the like information not found in the subject property.);
receiving, at the one or more processors, one or more aerial images of the subject property (¶ 16 wherein aerial images of the property are received);
extracting, by the one or more processors, feature values of the subject property using the same features used to train the machine learning algorithm, wherein at least one of the feature values is for the feature which is extracted from the one or more aerial images of the subject property (¶ 12, 16, 75, 77, 98, 121, 122, 215, 227 wherein a plurality of different features can be extracted from the aerial images, such as, but not limited to, square footage, renovations, additions that add square footage, and etc., which can further be obtained from other sources of information, such as, but not limited to, city records, MLS listings, social listings, and etc.
NOTE: For the purposes of compact prosecution, the Examiner has provided an alternate interpretation in view of Lookabaugh in the event that the applicant does not agree with this analysis to teach that a machine learning model can be trained using data from other properties and applied to a subject property to compare specific common characteristics and allow the machine learning model to determine, estimate, calculate, or the like information not found in the subject property.);
applying, by the one or more processors, the predictor variables from the feature values of the subject property to the trained machine learning algorithm to determine a year built of the subject property (¶ 12, 13, 14, 103, 104, 106, 119, 121, 154 wherein machine learning is trained to determine the square footage and qualitative build grade of a property using data mining of external sources, such as, but not limited to, property records, listings, photographic images (e.g., satellite/aerial, street view, videos), and etc.; ¶ 16, 215 wherein the system applies data from the aerial images to determine the street address of the property and to further determine the year it was built
NOTE: For the purposes of compact prosecution, the Examiner has provided an alternate interpretation in view of Lookabaugh in the event that the applicant does not agree with this analysis to teach that a machine learning model can be trained using data from other properties and applied to a subject property to compare specific common characteristics and allow the machine learning model to determine, estimate, calculate, or the like information not found in the subject property.);
applying, by the one or more processors, the predictor variables from the feature values of the subject property to the trained machine learning algorithm to determine a confidence level for the determination of the year built of the subject property (¶ 121 wherein the machine learning model is applied to determine a confidence of its results and, to improve upon its confidence, the system will search other sources of information, thereby increasing its confidence of the probability that the square footage has changed, i.e. a low confidence level results in a low probability that a renovation pertaining to a square footage has occurred, thereby having the machine learning model use other sources to increase its accuracy/confidence level. As stated by the PTAB, Rose discloses determining the square footage of a home using machine learning. The Examiner asserts that the probability of a renovation is a reflection of the machine learning’s confidence that a renovation has taken place due to its determination/calculation of the property’s square footage and reference to additional sources of information to increase its confidence of the property’s square footage. At, at least, ¶ 121 Rose discloses that to increase its confidence that a renovation has taken place, which, in turn, is a determination of the property’s square footage, the machine learning model can use, as a non-limiting example, street view and satellite images to determine/calculate the square footage of the home and, consequently, increase its confidence that a renovation has taken place due to an increase/decrease of the calculated square footage versus what is stored in the records, for example.; ¶ 12, 13, 14, 103, 104, 106, 119, 121, 154 wherein machine learning is trained to determine the square footage and qualitative build grade of a property using data mining of external sources, such as, but not limited to, property records, listings, photographic images (e.g., satellite/aerial, street view, videos), and etc.; ¶ 11, 13, 14, 15, 16, 70, 72, 75, 77, 78, 80, 121 wherein the machine learning algorithm, using the data is has gathered (e.g., aerial images), determines the probability of a renovation, as well as a change in the square footage and qualitative build grade of the property, with a level of confidence and learns over time in order to improve its prediction.; See also Fig. 18, 23; ¶ 7, 13, 88, 89, 121, 215, 218, 220, 227, 230, 233 regarding an example that supports the statements made above in that the system determines a confidence level and in response to an increase in confidence the probability that a change has occurred is also changed, i.e. the confidence level increases as more sources of information is used by the system to determine whether the probability should increase. ¶ 89 further discloses, “…the T3 threshold comprises a set threshold or a variable threshold, wherein flip probabilities above the T3 value are moderately indicative of a flip and require further evidence and/or analysis to increase certainty [i.e. confidence] of a flip before inspection…”, wherein hard documentary evidence is stored by the system and provided via a timeline (¶ 218)
NOTE: For the purposes of compact prosecution, the Examiner has provided an alternate interpretation in view of Lookabaugh in the event that the applicant does not agree with this analysis to teach that a machine learning model can be trained using data from other properties and applied to a subject property to compare specific common characteristics and allow the machine learning model to determine, estimate, calculate, or the like information not found in the subject property.);
In regards to:
generating, by the one or more processors, a profile of the subject property with the determined year built and confidence level for the determination of the year built; and
providing, by the one or more processors, the year built of the subject property […] for display
(Fig. 18, 23; ¶ 7, 88, 89, 119, 121, 215, 218, 220, 227, 230, 233 wherein the system generates and displays a profile of a subject property with the determined square footage, year built, and etc. that is based on the system’s confidence that a change has occurred, wherein the confidence is based on and increases in response to the system utilizing more and more external sources of information and/or simply having an inspector provide their findings by physically inspecting the property (which results in the system changing its initial confidence level that there is an X% probability of a change made to the property to a (for example)100% confidence that there is an X% probability of a change made to a property because building equipment, waste, and etc. has been witnessed). As another non-limiting example, the system determines a confidence level higher than 0% due to the system receiving an indication that a social media message with the description “Brand New Carport!” was tagged at a particular property, refers to the property’s profile that no previous history of such a renovation was made, and updates the property’s profile which prompts city officials to visit the property to determine that renovations have been made and issue a warning or fine for the unpermitted renovation, wherein the increase in confidence level also prompts the system to also determine the probability of a renovation having taken place. The system also compares its findings against a threshold and increases its confidence by searching through more external sources to determine if the probability of a renovation also increases. That is to say, the system searches and retrieves information from more and more external sources to increase its confidence level to determine whether the probability of a renovation will increase or decrease, wherein potential evidence is collected and where a timeline of unreported events with hard documentary evidence is provided).
Rose discloses a system and method for extrapolating unknown/not yet verified information about a home by referring to reliable/verifiable sources of information. Despite this, Rose fails to explicitly disclose whether it is old and well-known in the art to train a machine learning algorithm to use other sources of information in order to extrapolate this information, such as, referring to other homes. As an additional side note, in the event that the applicant disagrees that Rose does not determine or use year built information, the Examiner has also provided Lookabough to also teach this aspect of the invention, for the purposes of compact prosecution.
To be more specific, Rose fails to explicitly disclose:
obtaining, by one or more processors, (i) a first set of aerial images of a first set of properties, and (ii) an indication of a year built of each of the first set of properties
deriving, by the one or more processors, predictor variables from the feature values which are most likely to be predictive of year built by generating a visualization of the feature values to identify relationships between the feature values and the year built;
training, by the one or more processors, a machine learning algorithm to determine a year built of a property using the features values of the first set of properties and the year built of each of the first set of properties different from the first set of properties;
applying, by the one or more processors, the feature values of the subject property to the trained machine learning algorithm to determine a year built of the subject property. (i.e. the unknown or inaccurate property parameter)
applying, by the one or more processors, the feature values of the subject property to the trained machine learning algorithm to determine a confidence level for the determination of the year built of the subject property. (i.e. the unknown or inaccurate property parameter)
generating, by the one or more processors, a profile of the subject property with the determined year built and confidence level for the determination of the year built; and
providing, by the one or more processors, the year built of the subject property with the confidence level for the determination of the year built for display
However, Lookabough, which is also directed towards extrapolating unknown information about a home by referring to reliable sources of information, further teaches that other reliable well-known sources of information are other homes. Lookabough teaches that homes can have information that is common amongst themselves and, accordingly, it would have been obvious to compare one home with a similar (home) in order to extrapolate a particular piece of information for the home in question, such as, but not limited to, the year the home was built (see, at least, Page 30, Lines 3 – 24 wherein property attributes may be updated by identifying data form a source that reflects the property in cases which they disagree and wherein one of the many values that can be determined is the year that the property was built). Further still, Lookabough teaches that such processes can be performed by training a machine learning algorithm to learn what source of information can be referred to, as well as train the algorithm how to extrapolate the information.
With regards to “generating a visualization of the feature values to identify relationships between the feature values and the year built” ¶ 48 of the applicant’s specification discloses:
“For example, data may be visualized for the underlying relationships to determine which feature engineering steps should be assessed for performance improvement. This step may include manually entering user input, for example via user interface 123, which may include defining possible predictive variables for the machine learning model. Manual user input may also include manually including or excluding variables selection after running special feature engineering techniques. Manual user input may be guided by an interest to evaluate, for example, an interaction of two or more predictor variables (e.g., which data source the data came from)”.
With that said, see ¶ Page 30, Lines 3 – 24 discloses that the system is utilizes information that is relevant or more likely to provide a prediction to a particular outcome. As a non-limiting example, Lookabaugh discloses:
“For date values, such as ‘Last Sold Date’ or ‘Year Built’, the oldest, most recent, or some aggregate value can be taken, depending on the impact of the variable. For example, an investment criteria may specify that "flipped-homes" (those recently 20 purchased, remodeled, and often sold at a premium) are ineligible. In cases in which the "Last Sold Date" differs across sources, a conservative approach would be to use the most recent "Last Sold Date" when assessing this condition. The investment criteria could be evaluated with any new data to confirm the investment assessment measures.”
The Examiner asserts that this coincides with ¶ 50 of the applicant’s specification, which discloses:
“Variables may be evaluated in isolation to eliminate low predictive value variables, for example, by applying a cut-off value. For instance, some variables may have low predictive values for particular property parameters. For example, if a machine learning algorithm is being trained to determine a year built, it may be determined that variables such as property use information, kitchen size, and garage size have a low predictive value for determining the year built. Thus, in this example, the variables of property use information, kitchen size, and garage size may be eliminated during training so that the machine learning algorithm is wholly or partially trained without them. In another example, if a machine learning algorithm is being trained to determine a square footage, it may be determined that the variables of kitchen size, and garage size have a high predictive value for determining the square footage. Thus, in this second example, the variables of kitchen size, and garage size would not be eliminated while training the machine learning algorithm.”
(For additional support see: Page 16 Lines 22 – 32; Pages 17 – 18 Lines 23 – 1; Page 18 Lines 11 – 23; Page 19 Lines 11 – 17; Page 20 Lines 26 – 32; Page 21 Lines 26 – 37; Page 25 – 27 Lines 25 – 15; Page 30 Lines 3 – 24 wherein a plurality of properties within the vicinity of a subject property’s location are identified; Page 15 Lines 24 – 31; Pages 16 – 18 Lines 22 – 1; Page 18 Lines 11 – 23; Page 20 Lines 19 – 32; Page 21 Lines 26 – 37; Pages 26 – 27 Lines 25 – 15; Page 30 Lines 3 – 24; Page 36 – 37 Lines 34 – 8 wherein information of the plurality of properties is received)
More specifically, Lookabaugh teaches that a plurality of parameters about the subject property can be estimated based on analyzing a plurality of homes that share at least one common feature of the subject property that can be used to determine/predict a particular parameter for the subject property, i.e. the common feature has a relationship with the particular parameter as it is relevant and can be used for predicting the particular parameter. Lookabaugh teaches that this can be performed if information about the subject property is missing and, in order to fill in the missing information, the system estimates the particular parameter, which can be, but is not limited to, year built, square footage, and build grade, wherein the estimated parameter is a weighted average of the other properties’ particular parameter that is being estimated. Lookabaugh further teaches that that determination can be performed by using k-nearest neighbor, closest properties by Euclidean distance, common geographic region such as a town or county, latitude/longitude, nearby properties, and number of desired other homes to compare the subject property with.
(For support see: Pages 17 – 18 Lines 7 – 23; Page 19 Lines 11 – 25; Page 30 Lines 3 – 24; Pages 36 – 37 Lines 34 – 8)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include in the property analysis system and method of Rose additional sources of information, such as, other homes that share certain characteristics of a home in question, as taught by Lookabaugh, since providing as much information as possible about a particular home of interest provides home buyers who may have a low attention span and low tolerance for latency, but are interested with completing a real estate transaction, with rapid feedback pertaining to potential real estate transactions that can be critical in ensuring that would-be customer remain engaged, increase user experience, estimate the fair value of a home, and facilitate a customer’s decision making process (Pages 10 – 11 Lines 25 – 29; Page 12 Lines 5 – 33, Pages 14 – 15 Lines 32 – 7; Pages 15 Lines 32 – 21).
The combination of Rose and Lookabaugh discloses a system and method of utilizing a machine learning system that determines a confidence level that the probability of a renovation has taken place at a real estate property and provides evidence of its confidence to a user so that the system can convey to the user the probability of a particular parameter, attribute, characteristics, or the like of the property is present, such as, but not limited to, change in square footage or when a particular renovation/structure was built. Although this information and analysis is provided with the property’s profile so allow a user to view the property’s timeline with the documented evidence of when a renovation may have taken place and to determine whether a city official/inspector should be assigned to the property to confirm its findings, the combination of Rose and Lookabaugh fails to explicitly disclose whether it would have been obvious to one of ordinary skill in the art for a machine learning system to also display its confidence level to a particular finding.
To be more specific, the combination of Rose and Lookabaugh fails to explicitly disclose:
providing, by the one or more processors, the year built of the subject property with the confidence level for the determination of the year built for display
However, CaseGuard teaches:
“To give an example, an automatic transcription software program will provide a confidence level about the accuracy of the words that were said in the video or audio file. With these confidence levels, consumers can correct any mistakes or errors that may take place during the automation process. In this way, users of these software offerings can rest assured that they are putting out the highest quality of work at all times and avoiding any unnecessary errors.
Machine learning confidence levels provide users with a metric concerning the effectiveness of the automation process. This is typically done through percentages ranging from 0% to 100%. The higher the percentage of the confidence level, the more effective the automatic software program was in performing its specific function. These confidence levels can come in a range of structures or forms.
In the context of the automatic transcription process, this confidence level represents the software’s effectiveness in transcribing words from a given audio file. After transcribing an audio file, an automatic transcription software program will underline any words that have a low confidence level as they pertain to word detection, whether it be due to an accent or the volume of the speaker’s voice, and provide a specific accuracy based on percentages. Moreover, these underlined words will also have colors that correspond to specific accuracy ranges, (20%-40%, 40% –60%, 60%-80%, etc.) so that users can easily go back and correct any mistakes with as little time and effort as possible.
As using complex software can be both a confusing and daunting task, a machine learning’s confidence level is geared towards providing users with both a visual and analytical approach to achieving the highest standard of work in their respective professions or fields. Even if a word has a lower confidence level and has been underlined, users can still listen to the transcription to check if the software may have made a mistake.”
(Pages 2 – 3, 7)
In other words, one of ordinary skill in the art of machine learning would have found it obvious and motivated to display to a user the confidence value of a machine learning’s output to a user as this provides the user with additional insight and understanding of the machine learning’s output, regardless of whether output is with regards to square footage, year built, word usage, transcription, or etc. One of ordinary skill in the art of machine learning would have been motivated to adopt the teachings of CaseGuard as CaseGuard teaches that there is a benefit to displaying the confidence level of a machine learning’s output as this can provide the user reviewing the output with a sense of understanding, as well as whether there are any mistakes within the data, e.g., whether the hard evidence provided by the system of Rose is reliable or was properly analyzed by the system or whether human input, e.g., inspector input, is needed.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention that applying the known technique of displaying the confidence level for a machine learning’s output, as taught by CaseGuard, would have yielded predictable results and resulted in an improved system. It would have been recognized that applying the technique of CaseGuard to the teachings of the combination of Rose and Lookabaugh would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate the technique of displaying confidence levels for a machine learning’s output to a user. Further, applying this technique of CaseGuard to that of the combination of Rose and Lookabaugh would have been obvious by those of ordinary skill in the art as resulting in an improved system that would allow for a user to better understand the trustworthiness of the output and, in the case of the combination of Rose and Lookabaugh determine whether the prediction or estimations provided by the machine learning system can be relied upon and/or whether a human user may be needed to support the evidence and conclusions determined by the machine learning system.
In regards to claims 4, 10, 16, the combination of Rose, Lookabaugh, and CaseGuard discloses the computer-implemented method of claim 1 (the computer system of claim 7; the non-transitory computer-readable memory of claim 13), further comprising: applying, by the one or more processors, the feature values of the subject property to the trained machine learning algorithm to determine a garage size of the subject property (¶ 12, 16, 75, 163 wherein aerial images are used by the machine learning algorithm to determine if there has been a garage conversion made to the property, a change in the square footage of the property (which is also based on data gathered from the aforementioned plurality of data sources), and the frequency in which a garage door is opening and closing. As a result, the system is able to determine if the garage no longer exists because it has been converted to the room, thereby increasing the square footage of the home while decreasing the square footage of the home. Additionally, the system is also able to determine if an addition was made to the home and can further utilize aerial images and other data to determine street and driveway parking changes, e.g., determine that a property had vehicle parking originally then later determining that there is driveway parking, the determination of an increase in the size of the property, and the identification of a garage and its use. The system is configured to identify physical changes to the property.).
In regards to claims 5, 11, 17, the combination of Rose, Lookabaugh, and CaseGuard discloses the computer-implemented method of claim 1 (the computer system of claim 7; the non-transitory computer-readable memory of claim 13), further comprising: gathering measurement data of an exterior of a structure of the subject property based upon the aerial images (¶ 12, 75 wherein exterior images are used by the machine learning algorithm to determine if there has been a change in the square footage of the property (which is also based on data gathered from the aforementioned plurality of data sources)).
In regards to claims 6, 12, 18, the combination of Rose, Lookabaugh, and CaseGuard discloses the computer-implemented method of claim 1 (the computer system of claim 7; the non-transitory computer-readable memory of claim 15), further comprising:
receiving, at the one or more processors, home characteristic data for the subject property, wherein the features of the subject property include the home characteristic data
(¶ 12, 13, 14, 103, 104, 106, 119, 121, 215 wherein machine learning is trained to determine a property characteristic/feature, such as, but not limited to, square footage, year built, renovation information, and etc., of a property using data mining of external sources, such as, but not limited to, property records, listings, photographic images (e.g., satellite/aerial, street view, videos), and etc.).
Response to Arguments
Applicant's arguments filed 12/17/2025 have been fully considered but they are not persuasive.
Rejection under 35 USC 101
The rejection under 35 USC 101 has been maintained.
The applicant’s reliance on Example 39 is unpersuasive and the Examiner assert that the applicant has mischaracterized the facts of Example 39. The Examiner asserts that at no point does the USPTO state that “collecting a set of digital facial images from a database” is does not recite a judicial exemption. The applicant further mischaracterizes Example 39 by not taking into consideration the limitations as an ordered combination. There was no specific limitation or feature that the USPTO pointed to in order to demonstrate that Example 39 does not recite a judicial exemption and improperly compares Example 39 to the claimed invention.
Example 39 recites additional elements that improve technology, in this case, digital facial recognition, whereas, the claimed invention is not improving any technology, but reciting generic machine learning at a high level of generality and applying it to the abstract idea. The specification is not directed towards identifying an issue that arose in machine learning, resolving an issue that arose in machine learning, nor deeply rooted in machine learning or any technology, but directed towards real estate evaluation and assessment.
First, as was stated in the rejection, the claimed invention is directed towards steps that can be performed by humans, a human in their mind, and/or through the aid of pen and paper, e.g., having a human collect and look at a plurality of aerial images for a plurality of real estate properties, collect information that include an indication of the year built of those plurality of properties, look/observe the images to extract some information about the properties, identify and collect aerial images of a property of interest, look/observe the image of the property of interest, compare the information between the plurality of properties and property of interest to determine, guess, estimate, or the like the year built of the property of interest, creating a record/profile for the property, and displaying the results of the analysis along with its confidence level. In other words, the Examiner has provided an analysis that demonstrates that the claimed invention can, indeed, be performed by a human in their mind and/or with the aid of pen and paper (“Mental Processes”) and that it is also directed towards a marketing or sales activities or behaviors as it is recites real estate valuation (“Certain Methods of Organizing Human Activities”). This is further supported by ¶ 2 of the applicant’s specification.
Second, as stated above, Example 39 was not found eligible because digital facial image were collected, but because it was improving upon facial detection technology, which is a computer technology for identifying human faces in digital images. The background provides an explanation of issues that arose in the technology, provided a solution to resolve this technological issue, and is a technology that is deeply rooted in technology. The claimed invention, as previously discussed, is not directed towards such concepts. Additionally, because Example 39 did not claim mathematical relationships, formulas, calculations, or a fundamental economic practice, it was found that Example 3 did not recite a judicial exception. Moreover, Example 39 is deeply rooted in technology and directed towards improving the technology, thereby, not directed towards a mental process as the invention cannot be practically performed in the human mind, unlike the claimed invention and evidenced by the example provided by the Examiner demonstrating that the claimed invention can, indeed, be performed by a human in their mind and/or with the aid of pen and paper.
Third, the claimed invention also recites “Mathematical Concepts” because it recites, inter alia, that the machine learning model is trained to determine a confidence value, which is supported by ¶ 65, 66, 95, 96 of the applicant’s specification. That is to say, the confidence value, as defined by the specification, is a mathematical calculation that the machine learning model performs in order to determine, i.e. calculate, a level of accuracy between its prediction and known information, e.g. “90% confident in the accuracy of the home characteristics determined or estimated.”
Example 46 also does not apply to the claimed invention because, similar to Example 39, Example 46 is directed towards improving technology and reciting additional elements that when considered as an ordered combination results in not reciting a judicial exemption, which, as discussed above, the claimed invention is not. That is to say, Example 46 was not found eligible because it was “displaying the analysis results for the animal on the display”, but to the ordered combination of additional elements.
With that said, when reviewing the claimed invention to determine if the additional elements results in the claimed invention becoming eligible, the Examiner provided an explanation that it does not and further provides an analysis that training a machine learning model, as recited in the claimed invention, is insufficient to overcome the rejection. As a result, Ex Parte Desjardins does not apply because, as discussed above, unlike Ex Parte Desjardins, the claimed invention is not directed towards improving machine learning, resolving an issue that arose in machine learning, or deeply rooted in machine learning or other technology, but, again, reciting generic technology at a high level of generality and applying it to the abstract idea. Ex Parte Desjardins was not found eligible simply because machine learning or training machine learning was recited, but because Ex Parte Desjardins was directed towards improving machine learning technology and provided an explanation of issues that arose in machine learning and resolving those issues. Simply reciting what comprises the training data, which is directed towards what the applicant believes, in their mind, is the “best” data to use to determine the year in which a subject property is built, is not an improvement to machine learning technology. The data that the applicant believes is the “best” data is not directed towards improving the training process, which is supported by the applicant’s specification at ¶ 45, 46, where the applicant is relying on generic and known training techniques, i.e. the specification is not directed towards improving the training techniques, but reciting generic training techniques at a high level of generality and only using data that the applicant believes is the best data to use in the training dataset. The claimed invention describes information that is to be included in training data, describing information that is to be used to compare characteristics of a subject property with other properties, then generically recites training and applying machine learning to determine a desired output, i.e. year built and confidence level, and outputting/providing a result.
Therefore, the Examiner asserts that the claimed invention is not improving any technology, let alone machine learning techniques. The Examiner asserts that the use of machine learning has been recited at a high level of generality and applied to the abstract idea for the benefits that machine learning provides, i.e. faster, more efficient, and etc. The claimed invention is directed towards the collection and comparison of information and, based on a rule(s), identify options, in this case, having a human collect and look at a plurality of aerial images for a plurality of real estate properties, collect information that include an indication of the year built of those plurality of properties, look/observe the images to extract some information about the properties, identify and collect aerial images of a property of interest, look/observe the image of the property of interest, compare the information between the plurality of properties and property of interest to determine, guess, estimate, or the like the year built of the property of interest, creating a record/profile for the property, and displaying the results of the analysis along with its confidence level.
Rejection under 35 USC 103
The Examiner asserts that the applicant’s arguments are directed towards newly amended limitations and are, therefore, considered moot. However, the Examiner has responded to the newly submitted amendments, which the arguments are directed to, in the rejection above, thereby addressing the applicant’s arguments.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure can be found in the attached PTO-892 Notice of References Cited.
Shamiss et al. (US PGPub 2019/0370837 A1); Guggenmos et al. (US Patent 12,254,030 B1); Fagnan et al. (WO 2022/047011 A1); Guo et al. (US Patent 11,373,257 B1); Kasower (US PGPub 2015/0026014 A1) – which are directed towards analyzing real estate properties to determine, predict, and/or estimate one or more characteristics of a real estate property of interest
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 GERARDO ARAQUE JR whose telephone number is (571)272-3747. The examiner can normally be reached Monday - Friday 8-4:30.
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, Sarah Monfeldt can be reached at 571-270-1833. 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.
GERARDO ARAQUE JR
Primary Examiner
Art Unit 3629
/GERARDO ARAQUE JR/Primary Examiner, Art Unit 3629 1/14/2026