DETAILED ACTION
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 .
Notice to Applicant
The following is a Final Office action. In response to Examiner’s Non-Final Rejection of 3/19/26, Applicant, on 6/18/26, amended claims. Claims 1-3, 5, 7, 10-12, 14-17, and 19-26 are pending in this application and have been rejected below.
Response to Amendment
Applicant’s amendments are acknowledged.
Examiner appreciates applicant’s remarks explaining the removed claim language of “input categories” was intended to be supported by FIG. 2A and a user selecting a “type of transaction” (e.g. equipment lease, equipment sale... financial loan). The 112a rejections are withdrawn in light of the amendments and Remarks (pages 10-11).
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 1-3, 5, 7, 10-12, 14-17, and 19-26 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without reciting significantly more.
Step One - First, pursuant to step 1 in MPEP 2106.03, the claim 1 is directed to a method which is a statutory category.
Step 2A, Prong One - MPEP 2106.04 - The claim 1 recites–
“A method of forecasting commercial financial transactions, the method comprising:
training… a first predictive algorithm on historical data from past financial transactions, industry data obtained from at least one third party and portfolio data from one or more users;
presenting, …, series of diagnostic questions to the user… based on skip logic and branching capabilities (Applicant’s [0036-0058] and FIG. 2A give example of user questioned “is the transaction for… Real Property?” and user can select “Real Property, Lease, Purchase” among other options);
receiving… answers to the series of diagnostic questions;
determining,… using a second predictive algorithm different from the first predictive algorithm, an identified transaction type based on the answers to the series of diagnostic questions, and recommending and providing access to a preconfigured analytic calculator corresponding to the identified transaction type, wherein the preconfigured analytic calculator comprises the first predictive algorithm tailored to the identified transaction type, and presents specific variables related to the identified transaction type (Applicant’s [0059] - analytic calculator for the scenario (e.g. Real Property, Lease- Renewal, etc) is provided; FIG. 2A, [0035] as published – equipment lease, equipment sale, software license, software purchase, real estate lease, financial loan; FIG. 2B, [0060] as published – variables can be : renewal term, rate change, cost, square footage);
receiving, …, a series of independent variables that represent attributes of a first financial transaction for a predetermined location, wherein the series of independent variables comprises the specific variables related to the identified transaction type;
scaling and normalizing… the series of independent variables;
assembling… the scaled and normalized series of independent variables;
applying… weightings to the assembled scaled and normalized series of independent variables ;
predicting…using the first predictive algorithm, a first value for the first financial transaction for the predetermined location by entering the weighted, assembled, scaled, and normalized series of independent variables into the predictive algorithm;
predicting…using the first predictive algorithm, a second value for a second financial transaction for the predetermined location using the first predictive algorithm, wherein the second financial transaction is different from the first financial transaction;
comparing… the first value for the first financial transaction to the second value for the second financial transaction;
coordinating displaying … the comparison between the first financial transaction and the second financial transaction.”
As drafted, this is, under its broadest reasonable interpretation, within the Abstract idea grouping of “certain methods of organizing human activity” (commercial or legal interactions –contracts or marketing or fundamental economic principles (determining financial values in future)) and/or “mathematical relationships” as here we training a business prediction algorithm from past financial transactions, industry data, and portfolio data (e.g. leases/loans in [0078 as published), using series of equations based on skip logic and branching capabilities (falling within “certain methods of organizing human activity”, Marketing, and Market Research, user survey question progressions), receiving answers from users on questions regarding transaction type has specific variables for the transaction type (e.g. FIG. 2A, [0035] as published – equipment lease, equipment sale, software license, software purchase, real estate lease, financial loan; FIG. 2B, [0060] as published – renewal term, rate change, cost, square footage), determining by a “second predictive algorithm”, an identified transaction type based on answers and recommending an appropriate analytic calculator with financial transactions based on the answers from the user (this appears to be FIG. 2A-B and [0033 as published “The analytic calculator of the present disclosure provides guidance to the user on modeling any given scenario and then presents preset options (e.g., an analytic calculator library (AC library)) based on algorithms and mathematical order of operations; [0096] as published “ a predictive algorithm tailored to a type of transaction can be used as a pre-trained algorithm for a similar transaction (e.g., real property sales can be used to train algorithms for real property leases)”); receiving variables that represent attributes of a first financial transaction fitting input categories (claim 3 says they can represent various things such as location, type of center, comparable assets); scaling and normalizing independent variables e(.g. [0087] as published – scale for “excellent” can be a score of 5; score for “above average” can be a score of 4); [0101] as published – transforming the series of independent variables, such that the features are within a specific range (scaling); weighting different variables (e.g. [0096] as published gives example of increasing or decreasing influence variable has on the prediction); to then predict a value for the financial transaction, are forecasting a first value for a financial transaction using first predictive algorithm and a second value ( [0104] as published a financial value), predict second value for second financial location, comparing the first value and second value (e.g. FIG. 2D-E, [0073-0075] as published – comparing Renewal 252 with Scenario 2, 254, and resulting financials, NPV (net present value), or dollars). In combination, the limitations present a series of questions that can follow flow of skipping and branching (as in a market survey), and give a user a financial calculator to evaluate using a financial prediction, the “type” of transaction they have given answers on (e.g. a specific financial instrument), and then providing two comparative financial predictions of value in the 2nd page of limitations. Regarding “training” here being “for predictive algorithm,” See also August 4, 2025 Kim Memo page 3, where “training, by the computer, the ANN based on the input data and a selected training algorithm… a backpropagation algorithm and a gradient descent algorithm” requires specific mathematical calculations by name, as here we have explicit “training a first predictive algorithm from past financial transactions, industry data and portfolio data… predicting by first predictive algorithm a first value by entering weighted, assembled, scaled, and normalized series of independent variables.” See also Updated July 2024 Subject Matter Eligibility Update, Example 47, claim 2; Example 48, claim 1 – series of mathematical calculations from a mixed speech signal, includes the “training”; here we use general training such as a “predictive algorithm” based on financial transactions in [0095-0096] as published. Accordingly, claim 1 is directed to an abstract idea because it is doing a series of mathematical calculations to make a financial prediction where a person’s answers result in predicting likely appropriate financial calculator. Notably, while a computer is added to the claim, addressed in Step 2A, prong two and step 2B below, Applicant’s specification [0095] as published discloses “machine learning” as only being one example/option as it states “a training data set can comprise a grouping of data points that can be used to train a predictive (e.g., machine learning) algorithm”. Should Applicant desire for claim 1 to require machine learning, it would need to be amended into the claim.
Step 2A, Prong Two - MPEP 2106.04 - This judicial exception is not integrated into a practical application. In particular, the claim 1 recites additional elements that are:
A method of forecasting commercial financial transactions, comprising:
training, by a computing system, a first predictive algorithm on historical data from past financial transactions, industry data obtained from at least one third party and portfolio data from one or more users;
presenting, by the computing system, series of diagnostic questions to the user using artificial intelligence based on skip logic and branching capabilities;
receiving, at the computing system from a user device, answers to the series of diagnostic questions;
…
[each limitation is “by the computing system” now];
…
coordinating displaying, by the computing system, a graphical user interface (GUI) comprising the comparison between the first financial transaction and the second financial transaction
The claim involves a computer, training, receiving from a “user device”, “artificial intelligence” just for the selection of questions for users following skip and branch logic/rules, and a GUI for displaying; and the claim is considered, when viewing the additional limitations individually or in combination, “apply it [the abstract idea] on a computer” (MPEP 2106.05f merely uses a computer as a tool to perform an abstract idea); and “Field of use” (MPEP 2106.05h- “field of use” for “training,” “artificial intelligence” for question selection for customers/users, and “GUI” display that is “by a computer”. Examiner further notes a sequence of questions for different financial loan options operates the same way in a manual business flow of a marketing survey as it does here, where it is merely “by artificial intelligence” computer for sequencing through some set of questions (e.g. three).
Applicant’s specification [0095] as published discloses “machine learning” as only being one example/option as it states “a training data set can comprise a grouping of data points that can be used to train a predictive (e.g., machine learning) algorithm”. To any extent Applicant amends “machine learning” to be required for the “training” and prediction, or the claim is interpreted in that manner, “training, by a computing system,” is considered “apply it [the abstract idea] on a computer” (MPEP 2106.05f –the claim involves a computer performing training and possibly “machine learning” to make financial prediction); “field of use” (MPEP 2106.05h).
Accordingly, the additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim also fails to recite any improvements to another technology or technical field, improvements to the functioning of the computer itself, use of a particular machine, effecting a transformation or reduction of a particular article to a different state or thing, and/or an additional element 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. See 84 Fed. Reg. 55. The claim is directed to an abstract idea.
Step 2B in MPEP 2106.05 - The claim does 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 a computing system, “artificial intelligence”, where “training” occurs to make a financial prediction; and displaying “a GUI” is treated as MPEP 2106.05(f) (Mere Instructions to Apply an Exception – “Thus, for example, claims that amount to nothing more than an instruction to apply the abstract idea using a generic computer do not render an abstract idea eligible.” Alice Corp., 134 S. Ct. at 235); and MPEP 2106.05h (field of use)). Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
In addition, at step 2B, “receiving at the computing system from a user device,” answers to the questions is considered a “conventional computer function” (See MPEP 2106.05d(II) Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321).
The claim fails to recite any improvements to another technology or technical field, improvements to the functioning of the computer itself, use of a particular machine, effecting a transformation or reduction of a particular article to a different state or thing, adding unconventional steps that confine the claim to a particular useful application, and/or meaningful limitations beyond generally linking the use of an abstract idea to a particular environment. See 84 Fed. Reg. 55. The claim is not patent eligible. Viewed individually or as a whole, these additional claim element(s) do not provide meaningful limitation(s) to transform the abstract idea into a patent eligible application of the abstract idea such that the claim(s) amounts to significantly more than the abstract idea itself.
Independent claim 10 is directed to an apparatus at step 1, which is a statutory category. Claim 10 recites similar limitations as claim 1 and is rejected for the same reasons at step 2a, prong one, 2a, prong 2, and step 2b. The additional limitations, of processor, memory including instructions causing a computer to perform functions, are all part of “apply it on a computer” (MPEP 2106.05f) at step 2a, prong 2 and step 2b. The claim is not patent eligible.
Independent claim 19 is directed to an apparatus at step 1, which is a statutory category. Claim 19 recites similar limitations as claim 1 and claim 10 and is rejected for the same reasons at step 2a, prong one, 2a, prong 2, and step 2b. The additional limitations, of “computing system”, are part of “apply it on a computer” (MPEP 2106.05f) at step 2a, prong 2 and step 2b. The claim is not patent eligible.
Claim 2, 11, 20 narrow the abstract idea by giving various mathematical algorithms that are used.
Claims 3, 12, 21 narrow the abstract idea by stating descriptions of what variables represent (e.g. location, asset, lease, landlord, etc).
Claims 5, 14, 22 narrow the abstract idea by stating a mathematical operation of making a composite variable from combining other variables.
Claims 15, 23 narrow the abstract idea by stating the mathematical operations are repeated for different transactions/portfolios.
Claims 7, 16, 24 narrow the abstract idea by stating net present value is predicted for a plurality of scenarios.
Claims 17, 25 narrows the abstract idea by stating that the score is based on a “comparable”/similar financial transaction.
Claim 26 requires: “a) converting transaction data into vector format [EITHER before transaction labeled OR before transaction fed into algorithm], b) by concatenating transaction elements/variables, c) training is EITHER labeled, unlabeled, or a mixture. Portion a), b), c) falls within directed to an abstract idea and further narrowing the abstract idea, as this is a mathematical relationship of converting transaction data (e.g. dollars, cost) into a vector (a mathematical value), where a person “labels” training data to confirm it represents value for a past financial loan. To the extent this is “for training” and “before” labeling/ fed into algorithm, this is considered “apply it [the abstract idea] on a computer” (MPEP 2106.05f –the claim involves a computer performing training and possibly “machine learning” to make financial prediction); “field of use” (MPEP 2106.05h). Claim 26 is a method claim that has contingent limitations in d). The method claim states that the training data in c) is EITHER labeled, unlabeled, or a mixture. Accordingly, the last limitation (d) regarding the alternative of “unlabeled test data” is not required at this time. Even if it was, this would not help with eligibility as it is not directed to improving the machine learning itself. Rather, the claim is just scoring how accurate the financial prediction is; and this is the same as in the specification [0095] - “In some embodiments, unlabeled training data can be referred to as a test data set. In these embodiments, all or a portion of a test data set can be fed into a trained predictive algorithm to determine whether the training produced an accurate algorithm.” The contingent limitation [even if positively recited] is just used for scoring accuracy, which is related to the financial elements of a loan (or other financial type) being assessed. see also [0031, 0033] as published “The actual value of the financial elements can be difficult to assess accurately. For example, if a person is faced with buying an automobile, they may be faced with the decision around a purchase or a lease… The predictive analytical model for financial transactions (also referred to herein as an analytic calculator), as described herein, allows a user to model a specific financial transaction and to create various scenarios to determine the outcomes of any given set of variables.”
Therefore, the claim(s) are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
For more information on 101 rejections, see MPEP 2106.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-3, 7, 10-12, 15-17, 19-21, and 23-26 are rejected under 35 U.S.C. 103 as being unpatentable over Packes (US 2015/0317640) Lyons (US 20110184884, and Zurick (US 20190295163).
Concerning claim 1, Packes discloses:
A method of forecasting commercial financial transactions, the method (Packes –See par 18-20 - REP (“Real Estate Evaluating Platform”) may be utilized to predict pricing of real estate or predict value of buildings, neighborhoods, whole market, commercial spaces, or offices spaces), comprising:
training, by a computing system (Packes – See par 88 - The REP coordinator facilitates the operation of the REP via a computer system (e.g., one or more cloud computing systems); See par 89 – REP Coordinator includes a processor 801 that executes program instructions), a first predictive algorithm on historical data from past financial transactions, industry data obtained from at least one third party… (Packes – see par 35 - reducing property value variation associated with time (e.g., inflation, housing market trends, etc.) in historical data may lead to better results when training and/or retraining a neural network to estimate property value based on differences in attribute values. Historical data may be obtained for a specified estimation time frame (e.g., the last year) for properties that have data regarding property values during the estimation time frame (e.g., properties that were sold during the last year and have a selling price, properties whose property values were evaluated during a previous preparation iteration, etc.). The obtained historical data may be sliced for each estimation time period (e.g., for each month during the last year)... Properties comparable (e.g., based on similarity of attribute values) to the best performing subset of the data set may be evaluated using the first set of neural networks to estimate property values for the time period (e.g., for the month) associated with the slice; With a good representation (e.g., based on the estimated missing sale values) in time, the prediction system using recurrent neuronal networks, for example, may be trained to predict the price evolution for the future.)
Packes does not appear to consider the next part of the limitation regarding portfolio data.
Lyons discloses:
training, by a computing system, a first predictive algorithm on historical data from past financial transactions, industry data obtained from at least one third party “and portfolio data” from one or more users (Applicant’s specification [0078] as published – Using the specific values of that scenario, the user can now calculate the financial provisions, along with other applicable provisions (as desired) against other portfolios of similar leases. With additional reference to FIG. 3, this may consist of a data set from the user’s own lease portfolio.
Lyons discloses the limitations based on broadest reasonable interpretation in light of the specification – see par 16 - The predictive models are trained on historical data derived from a plurality of mortgage account profiles for a plurality of mortgages within the mortgage portfolio. See par 43 - The system 100 includes a source of historical data 118 that relates to a mortgage portfolio. A mortgage portfolio generally will include a multiplicity of mortgages held by a lender).
Packes discloses estimating value of real estate properties from user information (par 33), and correcting property values related to economic data as needed (See par 121). Lyons discloses modifying loans (see par 42) and collecting additional data including credit card data, demographic data (See par 118).
Zurick discloses:
presenting, by the computing system, series of diagnostic questions to the user using artificial intelligence based on skip logic and branching capabilities (Zurick – see par 41 - Natural language processing (“NLP”) is a sub-field of artificial intelligence that is focused on enabling computers to understand and process human languages, to get computers closer to a human-level understanding of language. The Counseling Dialog Rules include a series of if/then statements and other rules that regulate conversations created using a chatbot ; see par 42- platform of the invention connects with borrowers using a proprietary chatbot; See par 89 - FIGS. 4 and 5 are exemplary flowcharts (that include branching, and not including every question) of conversation rules for generating a chatbot conversation with a borrower assessing the borrowers need and eligibility for various options. While these flowcharts appear fairly straightforward, they are particularly designed in a manner that simplifies the conversation into short questions requiring short answers while also obtaining all of the information necessary to assess a borrower's eligibility for a wide variety of delay, deferment and forgiveness programs.);
receiving, at the computing system from a user device, answers to the series of diagnostic questions (Zurick – see par 52 - The chatbot advises the borrower of available repayment options, and records and transmits the borrower's preferences and/or selections, all in real time with no processing delays. A borrower's immediate needs are identified through a series of questions posed by the chatbot and answered by the borrower. The chatbot provides guidance and assistance to help resolve any issues the borrower may have meeting his or her current repayment obligations. The chatbot updates borrower contact information, provides loan balance and payment information, offers assistance to users having difficulty repaying, and guides a borrower to his or her best option to meet his/her situation. )
Lyons and Zurick disclose:
determining, by the computing using a second predictive algorithm different from the first predictive algorithm, an identified transaction type based on the answers to the series of diagnostic questions, and recommending and providing access to a preconfigured analytic calculator corresponding to the identified transaction type, wherein the preconfigured analytic calculator comprises the first predictive algorithm tailored to the identified transaction type, and presents specific variables related to the identified transaction type (Applicant’s [0059] - analytic calculator for the scenario (e.g. Real Property, Lease- Renewal, etc) is provided; FIG. 2A, [0035] as published – equipment lease, equipment sale, software license, software purchase, real estate lease, financial loan; FIG. 2B, [0060] as published – variables can be : renewal term, rate change, cost, square footage; ; [0096] as published “ a predictive algorithm tailored to a type of transaction can be used as a pre-trained algorithm for a similar transaction (e.g., real property sales can be used to train algorithms for real property leases).
Zurick discloses the limitations based on broadest reasonable interpretation in light of the specification – see par 52 - The chatbot advises the borrower of available repayment options, and records and transmits the borrower's preferences and/or selections, all in real time with no processing delays. A borrower's immediate needs are identified through a series of questions posed by the chatbot and answered by the borrower. The chatbot updates borrower contact information, provides loan balance and payment information, offers assistance to users having difficulty repaying, and guides a borrower to his or her best option to meet his/her situation. This includes obtaining information from the borrower which is used to determine what options are available to reduce or delay monthly payments and evaluating whether the borrower qualifies for loan forgiveness or other options. see par 89, FIGS. 3-5 - While these flowcharts appear fairly straightforward, they are particularly designed in a manner that simplifies the conversation into short questions requiring short answers while also obtaining all of the information necessary to assess a borrower's eligibility for a wide variety of delay, deferment and forgiveness programs. providing correct information to a borrower so that he or she may select the best available option for repayment.
see also Lyons discloses the limitations based on broadest reasonable interpretation in light of the specification – See par 174 - FIG. 9 shows four screen shots 910, 920, 930, 940 with equations and corresponding to the logic of the calculations for various metrics of interest (e.g. “Foreclosure, Monthly Payment, Future Home Value, and/or NPV = Revenue – Loss – Restructuring Expense”). For example one of the screen-shots 940 shows a Standardized lifetime value calculation for a mortgage, an action based predictor model (action effect model) built as a component of the subject matter described herein and a lender model (model built by the lender) that are part of the network of models forming the decision model; see par 179-183 – Net Present Value of Profit; see par 177 - The implemented code of original monthly payment and the new monthly payment after loan modification are both based on this formula, with some adjustment. see table after par 178 – e.g. Monthly Interest Rate; New Monthly payment before and after modification; see par 222, FIG. 13 shows treatment mix scenarios and the values of some of the metrics of interest corresponding to a number of constrained and unconstrained optimization runs. FIG. 13 is a screenshot 1300 of a report depicting six optimization runs 1310, 1312, 1314, 1316, 1318, 1320. One of the optimization runs 1310 is an unconstrained optimization. The other five optimization runs 1312, 1314, 1316, 1318, 1320 are constrained. The columns associated with each of the optimization runs 1310, 1312, 1314, 1316, 1318, 1320 show details as to how the treatment mix is distributed for each of these optimization runs and what are the values of some of the key metrics for each of these optimization runs.);
Packes, Lyons, and Zurick disclose:
receiving, at the computing system (Packes –See par 88 - The REP coordinator facilitates the operation of the REP via a computer system (e.g., one or more cloud computing systems); See par 89 – REP Coordinator includes a processor 801 that executes program instructions; see par 99 - Instructions for performing these processes may also be embodied as machine- or computer-readable code recorded on a machine- or computer-readable medium. In some embodiments, the computer-readable medium may be a non-transitory computer-readable medium. Examples of such a non-transitory computer-readable medium include, but are not limited to, a read only memory, … and a data storage device), a series of independent variables that represent attributes of a first financial transaction for a predetermined location, wherein the series of independent variables comprises the specific variables related to the identified transaction type (Packes – See par 18 - The REP may be utilized to predict the pricing of (e.g., urban) real estate, both at the time of the inquiry, and into the foreseeable future. Existing pricing schemes are geared to the horizontal modes of development in suburban and rural real estate markets and are inaccurate in multi-family and hi-rise development markets such as exist in cities all around the world. In some embodiments, the REP may be utilized to predict the value of individual apartment units for rental and sale, to predict the value of buildings, of neighborhoods, and/or of the whole market (e.g., as defined by any borough or boroughs with multifamily development). In some embodiments, the REP may be utilized to predict the value of commercial and office spaces (e.g., in vertical, or hi-rise, development structures). See par 31 - An attribute set selection may be obtained at step 105 of process 100. In one embodiment, real estate properties may have different attributes based on the unit type. For example, a condominium may have different attributes compared with a commercial unit (See Table after par 31 – e.g. “city, zip, state, neighborhood”)),
scaling and normalizing, by the computing system, the series of independent variables ([0087] as published – scale for “excellent” can be a score of 5; score for “above average” can be a score of 4); [0101] as published – transforming the series of independent variables, such that the features are within a specific range (scaling)). Packes discloses the limitations based on broadest reasonable interpretation in light of the specification – See par 38 - In one implementation, attribute values may be normalized. For example, numerical values may be converted to a 0 to 1 interval, where 1 is equivalent to the biggest original value and 0 is equivalent to the smallest original value.);
assembling, by the computing system, the scaled and normalized series of independent variables (Packes – See par 38 - In one implementation, attribute values may be normalized. For example, numerical values may be converted to a 0 to 1 interval, where 1 is equivalent to the biggest original value and 0 is equivalent to the smallest original value; see par 57, FIG. 4 - FIG. 4 shows a logic flow diagram illustrating a process 400 for estimating value (e.g., using a real estate value estimating (RVE) component) in accordance with some embodiments of the REP. FIG. 4 provides an example of how a set of neural networks may be used to estimate the value (e.g., property price, rental price, etc.) of a real estate property. In one implementation, the user may specify attribute values for any of the attributes discussed with regard to step 105 of process 100.);
applying, by the computing system, weightings to the assembled scaled and normalized series of independent variables ([0096] as published states “a weight of one or more variables can be adjusted. In this way, the influence the one or more variables have on a prediction can be increased or decreased.” Packes discloses the limitations based on broadest reasonable interpretation in light of the specification – see par 32 - a grouping process may be employed to group multiple attributes (e.g., attributes with a low importance factor) to create a single new attribute with a higher importance factor or importance index value). Attributes with higher importance or weight may be used by the REP in priority over other attributes when training a particular type of neural network for a particular use. If, for example, the REP is to create a neural network using only 5 inputs, the REP may be operative to select or receive a selection of 5 inputs with the highest importance factor for that neural network type. Such a limitation of the number of inputs may be dictated, for example, by any suitable information provided by a user in an estimation process enabled by the REP);
predicting, by the computing system, using the first predictive algorithm, a first value for the first financial transaction for the predetermined location by entering the weighted, assembled, scaled, and normalized series of independent variables into the first predictive algorithm (Packes – See par 33 - The goal of the training process may be to teach a neural network for pattern recognition. The best performances may be achieved with neural networks that may be specialized in the recognition of a limited number of patterns (e.g., a limited variation of the sale price). The unit localization may be one of, if not the, most important factors in sale price variation. By grouping and limiting the number of neighborhoods, the sale price for the units can be limited to a smaller range. For best training performances, the groups may be kept as small as possible. See par 61 – value of property may be estimated at step 417, 421; attribute values may be converted into numerical values and/or normalized prior to providing to neural network; see par 62 - When there are no more neural networks to utilize (e.g., as determined at step 413), then the overall result given by the neural networks in the selected set of neural networks may be calculated at step 425 of process 400. For example, the overall result may be displayed to the user. In one embodiment, the overall estimated property value may be calculated as the average of estimated property values from each of the neural networks in the selected set of neural networks (e.g., an average of each property value estimated at step 421)).
Packes discloses that it can consider economic data including interest rates (See par 34) and training neural networks on mortgage rates (See par 121).
Lyons discloses:
predicting, by the computing system, using the first predictive algorithm, a second value for a second financial transaction for the predetermined location using the first predictive algorithm, wherein the second financial transaction is different from the first financial transaction (Lyons – See FIG. 2, par 120 – decision model 200 is summation of analytical relationships being modeled, connecting lender actions and total portfolio; see par 124 – metric to optimize can be NPV (net present value); see par 126 – actions depicted can be Interest Reduction, Principal Reduction, and Term extension ; Refinance (with different treatments here depending on refinance terms)… or Change of Type of Loan (e.g. from ARM to conventional));
comparing, by the computing system, the first value for the first financial transaction to the second value for the second financial transaction (Lyons – See par 153 - FIG. 5 shows a screen shot of a treatment editor 600 for deriving treatments in a case when only three actions are considered: Principal Reduction 610, Term Extension 620 and Interest Rate Reduction 630. As shown in FIG. 5, various combinations of parameters for treatments are determined. Each combination of different parameters (namely different values for principal reduction, interest reduction, and term extension) is given a unique ID 640; see par 211 – optimization runs corresponding to different optimization scenarios are completed; See par 222-2223 - Different runs with global constraints are compared in order to converge on the preferred situation that will be implemented as the optimized policy for the mortgage portfolio as shown in FIG. 14).
Packes discloses having a graphical user interface (See Par 30, FIG. 3, par 73, FIG. 6A) for estimating value of a property (See par 73, FIG. 6A).
Lyons discloses a GUI with the comparison from the previous limitations:
coordinating displaying a graphical user interface (GUI) comprising the comparison between the first financial transaction and the second financial transaction (Lyons – See par 153 - FIG. 5 shows a screen shot of a treatment editor 600 for deriving treatments in a case when only three actions are considered: Principal Reduction 610, Term Extension 620 and Interest Rate Reduction 630; See also FIG. 9 – showing different analyses on transactions (e.g. NPV, Monthly Payment, etc) and FIG. 13 – showing different scenarios).
Packes, and Lyons are analogous art as they are directed to analyzing financials related to real estate property (see Packes Abstract, par 18; Lyons Abstract). Packes, Lyons, and Zurick are analogous art as they are directed to analyzing financials related to financial loans (see Packes Abstract, par 71; Lyons Abstract, par 42; Zurick Abstract). 1) Packes discloses looking at historical data and comparable properties when estimating property values (See par 35, 78). Packes discloses that it can consider economic data including interest rates (See par 34) and training neural networks on mortgage rates (See par 121). Lyons improves upon Packes by disclosing analyzing a portfolio of mortgages (See par 16, 43); disclosing analyzing different treatments in changes to loans or refinancing terms (see par 120-126), evaluating different combinations of parameters for treatments and using logic of calculations for various metrics of interest including NPV (See par 153, 174-183, FIG. 5, FIG. 9, FIG. 13) and displaying different GUIs showing comparisons of treatments (See FIG. 2, FIG. 5, 13-15). One of ordinary skill in the art would be motivated to further include a portfolio of mortgages; and include a portfolio of property sites with various loan modifications being altered and using logic of calculations for various metrics of interest including NPV to efficiently improve upon the estimated values of property by neural networks in Packes. 2) Packes discloses estimating value of real estate properties from user information (par 33), and correcting property values related to economic data as needed (See par 121). Lyons discloses collecting additional data including credit card data, demographic data (See par 118). Zurick improves upon Packes and Lyons by disclosing having series of information requests that can skip or branch off by a chatbot for obtaining information for the loans. One of ordinary skill in the art would be motivated to use chatbots for guiding borrowers to best options for their situation for reducing monthly payments to efficiently improve upon the estimated values of property by neural networks in Packes and the disclosure of refinancing loans in Lyons (par 42).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the predictions for real estate in Packes to further include analyzing portfolios of properties and consider modifying loans as disclosed in Lyons, and to further evaluate different loan modifications and using logic of calculations for various metrics of interest including NPV as disclosed in Lyons, since the claimed invention is merely a combination of old elements, and in combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable and there is a reasonable expectation of success.
Concerning independent claim 10, Packes and Lyons and Zurick disclose:
A device for forecasting commercial financial transaction scores (Packes –See par 18-20 - REP (“Real Estate Evaluating Platform”) may be utilized to predict pricing of real estate or predict value of buildings, neighborhoods, whole market, commercial spaces, or offices spaces), comprising:
at least one processor (Packes – See par 88 - The REP coordinator facilitates the operation of the REP via a computer system (e.g., one or more cloud computing systems); See par 89 – REP Coordinator includes a processor 801 that executes program instructions); and
a memory, coupled to the at least one processor, the memory including instructions causing the at least one processor to: (Packes See par 89, FIG. 8 – REP Coordinator includes a processor 801 that executes program instructions; processor 801 is coupled to memory 820 and storage device 819; see par 99 - Instructions for performing these processes may also be embodied as machine- or computer-readable code recorded on a machine- or computer-readable medium. In some embodiments, the computer-readable medium may be a non-transitory computer-readable medium. Examples of such a non-transitory computer-readable medium include, but are not limited to, a read only memory, … and a data storage device).
The remaining limitations are similar to claim 1 above.
Claim 10 is rejected for the same reasons.
It would be obvious to combine Packes and Lyons and Zurick for the same reasons as claim 1.
Concerning independent claim 19, Packes and Lyons and Zurick disclose:
An apparatus for forecasting commercial financial transaction scores, the apparatus configured to (Packes –See par 18-20 - REP (“Real Estate Evaluating Platform”) may be utilized to predict pricing of real estate or predict value of buildings, neighborhoods, whole market, commercial spaces, or offices spaces; See par 88 - The REP coordinator facilitates the operation of the REP via a computer system (e.g., one or more cloud computing systems)), comprising:
receiving, at a computing system (Packes –See par 88 - The REP coordinator facilitates the operation of the REP via a computer system (e.g., one or more cloud computing systems); See par 89 – REP Coordinator includes a processor 801 that executes program instructions; see par 99 - Instructions for performing these processes may also be embodied as machine- or computer-readable code recorded on a machine- or computer-readable medium. In some embodiments, the computer-readable medium may be a non-transitory computer-readable medium. Examples of such a non-transitory computer-readable medium include, but are not limited to, a read only memory, … and a data storage device).
The remaining limitations are similar to claim 1 above.
Claim 19 is rejected for the same reasons.
It would be obvious to combine Packes and Lyons and Zurick for the same reasons as claim 1.
Concerning claims 2, 11, and 20, Packes discloses “the obtained training method parameters may include the number of neural networks (e.g., 10 neural networks to create initially, 5 best performing neural networks to select for further analysis and/or retraining, etc.) for the set of neural networks (e.g., as may be described below with respect to step 141).” Packes also discloses having a feedforward neural network to provide an estimated output (e.g. an estimated value for a real estate property) (See par 121). However, Packes does not explicitly recite one of the alternative algorithms recited.
Lyons discloses:
The method of claim 1, wherein the predictive algorithm comprises at least one of ordinary least squares, random forests, and decision trees (Lyons – see par 37-38, FIGS. 17-18 – strategy decision trees; See par 234 - A classification tree can be built to express exactly what the optimized actions for each account are. To this end, the decision keys will be used as predictors (independent variables in the classification tree) and the optimal treatment that is known at this point will be used as a categorical dependent (response) variable; see par 236 - FIG. 17 as a strategy tree 1700. The terminal nodes, such as node 1710, of this strategy tree 1700 are treatments corresponding to an optimized strategy. The split points, such as split point 1720, of this tree 1700 are decision keys (e.g., FICO Score, current LTV, etc). see par 237 - The tree in FIG. 17 is a real life example of a tree that corresponds to an optimized strategy defined by an optimization scenario with the following settings: [0238] Objective: maximize NPV of the portfolio [0239] Treatments: combination of principal reduction, interest reduction, and term extension).
It would be obvious to combine Packes and Lyons for the same reasons as claim 1. Packes discloses “the obtained training method parameters may include the number of neural networks (e.g., 10 neural networks to create initially, 5 best performing neural networks to select for further analysis and/or retraining, etc.) for the set of neural networks (e.g., as may be described below with respect to step 141).” Packes also discloses having a feedforward neural network to provide an estimated output (e.g. an estimated value for a real estate property) (See par 121). Lyons improves upon Packes by disclosing using decision trees with decision keys used as predictors in the tree (See par 17-18, 234-239, FIG. 17). One of ordinary skill in the art would be motivated to further include decision trees to efficiently improve upon the estimated values of property by neural networks in Packes.
Concerning claims 3 and 12 and 21, Packes discloses
The method of claim 1, wherein the series of independent variables comprises at least one of market, trade area, location, type of center, asset, lease, landlord, comparable assets, negotiator, and negotiation strategy (limitations in the alternative -Packes See par 31 - An attribute set selection may be obtained at step 105 of process 100. In one embodiment, real estate properties may have different attributes based on the unit type. For example, a condominium may have different attributes compared with a commercial unit (See Table after par 31 – e.g. “city, zip, state, neighborhood”) – disclosing alternative of “location”).
Concerning claims 7 and 16 and 24, Packes discloses looking at property price, rental price, future pricing, direction of the market (See par 84).
Lyons discloses:
The method of claim 1, wherein the predicting the first value comprises predicting net present value for the first financial transaction for a plurality of scenarios (Lyons – see par 54 - Many of the predictive models, such as 210 and 220, predict intermediate results which are used to predict the overall metric. In this particular example implementation, the overall metric 230 is the net present value (NPV) of the mortgage portfolio. see par 204-205 - In selecting the settings for a given scenario of FIG. 10 (showing 15 different scenarios), the mortgage portfolio manager will consider various mixes of treatments, objectives and constraints as shown in FIG. 11. More specifically, FIG. 11 shows three screenshots 1110, 1120, 1130 that present selectable options to a user, such as a Mortgage Portfolio Manager. A screenshot is one presentation associated with a graphical user interface associated with the subject matter described herein. Screenshot 1110 presents four objective functions from which one is selected. In screenshot 1110, the objective function (NPV) of the optimization is chosen or selected as depicted by the highlighting).
It would be obvious to combine Packes and Lyons for the same reasons as claim 1. In addition, Packes discloses looking at property price, rental price, future pricing, direction of the market (See par 84). Lyons improves upon Packes by disclosing net present value calculation that includes scenarios. One of ordinary skill in the art would be motivated to further include net present value calculations comprising scenarios to efficiently improve upon the estimated values of property by neural networks and the future pricing (see par 84) in Packes.
Concerning claims 15 and 23, Packes discloses:
The device of claim 11, wherein the method is repeated for a series of first commercial financial transactions (Packes – see par 19 - the REP may be utilized to predict the value of individual apartment units for rental and sale, to predict the value of buildings, of neighborhoods, and/or of the whole market (e.g., as defined by any borough or boroughs with multifamily development). See par 31 – attribute set includes estimated values of each unit in a multi-unit building; see par 66 - the appropriate set of neural networks may be selected based on the unit type (e.g., one set of neural networks may be used to predict the value for a condominium, another set of neural networks may be used to predict the value for a commercial unit, and another set of neural networks may be used to predict the value for a multi-unit building;
See also Lyons – see par 215-218 - FIG. 12 is a tornado diagram 1200 where each bar corresponds to a variable (e.g., X.sub.1 through X.sub.6) and represents the range of change of an objective's value V (e.g., NPV) resulting from that variable's variation in a specified domain between a minimum and a maximum value; entire process may then be repeated until the analyst and the decision maker are satisfied with the final model.)
It would be obvious to combine Packes and Catalano and Lyons for the same reasons as claim 1.
Concerning claims 17 and 25, Packes discloses:
The method of claim 1, wherein the first value is based on comparable financial transactions (Packes – see par 35 - For each slice of the obtained historical data, a first set of neural networks may be generated (e.g., using the NNG component and process 100) using the slice as a data set, and the best performing subset (e.g., a strict or proper subset) of the data set may be determined (e.g., 10% of records for which the first set of neural networks gives the smallest output error (e.g., at step 141)). Properties comparable (e.g., based on similarity of attribute values) to the best performing subset of the data set may be evaluated using the first set of neural networks to estimate property values for the time period (e.g., for the month) associated with the slice).
Concerning claim 26, Packes, Lyons, and Zurick disclose:
The method of claim 1, wherein the training of the first predictive algorithm comprises converting transaction data into vector format before the transaction data is labeled or before the transaction data is fed into the first predictive algorithm by concatenating one or more transaction elements or one or more variables (Packes – see par 71 (page 14, col. 1) -Training mechanism for neural network can be summaries as follows:… 2) the external input (x(t) . . . , x(t-d)) at instant t and the neurons context activations at instant t may be concatenated to determine the input vector u(t) to the network, which may be propagated towards the output of the network), wherein a training data set for the first predictive algorithm comprises labeled data, unlabeled data, or a mixture thereof (Packes – see par 33 - The training of the neural networks may be a supervised process. see par 66 - Based on the set of unit characteristics of attributes used for training a network may thus be specialized for one or more specific types of units. A neural network may be trained in a supervised process using as a baseline for output one of the unit characteristics, such as price or days on the market.), and wherein at least a portion of an unlabeled test data set is fed into the trained first predictive algorithm to determine whether the training produced an accurate first predictive algorithm (Claim 26 is a method claim that has contingent limitations in d). The method claim states that the training data in c) is EITHER labeled, unlabeled, or a mixture. Accordingly, the last limitation (d) regarding the alternative of “unlabeled test data” is not required at this time.
For purposes of compact prosecution, art is still applied: Packes – see par 120 - at step 906, process 900 may determine if the results of the testing of step 904 are acceptable. For example, as described above with respect to FIG. 1, test results and/or the performance of a neural network may be analyzed (e.g., at step 141 and/or step 153 and/or step 169) to determine if the neural network is acceptable for use (e.g., if the average of the percentage testing errors of all records is less than a threshold amount).).
Claims 5 and 14 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Packes (US 2015/0317640) and Lyons (US 2011/0184884) and Zurick (US 20190295163), as applied above to claims 1-3, 7, 10-12, 15-17, 19-21, and 23-26, and further in view of Cozine US 2018/0225593.
Concerning claim 5 and 14 and 22, Packes discloses:
The method of claim 1, wherein:
the method further comprises:
generating a composite score of the first financial transaction by combining one or more of the series of independent variables, (Packes– see par 32 - a grouping process may be employed to group multiple attributes (e.g., attributes with a low importance factor) to create a single new attribute with a higher importance factor or importance index value; For example, by combining multiple low importance factor "sports" amenity attributes, such as three sports amenities for "swimming pool?", "gym?", and "track?" into a single high importance factor "combined sports amenities" attribute, the value of such a high importance factor grouped or combined attribute may be operative to reflect the properties of each low importance factor attribute of the group (e.g., if each of the three importance factor weight attribute's property was "yes", the value of the grouped attribute may be a 9 (e.g., the highest value), whereas if none of three low importance factor attribute's property was "yes", the value of the grouped attribute may be a 0 (e.g., the lowest value); See par 59 - the appropriate set of neural networks may be selected based on the type of value desired (e.g., one set of neural networks may be used to predict property prices, while another set of neural networks may be used to predict rental prices)).
Packes discloses repeating portions of the training process (See par 32), but does not appear to scale, normalize, weight the composite score as best understood.
Cozine discloses:
scaling and normalizing the composite score (the claim appears to only be supported by [0105] as published stating: “n an example, the series of independent variables may include, but is not limited to at least one composite variable of multiple variables.” Cozine discloses the limitations based on broadest reasonable interpretation in light of the specification – See par 51 - one or more servers of the automated modeling system can aggregate data feeds (e.g., the property records received via various data connections) from various data sources into main property attribute, transaction and location data structure 340. This aggregation can include a series of steps for combining different property records having different format into a single database schema including the normalizing of data from all providers into that schema. See par 59 – segmentation 375 involves grouping geographic data objects, relative to certain attributes (e.g. property type, price quartile, etc); See par 64 - Since it is impractical to value more that 80 million properties daily, IHI 387 may be used to time shift stored AVM valuations between valuation dates as the system continuously cycles through segments and refreshes valuations of all properties in every segment as the system cycles through them. see par 74 - In some embodiments, the statistics are normalized across all Census Tracts after weighting within Census Tracts, if required. Then, Euclidean distances between all possible pairs of Census Tracts in each county are computed for all counties computing the distances from the weighted and normalized Census statistics. The method can be used with numerical attributes in any number of dimensions from any source and some sources other than US Census Bureau statistics may be used in accordance with various embodiments. );
assembling the scaled and normalized composite score (Cozine – See par 64 - Since it is impractical to value more that 80 million properties daily, IHI 387 may be used to time shift stored AVM valuations between valuation dates as the system continuously cycles through segments and refreshes valuations of all properties in every segment as the system cycles through them. see par 65 - At any rate, in various embodiments, the clustering/segmentation process 375 always creates segments with a sufficient number of transactions to support a robust model; see par 74 - In some embodiments, the statistics are normalized across all Census Tracts after weighting within Census Tracts, if required.);
applying weightings to the scaled and normalized composite score (Cozine – see par 65 - create robust training sets for KARL 390. In some embodiments, KARL 390 is an AVM which produces valuation estimates and attribute weights. See par 66 - To improve the accuracy of comparable pricing within the automatic ES application 391, the values of comparables is time shifted if necessary using the IHI 387 index for the segment. In addition, KARL 390 produces attribute weightings that quantify the relative importance of property attributes within each segment which ES 391 uses to more accurately adjust comparable properties for attribute differences compared to the subject property.); and
applying weightings to the assembled, scaled, and normalized composite score (Cozine – see par 65 - create robust training sets for KARL 390. In some embodiments, KARL 390 is an AVM which produces valuation estimates and attribute weights. See par 66 - In addition, KARL 390 produces attribute weightings that quantify the relative importance of property attributes within each segment which ES 391 uses to more accurately adjust comparable properties for attribute differences compared to the subject property.); and
predicting the first value for the first financial transaction comprises predicting the first value for the first financial transaction by entering the weighted, assembled, scaled, and normalized series of independent variables and the weighted, assembled, scaled, and normalized composite score into the first predictive algorithm (Cozine – See FIG. 3 par 60 - ES 391 is a computer-executed algorithm using “Comparable Sales Methodologies” that infer the value of a subject property by referring to transaction values for nearly identical properties. See par 66 - To improve the accuracy of comparable pricing within the automatic ES application 391, the values of comparables is time shifted if necessary using the IHI 387 index for the segment. In addition, KARL 390 produces attribute weightings that quantify the relative importance of property attributes within each segment which ES 391 uses to more accurately adjust comparable properties for attribute differences compared to the subject property).
Packes, Lyons, Zurick, and Cozine are analogous art as they are directed to analyzing financials (see Packes Abstract, par 18; Lyons Abstract, par 42; Zurick Abstract; See Cozine – predict price of property). Packes discloses repeating portions of the training process (See par 32), and combining multiple factors into a “grouped or combined attribute.” Cozine improves upon Packes, Zurick, and Lyons by disclosing aggregating and combining records and normalizing records (See par 51), pulling data from certain dates (a form of scaling), and weighting property attributes that quantity the relative importance within a “segment” or composite of data (see par 66). One of ordinary skill in the art would be motivated to further include scaling, normalizing, and weighting for segmented/composite data to efficiently improve upon the estimated values of property by neural networks in Packes.
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the predictions for real estate in Packes to further include analyzing portfolios of properties and to further evaluate different loan modifications as disclosed in Lyons, to further include chatbots for guiding users in Zurick, and further include scaling, normalizing, and weightings for composite/segmented data as disclosed in Cozine, since the claimed invention is merely a combination of old elements, and in combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable and there is a reasonable expectation of success.
Response to Arguments
Applicant's arguments filed 6/18/26 have been fully considered but they are not persuasive and/or are moot in view of the new rejections.
With regards to the 101 rejection, Applicant argues that the claims only “involve” the abstract idea and are not “directed to an abstract idea” similar to Example 39. Remarks, pages 15-16. In response, Examiner respectfully disagrees. Examiner also notes that this is not similar to Abstract Idea Example 39, in the January 2019 Guidance, was an improvement with specific steps of how the “training” of a neural network occurred, along with a disclosure discussing the technical issues with the image analysis that was occurring. See MPEP 2106.04(a)(1)(vii) where details related to a two-stage training system are present – e.g. “training the neural network in a first stage using the first training set, creating a second training set including digital non-facial images that are incorrectly detected as facial images in the first stage of training; and training the neural network in a second stage using the second training set.”. See also 2019 Revised Patent Subject Matter Eligibility Guidance, available at https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility-examination-guidance-date , slide 105 detailing how Example 39 is for improving facial detection; slide 108 Explaining what Applicant invented is to address false positives in expanded training set by performing iterative training algorithm, to provide a “robust face detection model that can detect faces in distorted images while limiting the number of false positives.” In contrast here, we only have “training, by a computing system, a first predictive algorithm” in claim 1 and “determining, by the computing system using a second predictive algorithm different from the first predictive algorithm”, such that the 1st and predictive algorithms are for predicting values of financial transactions – for comparative financial values that are predicted (where “second predictive algorithm” is based on [0096] as published – real property sales for real property leases; and Applicant’s example [0074] as published and FIG. 2D – two scenarios of financial predictions compared). We do not have a similar situation here of improving a technical process of image facial detection, and the arguments are not persuasive. See also MPEP 2106.04(d)(1) “If the specification explicitly sets forth an improvement but in a conclusory manner (i.e., a bare assertion of an improvement without the detail necessary to be apparent to a person of ordinary skill in the art), the examiner should not determine the claim improves technology.” Applicant’s claims also recite explicit mathematical calculations throughout the claims for making financial value predictions.
Applicant argues the claims cannot be practically performed in the human mind. Remarks, pages 16-17. In response, Examiner respectfully disagrees. First, other abstract idea groupings (Certain Methods of Organizing Human Activity; and Mathematical relationships) were applied. Second, Applicant’s argument is based on “unpractically large amount of data” from new limitations of questions presented to user based on skip logic and branching algorithms” as well as other limitations. Remarks, page 17. Examiner respectfully disagrees. Requiring a computer alone does not make the claim eligible. See MPEP 2106.04(a)(2)(III)(C) “An example of a case in which a computer was used as a tool to perform a mental process is…The patentee in Mortgage Grader claimed a computer-implemented system for enabling borrowers to anonymously shop for loan packages offered by a plurality of lenders, comprising a database that stores loan package data from the lenders, and a computer system providing an interface and a grading module.” In addition, there is no requirement of “large” amounts of data, and the additional elements have no restriction on number of questions. Applicant then argues a number of aspects of the specification and/or new claim 26. Remarks, page 18-19. These aspects are not required by the claim, and even claim 26 has many alternatives. Arguments are moot over the new rejections necessitated by the amendments.
Applicant argues with respect to step 2a, prong 2 that there is an improvement in computer functionality since the claim now is similar to Enfish, Finjan, and Core Wireless because the claim recites particular computer-implemented architecture with question flow using artificial intelligence based on skip logic and branching capabilities, and a second prediction along with access to an analytic calculator for the identified transaction type. Remarks, pages 19-21. In response, Examiner respectfully disagrees. First, this is moot in view of the new rejection. Second, this is not persuasive as this is just displaying the result of the abstract idea of financial analysis using math “on a computer display”; where “question sequences” operate in the same way manually as when they do “by artificial intelligence”. Examiner further notes a sequence of questions for different financial loan options operates the same way in a manual business flow as it does here where it is merely “by artificial intelligence” computer for sequencing through some set of questions (e.g. three). [0041] as filed (0059 as published), argued by Applicant, states “Upon answering the complete set of questions, the calculator 200 may then recommend the appropriate preconfigured solution for the specific transaction that is contemplated. For example, in response to a user making the above-mentioned example selections (i.e., real property, lease, renewal, and lessee), the calculator 200 may provide the single scenario analytic calculator 202 for a real property lease (see FIG. 2B).” Merely providing the appropriate calculator for the appropriate financial type deemed relevant (here a “renewal”) further supports the limitations are , individually or as a whole, “apply it [abstract idea] on a computer” (MPEP 2106.05f) and “field of use” (MPEP 2106.05h). MPEP 2106 summarizes Core Wireless decision as “An improved user interface for electronic devices that displays an application summary of unlaunched applications, where the particular data in the summary is selectable by a user to launch the respective application. Core Wireless Licensing S.A.R.L., v. LG Electronics, Inc., 880 F.3d 1356. By displaying only a limited list of common functions and data from which to choose, the invention spared users from time-consuming operations of navigating to, opening up, and then navigating within, each separate application. Id. The invention thus increased the efficiency with which users could navigate through various views and windows. Id. The claims here are not similar to Core Wireless.
Enfish was eligible because, as stated in MPEP 2106.05(a) “In Enfish, the court evaluated the patent eligibility of claims related to a self-referential database. Id. The court concluded the claims were not directed to an abstract idea, but rather an improvement to computer functionality. Id. It was the specification’s discussion of the prior art and how the invention improved the way the computer stores and retrieves data in memory in combination with the specific data structure recited in the claims that demonstrated eligibility. 822 F.3d at 1339, 118 USPQ2d at 1691.” There is nothing similar here in this claim.
Finjan was eligible at step 2A for “virus scan that generates a security profile identifying both hostile and potentially hostile operations.” Finjan also states that a behavior-based virus scan constitutes an improvement in computer functionality because, “in contrast to traditional “code-matching” systems, which simply look for the presence of known viruses, “behavior-based” scans can analyze a downloadable’s code and determine whether it performs potentially dangerous or unwanted operations” and the profile approach enables more flexible and nuanced virus filtering. See Slip opinion at pages 6-7. There is nothing similar here in this claim.
Applicant argues the claims improve another technology similar to McRO, Diehr, and Thales decisions, because here there is a “specific” improvement in artificial intelligence and selecting an appropriate financial calculator based on going “beyond simple input/output calculator tools by leveraging other data sets, comparing modeled outcomes to historical outcomes, presenting preset options based on algorithms and mathematical order of operations, comparing multiple contracts, assessing portfolios of contracts of a specific transaction type, and saving calculator results to structured modules for further analysis.” Remarks, page 23-25, quoting Specification [0033]. In response, Examiner respectfully disagrees. First, eligibility based on 101 is not simply whether any “specific” limitations are recited in the claim – it needs to be a particular solution to “improve a computer or other technology.” Rather, McRo, as explained in MPEP 2106.05(a)(II)(“Improvements to Any Other Technology of Technical Field”), had a specific way to solve the problem of producing accurate and realistic lip synchronization and facial expressions in animated characters. As stated in MPEP 2106.05(a)(I)(“Improvement to Computer Functionality”),
in computer-related technologies, the examiner should determine whether the claim purports to improve computer capabilities or, instead, invokes computers merely as a tool. Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016).” See MPEP 2106.05(a)(I). Second, the quotation of much of [0033] of the specification has many aspects not required by the claims. Here there are no technical details on the training, or what is in the calculator other than “presents specific variables”. MPEP 2106.05(a)(II) stated Diehr was eligible because “Particular computerized method of operating a rubber molding press, e.g., a modification of conventional rubber-molding processes to utilize a thermocouple inside the mold to constantly monitor the temperature and thus reduce under- and over-curing problems common in the art.” This is not the situation here. MPEP 2106.05(a)(II) stated Thales was eligible because “vii. Particular configuration of inertial sensors and a particular method of using the raw data from the sensors.” There is nothing similar here – there is not even a first sensor for physical measurements. See also MPEP 2106.05(a)(I) example not sufficient to show improvement in computer functionality - Accelerating a process of analyzing audit log data when the increased speed comes solely from the capabilities of a general-purpose computer, FairWarning IP, LLC v. Iatric Sys., 839 F.3d 1089.
Applicant then argues the claims are now similar to Desjardins, as the claim now recites “presenting, by the computing system, a series of diagnostic questions to the user using artificial intelligence based on skip logic and branching capabilities.” Remarks, pages 25-27. In response, Examiner respectfully disagrees. First, argument is moot over revised rejection. Second, these limitations pertain to a user making a couple selections – e.g. Real Property, Lease, Renewal as in Applicant’s examples [0036-0059] as published and FIG. 2A; if Applicant selects Real Property, Lease, Renewal, they are given access to calculator with variables related to answered transaction type. Second, MPEP 2106.04(a) explains “ if the specification explicitly sets forth an improvement but in a conclusory manner (i.e., a bare assertion of an improvement without the detail necessary to be apparent to a person of ordinary skill in the art), the examiner should not determine the claim improves technology.” Here, just providing the financial calculator (e.g. Renewal) based on the selection, is not viewed as an improvement to artificial intelligence. There are not any further technical details required in the claim, or reflected in the claim from the specification, on the skip logic and capabilities of branching questions, and/or the artificial intelligence. Applicant then argues the training data being labeled, unlabeled OR a mixture. However, these aspects are not claimed and a number of aspects are from the specification and/or new claim 26. Remarks, page 27-28. These aspects are not required by the claim, and even claim 26 has many alternatives. Arguments are moot over the new rejections necessitated by the amendments.
Applicant argue that all the claim limitations result in “significantly more”. Remarks pages 29-30. In response, Examiner respectfully disagrees. Many limitations Applicant points to are part of the “directed to an abstract idea”; other limitations are part of “apply it [abstract idea] on a computer” (MPEP 2106.05f) as discussed in the revised rejection.
Applicant then appears to argue “conventional” evidence is needed to have a 101 rejection here for many of the limitations. Remarks, pages 30-34. Examiner respectfully disagrees. With regards to step 2B, only those additional elements (analyzed under 2B) that are deemed “conventional” need to comply with Berkheimer. When elements are just part of “apply it” [abstract idea] on a computer, under MPEP 2106.05(f) or “field of use” (MPEP 2106.05h), no evidence is needed.
Applicant’s arguments with respect to 103 are moot in view of the revised rejection.
Conclusion
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 IVAN R GOLDBERG whose telephone number is (571)270-7949. The examiner can normally be reached 830AM - 430PM.
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, Anita Coupe can be reached at 571-270-3614. 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.
/IVAN R GOLDBERG/Primary Examiner, Art Unit 3619
Prosecution Insights