IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`In re Patent of:
`U.S. Patent No.:
`Issue Date:
`
`Appl. Serial No.:
`Filing Date:
`Title:
`
`Cheng et al.
`7,970,674
`June 28, 201 1
`
`Attorney Docket No. 30693-0090IPI
`
`11/347,024
`February 3, 2006
`AUTOMATICALLY DETERMINING A CURRENT VALUE FOR A
`
`REAL ESTATE PROPERTY, SUCH AS A HOME, THAT IS
`
`TAILORED TO INPUT FROM A HUMAN USER, SUCH AS ITS
`OWNER
`
`Mail Stop Patent Board
`Patent Trial and Appeal Board
`U.S. Patent and Trademark Office
`P.O. Box 1450
`
`Alexandria, VA 22313-1450
`
`DECLARATION OF JOHN KILPATRICK UNDER 37 C.F.R. § 1.132
`
`I, John Kilpatrick, declare as follows:
`
`I hold a Ph.D. in Real Estate Finance and am a certified (general) real estate
`1.
`appraiser in Washington State as well as in 49 of the 50 states.
`I head up Greenfield Advisors
`(formerly known as Mundy Associates), a 37 year-old real estate appraisal and consulting firm
`headquartered in Seattle. ' For more than three decades, our firm has been a leading authority on
`difficult real estate appraisal problems.
`
`I am the author or editor of four books on real estate and a contributing author to
`2.
`three others, most recently Brownfields: A Comprehensive Guide to Redeveloping
`I am
`Contaminated Property, Third Edition (2010), from the American Bar Association.
`presently or have recently been a consultant on complex real estate valuation issues to such
`clients as the Federal Housing Finance Authority, the U.S. General Services Administration, the
`U.S. Army, and numerous corporations, trusts, and private investors. At the invitation of the
`Japan Real Estate Institute, I co-authored the authoritative guide on the appraisal of
`environmentally impaired properties for use in that nation, published in the October 2003 issue
`ofthe Journal ofthe Japan Real Estate Institute.
`I am a member ofthe Editorial Board of The
`Appraisal Journal, the Editorial Board ofthe Journal of Sustainable Real Estate, and a reviewer
`for the Journal of Real Estate Research.
`I was a featured speaker on valuation methods at the
`2012 annual meeting ofthe Appraisal Institute, and will also be a featured speaker on this topic
`at the upcoming 2013 meetings.
`I am regularly invited to be a presenter on real estate valuation
`topics at meetings of various academic societies, such as the 2013 American Real Estate Society
`annual meetings, held last month.
`I have also been an invited lecturer on real estate valuation to
`the Asian Real Estate Society (twice) and at various universities and meetings in the United
`States, Canada, and Europe.
`
`56920-8901/1.li(jAL26969875.2
`
`Zillow 200] - Microstralcgy. Inc. v Zillnw. Inc
`. Case l|’R20l3-00034
`
`

`
`In 2004, I became one of a handful of appraisers in the United States to be
`3.
`designated as a Nationally Certified Appraisal Standards Instructor by the Appraisal Standards
`Board in Washington, D.C. Also in 2004, I was honored by my peers in the industry by being
`nominated for a seat on the Appraisal Qualifications Board and by being named a Member
`(later elevated to be a Fellow) of the Faculty of Valuation of the British Royal Institution of
`Chartered Surveyors.
`I am also a Fellow of the American Real Estate Society.
`I have been
`professionally engaged in real estate finance, appraisal, development, and teaching for the past
`three decades.
`I have been accepted as an expert witness in various state and federal courts, the
`federal court of claims, and the federal tax court on matters relating to real estate appraisal,
`appraisal standards, automated valuation models and other mass appraisal techniques, and other
`issues regarding real estate markets, finance, and economics.
`
`4.
`
`5.
`
`6.
`project.
`
`A copy of my C.V. is attached hereto as Exhibit A.
`
`I have no financial interest in either party or in the outcome of this proceeding.
`
`I am billing Zillow's counsel my standard hourly rate for my consulting on this
`
`Overview of Opinion
`
`I was asked to review the Decision of the Patent and Appeal Board, entered on
`7.
`April 2, 2013, in the above referenced matter ("the Decision"). In that regard, I have also
`reviewed the Zillow Patent, US 7,970,674 (the "'674 Patent"), the Dugan Patent, US 5,857,174
`(the "Dugan Patent"), and U.S. Patent Publication No. 2005/015465 ("the Kim Application").
`have reviewed these from the perspective of a real estate expert, not as a patent attorney or
`expert on patent law.
`I have been advised, however, as to the principles relating to patent
`validity, including the principles of anticipation and obviousness.
`In considering these issues, I
`have focused on how the terms and disclosures of the '674 Patent, the Dugan Patent, and the
`Kim Application would be interpreted by persons of skill in the art as of February 3, 2006, the
`filing date of the '674 patent.
`
`I
`
`8.
`
`The Decision asserts that there is a reasonable likelihood that the Petitioner would
`
`prevail on showing that certain claims of the '674 Patent are unpatentable due to the Dugan
`Patent alone or in combination with other patents (primarily the Kim Patent).
`I do not believe
`the Dugan Patent teaches the fundamental invention of the '674 Patent — allowing a property
`owner to have access to an automatic valuation model to refine or revise an existing valuation
`for his or her home, by inputting or adjusting aspects of information about the home (e.g., the
`number of bedrooms). The differences between Dugan and the '674 Patent can be summarized
`as follows: (1) Dugan utilizes a different meaning of"valuation" than used in the '674 Patent;
`(2) Dugan does not employ an "automatic" valuation model; and (3) Dugan does not provide the
`seller direct access to any valuation model (any revisions to input appear to be done with
`appraiser involvement). This last difference is significant, because general appraisal
`methodology has always involved the appraiser supervising the valuation process.
`In the '674
`Patent, a homeowner is able to revise an existing valuation without appraiser involvement.
`
`56920-890l/l.EGAL2696987S.2
`
`

`
`My Understanding ofPatent Principles
`
`I have
`I am not a patent lawyer, and do not have formal training in patent law.
`9.
`been advised, however, as to some of the principles relating to patent validity that are reflected
`in the Decision.
`I understand, for example, that patent validity is determined on a claim-by-
`clam basis. Some claims in a patent may include an introductory portion, referred to as the
`preamble, and then a body, which includes the claim elements.
`I understand the terms of the
`claim should be interpreted based on how a person of ordinary skill in the art would have
`understood them, in light of the patent disclosure, at the time the patent application was filed.
`With respect to the '674 Patent, for example, I understand that the terms should be interpreted as
`somebody of ordinary skill in the real estate appraisal art would have interpreted them as of
`February 3, 2006.
`I understand that the claim elements are presumed to constitute limitations
`on scope, and the preamble is presumed not to be a limitation, but may be a limitation where
`terms in the claim elements are introduced in the preamble. An independent claim stands on its
`own, while a dependent claim incorporates all of the limitations of any claim on which it
`depends.
`
`I understand that a claim is "anticipated" where all of the elements of the claim
`10.
`are disclosed in a single prior art reference.
`I understand that a claim may be invalid as obvious
`if any differences between the claim limitations and one or more prior art reference would have
`been obvious to a person of ordinary skill in the art at the time the application was filed. For
`my analysis, I have focused primarily on whether the Dugan Patent discloses or teaches the
`invention claimed in the '674 Patent.
`In doing so, I have relied on how a person of ordinary
`skill in the art in 2006 would have understood the '674 invention, in light of the claim terms and
`the disclosure in the patent specification.
`
`The '6 74 Patent
`
`The '674 Patent is directed to methods and computer-readable media that "procure
`1 1.
`information about a .
`.
`. property from its owner .
`.
`. to refine an automatic valuation of the .
`.
`.
`property .
`.
`. ." '674 Patent Abstract.
`
`The '674 Patent discloses an automatic valuation model that can generate
`12.
`valuations based on the characteristics of a subject property and sales data relating to
`comparable properties. Automatic valuation models are designed to work without appraiser
`involvement. While they can generate a value for a property, they are not considered a true
`"appraisal." An appraisal is conducted by an appraiser, while an automatic valuation model
`works "automatically" based upon a data set, without appraiser involvement. An automatic
`valuation can value multiple properties essentially simultaneously, while in an appraisal, an
`appraiser is focused on valuing a single “subject” property.
`
`The '674 Patent distinguishes appraisals from "automatic valuations [that are]
`13.
`. performed based upon the contents ofa public database, and without input from
`.
`generally .
`each home's owner or other information not in the public database." '674 Patent at 1:45-49.
`The '674 Patent model generates a value based upon a large set of sales data, through a
`regression analysis (i.e., a least-square error method) or a decision tree forest based model. See
`Figure 4B, steps 458-461, and col. 8:18-24 and 9:13-l4.
`
`56920-8901/LEGAL269698752
`
`_ 3 _
`
`

`
`Such "automatic" valuation models existed in the prior art. The '674 Patent is
`14.
`focused, however, on allowing a homeowner to refine a valuation from such an automatic
`model through the revision of the characteristics or attributes information about the subject
`property or, in other circumstances, by identifying more relevant comparable sales. There is no
`appraiser involvement.
`
`More specifically, the ‘674 Patent discloses the display of an initial value of a
`15.
`home to its owner, and then soliciting updated information concerning home attributes,
`improvements, and other factors that might affect the value of the home from the owner, and
`then displaying a refined valuation. See Figure 14. The patent explains that "by enabling a user
`to refine a valuation of his or her home based upon information about the home known to the
`user, the [invention] in many cases makes the valuation more accurate than otherwise be
`possible and/or helps the user to more fully accept the valuation as appropriate." '674 Patent at
`3:18-22.
`
`I understand that the Decision found claims 15 and 17 are anticipated by Dugan,
`16.
`that claims 2, 5-10, 13, 14, 16, 26, 27, 29-33, 35-37, 39 and 40 are obvious in light ofDugan
`and Kim, and claims 11, 12, 28, 34 and 38 are obvious in light of Dugan, Kim and at least one
`other reference. Claim 15 is directed to a method for refining an automatic valuation of a home,
`from a user knowledgeable about the distinguished home (e.g., the owner), that includes
`obtaining input adjusting at least one aspect of information about the home used in its valuation,
`and automatically determining a refined valuation based on the user input. Claim 17 depends
`on claim 15, and also involves displaying the refined valuation to a user other than the user
`providing the input.
`
`Claim 2 has similar limitations, but is directed to a computer readable medium
`17.
`that causes a computer system to perform the method, and also specifies that the system
`procures infonnation from the home owner to refine an automatic valuation, and displays to the
`owner the refined valuation.
`
`Claims 5-10 are dependent on claim 2, and describe the nature of the input from
`18.
`the home owner (e.g., altered property attributes, improvement, recent comparable sales).
`Claims 1 1 and 12 depend on claims 2 and 8, but also provide for the displaying ofa map that
`allows an owner to select or identify recent comparable sales.
`
`Other dependent claims have similar limitations, and include claims that are
`19.
`specifically focused on details relating to the identification ofcomparable sales. My report does
`not address every detail of the claims, but assesses broadly whether Dugan discloses the thrust
`ofthe '674 invention, and why someone of ordinary skill in the art would not look to Dugan to
`arrive at the '674 invention.
`8
`
`The Dugan Patent
`
`The Dugan Patent does not focus on "automatic" valuation models, much less
`20.
`letting a homeowner directly access such a model to revise an existing valuation. Instead,
`Dugan focuses, essentially, on providing a tool for the more conventional appraisal approach,
`where a single home is appraised by an appraiser based on comparable sales. Dugan also
`
`S6920-8901/LEGAL269698752
`
`_ 4 _
`
`

`
`focuses on a significantly different valuation — a valuation based on the individual preferences
`of a buyer or seller. A traditional sales comparison approach], as used in a traditional appraisal,
`starts by selecting a series of comparable sales (typically 3 or so for the appraisal of a single
`family residence). Since "comparable" does not mean "identical," there will typically be
`differences between each of the comparables and the subject property being appraised. These
`differences require adjustments to the sales price of the comparables. Good appraisal practice
`dictates the use of comparables with minimal adjustments. Fannie Mae, for example, requires
`that "gross" adjustments to any one comparable not exceed 25%, and "net" adjustments not
`exceed 15%2. After all adjustments are made, each comparable has an "adjusted sales price"
`which may be very different from the actual sales price. The appraiser will then weight these
`adjusted sales prices so that the sum of the weights equals 100%} While the Uniform Standards
`of Professional Appraisal Practice ("USPAP"), which can be found through this link --
`www.uspap.org -- is silent to the specific manner of determining weights, other authoritative
`guidance can be found in specific situations. For example, the IRS requires that the appraiser
`put the most weight on the comparable with the least adjustment4‘ Thus, the value determined
`by the sales comparison approach is the weighted average of the adjusted sales prices.
`
`Dugan discloses a method to combine both the adjustments as well as the
`21.
`weighting with an index which he refers to as his "Ideal Point System" ("IPS"). This IPS is at
`the heart of the Dugan patent; nothing like it appears in the '674 patent.
`
`The Dugan Patent mentions, in passing, that the operator may revise the record of
`22.
`either the subject property or one of the comparable properties (see step 50 and discussion at
`Column 8, line 19). Apparently, this revision can be to correct errors in the characteristics of
`the subject or comparable or to revise the IPS rankings. However, the option of revising
`characteristics data is clearly not an objective of his patent, as it is not listed among the
`objectives at Column 4. Indeed, the Dugan Patent does not depend on this option, as is noted in
`Column 14, Lines 44-54. Specifically, "[i]t is also recognized that the system 10 may be
`configured so that an appraiser is unable to modify records stored at a national database." At
`the least, Dugan does not disclose the owner revising the database records.
`
`Indeed, the core objective of the Dugan Patent is to be able to assign numbers to
`23.
`this IPS index. This index allows the Dugan patent to combine both the sales adjustment
`process along with the weighting process for a given subject property. He notes that the
`modification step 50 is where the IPS values are input and/or updated for the comparables and
`the subject property (see Column 8, lines 22-24). The crux ofthe Dugan Patent is that, using
`his index, the adjusted sales prices (156) ofthe comparables will vary depending on the subject
`property (see Column 12, lines 61-67 and Column 13, lines 1-3).
`
`Dugan notes that his appraisal will result in varying mathematical outcomes
`24.
`depending on who uses it. For example, for a given subject property with a fixed set of
`
`' Note that Dugan confusingly calls this the "Market Data Approach," a leftover from some older, outdated verbiage.
`USPAP uses the more accurate term "Sales Comparison Approach", since other approaches, such as Cost and
`Income, will also use "market data."
`2 Fannie Mae Selling Guide, June 30, 2010, Section B4-1.4-l7
`3 Note that even a simple averaging ofthe adjusted sales prices is a weighting, with all ofthe weights being equal.
`4 See IRS Publication 561
`
`56920-890]/LEGAL269698752
`
`_ 5 _
`
`

`
`characteristics, and a given set of comparables (also with a fixed set of characteristics), the
`buyer may have one set of preferences and the seller another, resulting in different IPS indexes
`and thus different values (see Column 13, lines 39-49).
`
`In short, the key element in the Dugan patent is the ability to assign an IPS value
`25.
`to each property, and use this weighted index to replace the traditional sales adjustment grid
`(where an appraiser weights comparable properties). Dugan’s approach produces value
`determinations that vary significantly between users having different preferences.
`
`A Summary of the Dtfierences Between Dugan and the '6 74 Patent
`
`The '674 Patent claims, including independent claim 15, can be distinguished
`26.
`from Dugan because: (1) they claim an automatic valuation model, which would be construed
`by somebody of ordinary skill in the art on February 3, 2006 as a model that can perfonn mass
`valuations using large data sets without any appraiser involvement; and (2) they recite obtaining
`user input to automatically determine a refined valuation, which would be understood by a
`person of ordinary skill in the art as allowing a user to input the data directly into the automatic
`(aka "automated") valuation model (AVM) (as opposed to simply reporting information to an
`appraiser working on a conventional appraisal).
`
`The most fundamental differences are that the '674 Patent discloses a system
`27.
`(referred to as a "facility") that allows a user to directly input information into an AVM. See,
`e.g., '674 Patent at 2:58-59 ("a software facility for automatically determining a current value
`for a home .
`.
`. that is tailored to input from its owner .
`.
`. ."); 3:64-66 ("[A] home's attributes
`may be inputted by a person familiar with them, such as the owner. .
`. ."). The patent expressly
`discloses that the invention "does not require the services ofa professional appraiser .
`.
`. ."
`'674 Patent at 4:21-22 (emphasis added).
`
`There is also no question that the ‘674 Patent is directed to an AVM. See, e.g.,
`28.
`'674 Patent at 4:48-49 (disclosing that the facility "automatically determine[s] current values for
`homes in a geographic area"); 5:52-57 ("such request may be individually issued by users, or
`issued by a program, such as a program that automatically requests valuations for all homes in
`the geographic area .
`.
`. .")
`
`The '674 Patent thus claims a very different invention than what is disclosed in
`29.
`Dugan, which is essentially a tool that an appraiser can use to identify and weight comparables.
`The '674 Patent does not create any sort of index, such as Dugan's IPS. The '674 Patent instead
`begins with a set of characteristics of the subject property. Then, it extracts from the nearest
`sales those which are the most comparable based on those salient characteristics. Dugan would
`weight these comparables (and also perform adjustments) using a weighting system based on
`preferences (e.g., buyer preferences, seller preferences, some other subjective preferences, or
`some market weighted set of preferences input by an appraiser and preferably based on buyer or
`seller preferences (see Column 10, lines 9-12 and 57-58)). Dugan notes that the selection ofa
`given set of preferences will influence the outcome, and that two different sets of preferences
`will result in two different sets oflPS indices, and hence two different mathematical results.
`
`56920-890 l /LEGAL269698752
`
`

`
`The '674 Patent, in contrast, discloses weighting the comparables via an objective
`30.
`method (see Figure 4B, steps 458-461, and Column 8, lines 18-24, and Column 9, lines 13-14).
`Since the Dugan method allows for greater human input, the accuracy of the characteristic data
`is not as important (note that the Dugan patent continues unabated even if the characteristic data
`base cannot be modified). However, the '674 Patent relies on objective statistical measurement,
`and thus, from an appraisal perspective, the accuracy of the characteristic data is more critical.
`As such, and unlike the Dugan patent, the '674 Patent very specifically provides for a
`mechanism where property owners (or other knowledgeable parties) may correct characteristic
`data if it is mistaken (see Figure 14 steps 1402-1407). The '674 patent also discloses an
`opportunity for the home owner (or other interested party) to suggest similar comparables and
`thus arrive at a refined valuation. This refined valuation follows a similar tree-and-leaf pattern,
`but with the addition (and added weighting) of the home owner-suggested comparables (see
`Col. 15, line 56, thru Col. 16, line 35).
`
`Further, Dugan never discloses a homeowner (or even a seller) inputting data into
`31.
`a model without appraiser involvement. Indeed, the appraiser guidelines provide that an
`appraiser must not allow a homeowner's or client's objectives to bias the valuation. 5 Appraisers
`do not and would not turn a computer model over to a homeowner to create an appraised value.
`Instead, while the appraiser may accept input from the homeowner, the appraisal guidelines
`require that the appraiser stay involved in the appraisal process. This is specifically relevant to
`the conclusion in the Decision (p. 17) that it "would have been obvious to substitute a seller for
`the buyer and appraiser who inputs information in Dugan, because it would make the appraisal
`more accurate, for example, by allowing the seller to fix errors, and thus make the appraisal
`more trustworthy .
`.
`. ." This, however, would not be obvious to one of skill in the art, because
`an appraiser must control the appraisal process. He or she would oversee all aspects of the
`appraisal, and would not simply allow the seller to revise data sets unilaterally.
`
`In addition, the '674 Patent and the Dugan Patent arrive at very different types of
`32.
`valuations, based on the very different methodologies and data selection criteria.
`In the '674
`Patent, the methodology performs a sales adjustment process which combines the data selection
`heuristic of a typical appraiser (using the node-and-leaf to arrive at the closest comparables) and
`a least-square-error criteria for weighting those comparables in a manner which suggests a
`hedonic regression model. Both of these are well within the market value universe, and as such
`the '674 Patent arrives at a value which would be recognized as a form of market value by most
`appraisers and other real estate professionals.
`
`Conversely, Dugan arrives at something very different. First, the Dugan patent
`33.
`output is a moving target based on the preference set suggested by the user. Since market value
`is a unique finding for a given subject property, the fact that a number of different values can
`come out of Dugan based on varying preference sets suggests something very different. Indeed,
`what Dugan provides may be called Investment Value, which is most commonly defined as the
`
`5 See, for example, the discussion of“Scope of Work Acceptability” and the impact of bias toward a client in
`USPAP, Scope of Work Rule, Lines 406-429.
`
`S6920-8901/LEGAL269698752
`
`_ 7 _
`
`

`
`. value of a specific property to a particular investoré." From an appraisal perspective, this is
`". .
`clearly something different from market value.
`
`The '674 Patent specifically discloses a model that simply utilizes public sales
`34.
`information, and regression or related techniques, to determine the market value of a home or
`homes. In fact, the patent emphasizes the automatic nature of the valuation, using available
`data. See, e.g., '674 Patent Columns 6-] 1.
`
`This is very different from the Dugan Patent, which does not disclose any mass
`35.
`valuation fimctionality, and never discloses a homeowner being able to directly change
`attributes so as to obtain a revised valuation. Although Dugan discloses the use of an Artificial
`Intelligence (AI) type model, 2:32 — 3:18, Dugan criticizes that approach because it does not
`"consider the values of the buyers or sellers in appraising either the comparable properties or the
`subject property." Id. at 3:21-23. This same criticism could apply, however, to the '674 Patent.
`The '674 Patent does not use the personal preferences of buyers and sellers to weight or revise
`the values and instead uses a system that “because [it] does not require the services of a
`professional appraiser .
`.
`. can typically determine a home’s value .
`.
`. in a manner generally free
`from subjective bias.” (See Column 4, lines 22-24.). For all of these reasons, I do not believe
`the Dugan Patent is similar to or discloses the '674 invention including the invention in the
`claims at issue.
`I expand on these distinctions below.
`
`The Dugan Patent Does Not Disclose an "Automatic " Valuation
`
`the traditional
`In property valuation, there are basically two kinds of analyses:
`36.
`"manual" appraisal (which is still most commonly used for individual property valuations) and
`the Automated (or "Automatic") Valuation Model, or AVM, which is commonly used for mass
`valuations or at least has the capability of valuing multiple properties simultaneously without
`appraiser involvement.7‘
`
`The distinguishing feature ofthe AVM is the use of computer algorithms to
`37.
`generate values in a statistical fashion based on a large data set of sales, without any need for an
`appraiser to identify comparable properties. For example, the most common tool used in an
`AVM is the ordinary least squares regression, which utilizes a large number of comparables, or
`"comps," which are sorted in some fashion (geographically, or by degree of comparability) and
`then used to value each ofthe subject properties. Other AVM techniques can include time-
`series models (such as are used by common indices, such as the Federal Housing Finance
`Authority model or the CoreLogic Case Shiller model), repeat sales indices, or other variations.
`At Greenfield we have an AVM which is a geographically weighted function of local tax
`assessment data.
`
`No AVM is 100% automated or "automatic." One important component of any
`38.
`AVM is data cleaning, which can be thought of as front—end quality control. Both the subject
`property characteristics data and the com parable property data (both characteristics and sales
`data) must be reasonably accurate for the AVM to function properly. USPAP has certain
`criteria for such model development and calibration in USPAP Standards Rule 6.
`
`6 Appraisal ofReal Estate, l3"‘, op cit., 28.
`7 In the tax assessment field, the AVM is often called the "Computer Assisted Mass Appraisal, or CAMA.
`
`56920-8901/LEGAL269698752
`
`_ 3 -
`
`

`
`At the other end of the spectrum is the traditional manual appraisal. Thirty years
`39.
`ago, when I first appraised a house, the process was almost 100% manual (save for the use of a
`pocket calculator, copy machine, and an electric typewriter).
`I inspected the property, used a
`tape measure and camera to record data, hand-calculated adjustments, typed the findings onto a
`report form, and affixed photographs and maps manually to the pages and even affixed small
`decals with arrows and words like "subject" and "comparable" to the map.
`
`Today, the "manual" appraisal is highly automated. The field research uses a
`40.
`laser measuring tool, a digital camera, a GPS for location and mapping, and often a digital voice
`recording device which obviates the need for the appraiser to write anything down (or even type
`onto the report form!). Data is digitally downloaded, calculations are performed by computer
`models, and the report itself is generated as a digital document (with a digital signature) rather
`than printed. Nonetheless, despite all of this automation assistance, the traditional manual
`appraisal continues to be a very different process and report from a "fully" automated AVM. No
`user of appraisal services would ever confuse the two.
`
`This is important because Dugan does not disclose an "automatic valuation" as is
`41.
`claimed in the '674 Patent. Instead, Dugan focuses on a conventional appraiser-focused
`approach to appraisal, which is very different from the automatic valuation model in the '674
`Patent. More specifically, the '674 Patent specifically contrasts conventional appraisals and
`automatic valuations, on the basis that "such automatic valuations were generally performed
`based upon the contents of a public database, without input from each home's owner or other
`information no in the public database." '674 Patent at Col. 1:45-49. The '674 Patent is, at
`bottom, an improvement on existing AVMs, not an improvement, as Dugan purports to be, on
`conventional appraisals.
`
`The Dugan and '6 74 Patent Are Directed to Different Valuations
`
`Another difference between the Dugan Patent and the '674 Patent is the type of
`42.
`value which results from the two different appraisal processes. For example, the USPAP
`differentiates between "type" and "definition" of values. In the appraisal profession, it is
`recognized that an appraisal analysis may result in any one (or even more) of many different
`types of value, and that often these different results may be materially different from one
`another. For example, "distressed sale value" which recognizes that a property may be sold in a
`hurry, may be very different from "market value" which typically assumes that the property is
`exposed for a market-normal period of time. Even "market value" itself may have varying
`definitions, depending on the situation, and the Appraisal Institute, in its highly-regarded text,
`The Appraisal of Real Estate 13"‘, discusses eight different types of value as examples, although
`not as an exhaustive list9.
`_
`
`From an appraisal perspective — and germane to this matter — the type and
`43.
`definition of value dictates such things as the methods used, the data collected, and the manner
`
`8 In response to the mandates ofthe 1989 Financial Institutions Reform, Recovery, and Enforcement Act, USPAP
`provides mandatory standards for the conduct of all Federally-related mortgage transactions, and adherence has been
`adopted as a matter oflicense law in all 50 states.
`9 The Appraisal of Real Estate 13"‘, (Chicago, The Appraisal Institute, 2008), 22-32. Earlier versions ofthis text
`listed even more types of value.
`
`56920-8901/LEGAL269698752
`
`_ 9 _
`
`

`
`of reconciliation. The same property examined utilizing two different types of value may arrive
`at two different quantitative estimates. For example, a "market value" appraisal of a single
`family residence for federal financing purposes will typically be driven by market sales
`comparables, using some variant of the sales comparison approach, and will then arrive at a
`conclusion which will often be described as the "most probable price." Conversely, a distressed
`sale value of that same residence may focus on the income approach, using market rental
`comparables, and will arrive at some estimate of value which is driven by what investors might
`pay in an auction scenario (the highest cash auction price, rather than the most probable market-
`exposed price with typical financing available).
`
`There is no one universally-accepted standard definition of market value. Even in
`44.
`federally-related mortgage financing transactions, there are slight differences in the definitions
`of market value used. Outside of the mortgage realm, the definitions of market value (or its
`ofien-used brother, "fair market value") may change drastically from one situation to another.
`Some definitions of value may cause the appraisal exercise to arrive at a higher value and some
`at a lower value, depending on the methodology, data, and reconciliation.
`
`The '674 Patent and the Dugan Patent are directed to different types of valuation.
`45.
`A person of ordinary skill in the art on February 3, 2006, would understand that the '674 Patent
`claims, when it refers to a "valuation" (or more precisely an "automatic valuation") refers to a
`market valuation based on a regression or similar process across a large data set. This market
`valuation is not based on individual preferences of the buyer or seller, but on the value
`generated by the data set and the model.
`
`In contrast, it is clear that Dugan is speaking of a valuation based on individual
`46.
`preferences (which may have no relationship to a market valuation). Indeed, Dugan admits that
`his IPS values "represent the values fo