(12) United States Patent
`Cheng et al.
`
`(10) Patent N0.:
`
`(45) Date of Patent:
`
`US 7,970,674 B2
`Jun. 28, 2011
`
`US007970674B2
`
`(54) AUTOMATICALLY DETERNIININGA
`CURRENT VALUE FOR A REAL ESTATE
`
`PROPERTY, SUCH AS A HOME, THAT IS
`TAILORED TO INPUT FROMA HUMAN
`
`USER, SUCH AS ITS OWNER
`
`Inventors: David Cheng, Seattle, WA (US); Stan
`Humphries, Sarnmamish, WA (US);
`Kyusik Chung, Seattle, WA (US); Dong
`Xiang, Sammamish, WA (US);
`Jonathan Burstein, Seattle, WA OJS)
`
`Assignee: Zillow, Inc., Seattle, WA (US)
`
`Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 1394 days.
`
`11/347,024
`
`Feb. 3, 2006
`
`Prior Publication Data
`
`US 2007/0198278 A1
`
`Aug. 23, 2007
`
`Int. Cl.
`
`(2006.01)
`G06Q 40/10
`U.S. Cl.
`........................................ .. 705/35; 705/313
`Field of Classification Search .................. .. 705/35,
`705/313
`
`See application file for complete search history.
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`ll/1994 Jostetal.
`5,361,201 A
`1/2001 Cheetham eta].
`6,178,406 Bl*
`5/2001 Naughton
`6,240,425 B1
`8/2003 Khedkar etal.
`6,609,118 Bl*
`7/2005 Sasajima
`6,915,206 B2
`7,289,965 Bl* 10/2007 Bradley etal. .................. .. 705/1
`
`.......... .. 705/36R
`
`............ .. 705/10
`
`7,461,265 B2
`7,487,114 B2
`7,567,262 B1*
`2003/0212565 A1
`2004/0073508 A1
`2005/0108084 A1*
`2005/0154657 A1*
`2007/0124235 A1
`2007/0185727 A1
`2007/0185906 A1
`2008/0077458 Al
`2009/0043637 A1*
`
`............ .. 345/632
`
`12/2008 Ellmore
`2/2009 Florance et al.
`7/2009 Clemens et al.
`11/2003 Badalietal.
`4/2004 Foster et al.
`705/10
`5/2005 Rarnamoortiet al.
`7/2005 Kimetal.
`..................... .. 705/30
`5/2007 Chakraborty et al.
`8/2007 Maetal.
`8/2007 Humphries et al.
`3/2008 Andersen et al.
`2/2009 Eder
`............................. .. 705/10
`
`OTHER PUBLICATIONS
`
`Vladimir Svetnik et al, Random Forest: A Classification and Regres-
`sion Tool for compound Classification and QSAR Modleing. J. Chem
`Info. Computer Science, 2003, Vol. 43. pp. 1947-1958*
`U.S. Appl. No. 11/927,623, filed Oct. 29, 2007, Humphries et al.
`U.S. Appl. No. 11/971,758, filed Jan. 9, 2008, Humphries et al,
`Quinlan, Ross J., “C45: Programs for Machine Learning,” Machine
`Learning, 1993, 302 pages, Morgan Kaufmann Publishers, San Fran-
`cisco, CA, USA.
`
`(Continued)
`
`Primary Examiner — Kirsten S Apple
`Assistant Examiner — Abdul Basit
`
`(74) Attorney, Agent, or Firm — Perkins Coie LLP
`
`(57)
`
`ABSTRACT
`
`A facility procuring information about a distinguished prop-
`erty from its owner that is usable to refine an automatic
`Valuation of the distinguished property is described. The
`facility displays information about the distinguished property
`used in the automatic valuation ofthe distinguished property.
`The facility obtains user input from the owner adjusting at
`least one aspect of infomiation about the distinguished prop-
`erty used in the automatic Valuation of the distinguished
`property. The facility then displays to the owner a refined
`Valuation of the distinguished property that is based on the
`adjustment of the obtained user input.
`
`40 Claims, 21 Drawing Sheets
`
`oomputar system
`CPU ;— 101
`
`100
`
`persistent storage
`
`/-
`
`computer-veadable
`media drive
`
`network connection
`
`TRULIA - EXHIBIT 1001
`
`TRULIA - EXHIBIT 1001
`
`

`
`US 7,970,674 B2
`Page 2
`
`OTHER PUBLICATIONS
`
`Mobasher, B. “Classification Via Decision Trees in WEKA,” DePaul
`University, Computer Science, Telecommunications, and Informa-
`tion Systems, ECT 584-Web Data Mining, 2005, http://1naya.cs.
`depaul.edu/~classes/Ect584/WEKNclassify.html, Spages [internet
`accessed on Dec. 6, 2007].
`Bennett, Kristin P., “Support Vector Machines: Hype or Hallelujah?”
`SIGKDD Explorations, Dec. 2000, pp. 1-12, vol. 2, issue 2, ACM
`SIGKDD.
`Hill. T. and Lewicki, P., “K-Nearest Neighbors.” Smtistics Methods
`and Applications, 2007, http://WWW.statsoft.com/textbook/stknn.
`htrnl, [internet accessed on [Dec. 6. 2007].
`Breirnan, L., “Random Forests,” Machine Learning, 45, pp. 5-32,
`2001, Kluwer Academic Publishers, The Netherlands.
`http://wWw.ics.uci.edu/~mlearn/databases/housing/housingnames,
`1 page [accessed Dec. 13, 2005].
`StatSoft, Inc., “Classification Trees,” http://vvWw.statsoft.com/teXt-
`booldstclatrehtml, pp. 1-20, © 1984-2003 [accessed Dec. 13,2005].
`Breirnan et al., “Random Forest.” Classification Description, http://
`WWW.stat.berkeley.edu/users/breiman/RandomForests/cc_home.
`htm. pp. 1-28 [accessed Dec. 13. 2005].
`Real-info.com, “What is an AVM,” www.real-info.com’products_
`avm.asp? Internet Archive Date: Oct. 30, 2005, 5 pages [accessed
`Mar. 21,2007].
`RealEstateABC.com, see paragraph headed “How do I make the
`estimate more accurate?” \vvWv.realestateabc.corn/home-values/,
`Internet Archive Dated: Apr. 5, 2006, 4 pages [accessed Mar. 20,
`2007].
`Standard & Poors, “Guidelines for the use of Automated Valuation
`Models for U.K. RMBS Transactions,” http://WvWv.rics.org/NIU
`rdonlyres/8Fcdd20c-7FAC-4549-86FB-3930CDOCBC05/0/'
`StandardandPoorsReportonAVMs.pdf, Published Feb. 20, 2004, 4
`pages.
`
`wWw.r-project.org, “The R Project for Statistical Computing,” http://
`webarchive.org/Web/20060102073515/vvw\v.r-project.org/main.
`shtrnl, 1 page [internet archive date: Jan. 2, 2006].
`“Centre for Mathematical Sciences,” Lund University, http://web.
`archive.org/web/20060101005103/http://www.maths.Ith.se/, 1 page
`[internet archive date: Jan. 1, 2006].
`http://Web.archive.org/web/
`“An
`Introduction
`to
`R,”
`20060118050840/http://cran.r-project.org/doc/manuals/R-intro.
`html, pp. 1-105 [intemet archive date: Jan. 18,2006].
`wWw.cran.r-project.org, “The Comprehensive R Archive Network,”
`http://web.archive.org/Web/20050830073913/cran.r-project.org/
`banner.shtrnl, pp. 1-2 [internet archive date: Aug. 30, 2005].
`Non-Final Office Action for U.S. Appl. No. 11/524,048, Mail Date
`Apr. 29, 2009, 10 pages.
`Non-Final Office Action for U.S. Appl. No. 11/524,047, Mail Date
`Oct. 28, 2009, 12 pages.
`Final Office Action for U.S. Appl. No. 11/524,048, Mail Date Dec. 8,
`2009, 12 pages.
`Non-Final Office Action for U.S. Appl. No. 11/927,623, Mail Date
`Dec. 28, 2010. 22 pages.
`Non-Final Office Action for U.S. Appl. No. 11/347,000, Mail Date
`Apr. 9, 2010, 29 pages.
`Tay et al., “Artificial Intelligence and the Mass Appraisal of Residen-
`tial Apartments,” Journal of Property Valuation and Investment, Feb.
`1, 1992, 17 pages.
`Meyer, Robert T., “The Learning of Multiattribute Judgment Poli-
`cies,” The Journal ofConsumer Research, vol. 14, No. 2, Sep. 1987,
`pp. 155-173.
`Non-Final Office Action for U.S. Appl. No. 11/347,000, Mail Date
`Oct. 27, 2010, 25 pages.
`
`* cited by examiner
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 1 of 21
`
`US 7,970,674 B2
`
`computer system
`
`persistent storage
`
`computer—readab|e
`media drive
`
`network connection
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 2 of 21
`
`US 7,970,674 B2
`
`select recent sales for
`geographic area
`
`forx=I ton
`
`construct tree x
`
`score tree x
`
`receive request for valuation
`identifying home
`
`apply trees, weighted by
`scores, to attributes of home
`identified in request
`
`

`
`
`
`1H9113cI°S'fl
`
`IIOz‘sz‘tint
`
`(I)
`'.:"(‘D
`as:->
`D)
`Ov-vs
`—
`
`N>
`
`E
`
`address
`
`bedrooms
`
`se||in rice 35$
`
`Henderson Coun
`
`recent sales table
`
`1 111 Main St., Hendricks, IL 62012
`
`2 96 Elm St., Hendricks, IL 62014
`
`3 140 Cottontail Rd., Baron, IL 62019
`
`4 6 Spratt Ln., Baron, IL 62019
`
`5 776 Fir St., Hendricks, IL 62014
`
`6 111 Industry Ave., Fenton IL 62017
`
`7 105 Elm St., Hendricks, IL 62014
`
`8 110 Muffet St., Baron, IL 62019
`
`9 156 Elm St., Hendricks, IL 62014
`
`10 142 Cottontail Rd.. Baron, IL 62019
`
`11 160 Prospect Bldv., Fenton IL 62017
`
`$132,500
`
`1/3/2005
`
`$201,000
`
`1/8/2005
`
`$98,750
`
`1/11/2005
`
`$106,500
`
`1/14/2005
`
`$251,000
`
`1/26/2005
`
`$240,000
`
`2/4/2005
`
`$230,000
`
`2/4/2005
`
`$74,900 2/14/2005
`
`$253,500 2/15/2005
`
`$102,000 2/18/2005
`
`$230,000 2/20/2005
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 4 of 21
`
`US 7,970,674 B2
`
`randomly select fraction of
`recent sales and fraction of
`attributes for tree
`
`create root node representing
`all basis sales and fully range
`of each basis attribute
`
`for each node of tree
`
`node should
`
`be split
`
`determine mean selling price of
`basis sales represented by
`node
`
`create pair of children for node,
`each representing an attribute
`subrange on a different side of
`the selected split point and
`node’s full range of other basis
`attributes. and all qualifying
`basis sales
`
`next node
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 5 of 21
`
`US 7,970,674 B2
`
`.
`.
`.
`population satisfies split
`threshold
`
`No
`
`return without identifying
`spm opportunity
`
`452
`
`453
`
`45
`
`4
`
`456
`
`4
`
`57
`
`Yes
`
`determine mean selling price among sales represented by node
`to obtain node mean selling price
`
`sum the squares of the differences between node mean selling
`price and the selling price of each sale represented by the node
`to obtain node overall squared error
`
`divide the overall squared error by the number of sales
`represented by the node - 1 to obtain node variance
`
`for each possible split opportunity in an attribute range
`represented by node
`
`for each side of the possible split opportunity, determine mean
`selling price among sales on that side to obtain split side mean
`selling price
`
`sum the squares of the differences between the selling price of
`each sale represented by the node and the split side mean
`selling price on the same side of the possible split opportunity to
`obtain possible split opportunity squared error
`
`divide the possible split opportunity squared error by the
`number of sales represented by node - 2 to obtain variance for
`possible split opportu nit
`
`next possible split opportunity
`
`select possible split opportunity having lowest variance
`
`462
`
`selected
`ossible split opportunity variance
`< node variance
`
`Yes
`
`return identifying selecte
`split opportunity
`
`.
`.
`.
`.
`return Vfmhout Iderlmymg
`Spm opponumty
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 6 of 21
`
`US 7,970,674 B2
`
`m
`
`address
`
`bedrooms
`
`sellin rice
`
`tree 1 basis table
`
`2 96 Elm St., Hendricks, IL 62014
`
`8 110 Muffet St., Baron, IL 62019
`
`9 156 Elm St., Hendricks, IL 62014
`
`11 160 Prospect Bldv., Fenton IL 62017
`
`13 118 Main St., Hendricks, IL 62012
`
`15 677 Fir St., Hendricks, IL 62014
`
`$201,000
`
`$74,900
`
`$253,500
`
`$230,000
`
`$211,000
`
`$238,000
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 7 of 21
`
`US 7,970,674 B2
`
`/ 600
`
`6°‘
`
`sa|es=2,8,9,11,13,15
`bedrooms = 1- oo
`
`view = no—yes
`
`FIG. 6
`
`‘/ 700
`
`sales = 2, 8, 9, 11, 13, 15
`bedrooms = 1- 00
`
`view = no-yes
`
`bedrooms 5 4
`
`bedrooms > 4
`
`sales = 8, 11
`bedrooms = 1-4
`
`view = no-yes
`valuation = $152,450
`
`703
`
`sales = 2, 9, 13, 15
`bedrooms = 5- oo
`
`View = no-yes
`
`view = yes
`
`713
`
`sales = 2, 13
`bedrooms = 5-00
`
`view = no
`valuation = $206,000
`
`FIG. 7
`
`705
`
`sales = 9, 15
`bedrooms = 5-00
`
`view = yes
`valuation = $245,750
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 8 of 21
`
`US 7,970,674 B2
`
`identify for scoring tree recent
`sales not used as a basis for
`
`constructing tree
`
`for each identified sale
`
`apply tree to attributes of sale
`to obtain value
`
`compare value to selling price
`to determine magnitude of
`error
`
`next identified sale
`
`calculate score inversely
`related to median error
`
`ma nitude
`
`

`
`
`
`
`
`Im‘sz‘lint1u911a([‘Sn
`
`(1)
`1:"(‘D
`as:->
`\o
`Ov-vs
`—
`
`N>
`
`m
`
`address
`
`bedrooms view sellin rice
`
`valuation
`
`error
`
`tree 1 scorin -
`
`1 111 Main St., Hendricks, IL 62012
`
`$132,500
`
`$152,450
`
`3 140 Cottontail Rd., Baron, IL 62019
`
`$98,750
`
`$152,450
`
`4 6 Spratt Ln., Baron, IL 62019
`
`$1 06,500
`
`$152,450
`
`5 776 Fir St., Hendricks, IL 62014
`
`$251,000
`
`$152,450
`
`6 111 Industry Ave., Fenton IL 62017
`
`$240,000
`
`$152,450
`
`7 105 Elm St., Hendricks, IL 62014
`
`$230,000
`
`$245,750
`
`10 142 Cottontail Rd., Baron, IL 62019
`
`$102,000
`
`$152,450
`
`

`
`/1000
`
`21505 SE 2nd St, sammamish, wp. 980?4 /" 1°C"
`ZESTIMATE : $455,399/Em-5 this?)
`
`1002
`
`Value Range: $25|II_.?'44 — $7’52,23E:
`
`Refine value of this home
`
`Mag cnmgar-able homes
`
`
`
`1H9113cI°S'fl
`
`IIOz‘sz‘tint
`
`IZJ00IJ99HS
`
`

`
`
`
`
`
`IIOz‘sz‘lint11191124‘SnIIOz‘sz‘lint11191124‘Sn
`
`
`
`
`
`
`
`
`(1)(1)
`
`'.="(‘D'.="(‘D
`
`as:->as:->
`
`>—a>—a
`
`>—a>—a
`
`O-sO-s
`
`—-—-
`
`N»
`N»
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 12 of 21
`
`US 7,970,674 B2
`
`I’.§€?. Z3
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 13 of 21
`
`US 7,970,674 B2
`
`display initial valuation of
`subject home
`
`solicit updated home attributes
`from owner
`
`display refined valuation that
`takes into account updated
`attributes
`
`solicit informationvabout
`_
`improvements to subject home
`from owner
`
`display refined valuation that
`takes improvements into
`account
`
`solicit other factors affecting
`value of home from owner
`
`display refined valuation that
`takes into account other factors
`
`solicit list of similar comps from
`owner
`
`display refined valuation that
`takes into account similar
`comps
`
`FIG. 14
`
`

`
`
`
`
`
`IIOz‘sz‘lint1u911a([‘Sn
`
`(1)
`'.:"(‘D
`as:->
`>—a
`-Ik
`O-s
`—
`
`N>
`
` *ws.2‘—2?s‘fW
`iazzw raimt. $3f§;’£:£§
`
`>
`
`4_ _
`4
`s
`=:;g;§’.:rzs=a§£:«a *£iv:.¥.z:~‘;: Ezxiii
`4\ {.
`
`

`
`U S. Patent
`
`Jun. 28, 2011
`
`Sheet 15 of 21
`
`US 7,970,674 B2
`
`«.
`éiawafii, ,2
`iwi mfiteaa asst? fame
`
`1
`
`“:22
`
`,
`11%;: fem?
`
`_
`
`

`
`Hwm
`
`atn.fiRamy..._i\
`
`,9
`
`eah.9;
`
`
`
`cm.5:2:2....3.3:2;:333cm8393“.3
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`LI.eehS
`
`iééxsmsfk
`sw sat
`
`22:5
`
`Um
`
`12f0
`
`US 7,970,674 B2
`
`T6.-$J;'._§,¥$s
`
`53+»; téssf
`
`L.
`
`35+.
`
`«W
`
`

`
`U S. Patent
`
`Jun. 28, 2011
`
`12f0001LI.8C.h__S
`
`US 7,970,674 B2
`
`fi.~:%?.i-'2-5: ms $r.5'5i
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 19 of 21
`
`US 7,970,674 B2
`
`, Overall Summary
`l Original Zestimate .......................... .. $555,':'27/I_:::(1)
`‘ Change Home Fact ............................ ., +$
`1,500/F1932
`Change Home Improvements ............... .. + $
`3.300/F1933
`, Your other estimated values ............... .. - $
`300 I
`1934
`Change based on cornparable homes .... .. +3; 2650/-
`1 NEW REVISED VALUE
`2
`$ 553,177 /-
`
`1980
`
`FIG. 19B
`
`: Residence:
`‘; # Bedrooms:
`"‘ # Baths:
`V
`511 ft!
`Lot size (sf):
`Year built:
`
`Single Family
`,
`
`2.5
`1653
`2356
`1955
`
`/-1942
`water *(nDne)
`view:
`Attached
`Garage:
`Architechtural style: Colonial
`Construction quality: Good
`Pool:
`No
`
`l
`
`1950j
`
`Total changes to home facts : $1500
`
`* (Previous info)
`
`FIG. 19C
`
`FIG. 19D
`
`1950
`3; Other Values Detailed Summary
`
`Orchard in back.................... ..+
`1‘ Need new fence .................... ..-
`Total other values
`l
`
`FIG. 19E
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`12f002LI.8C.h__S
`
`US 7,970,674 B2
`
`

`
`U.S. Patent
`
`Jun. 28, 2011
`
`Sheet 21 of 21
`
`US 7,970,674 B2
`
`IE
`
`Henderson Coun
`
`recent sales table for linear reression model
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`sing|e—family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`single-family
`
`

`
`US 7,970,674 B2
`
`1
`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
`
`_
`
`TECHNICAL FIELD
`
`The described technology is directed to the field of elec-
`tronic commerce techniques, and, more particularly, to the
`field of electronic commerce techniques relating to real
`estate.
`
`BACKGROUND
`
`In many roles, it can be useful to be able to accurately
`determine the value of real estate properties (“properties”),
`such as residential real estate properties (“homes”). As
`examples, by using accurate values for homes: taxing bodies
`can equitably set property tax levels; sellers and their agents
`can optimally set listing prices; and buyers and their agents
`can determine appropriate offer amounts.
`A variety of conventional approaches exist for valuing
`homes. Perhaps the most reliable is, for a home that was very
`recently sold, attributing its selling price as its value. Unfor-
`tunately, following the sale of a home, its current value can
`quickly diverge from its sale price. Accordingly, the sale price
`approach to valuing a home tends to be accurate for only a
`short period after the sale occurs. For that reason, at any given
`time, only a small percentage of homes can be accurately
`valued using the sale price approach.
`Another widely-used conventional approach to valuing
`homes is appraisal, where a professional appraiser deter-
`mines a value for a home by comparing some of its attributes
`to the attributes of similar nearby homes that have recently
`sold (“comps”). The appraiser arrives at an appraised value by
`subjectively adjusting the sale prices of the comps to reflect
`differences between the attributes of the comps and the
`attributes of the home being appraised. The accuracy of the
`appraisal approach can be adversely affected by the subjec-
`tivity involved. Also, appraisals can be expensive, can take
`days or weeks to complete, and may require physical access
`to the home by the appraiser.
`While it might be possible to design systems that automati-
`cally value homes, such automatic valuations would gener-
`ally be performed based upon the contents of a public data-
`base, and without input from each home’s owner or other
`information not in the public database. In such systems, fail-
`ing to consider such information may result in valuations that
`are significantly inaccurate in some instances.
`In view of the shortcomings of conventional approaches to
`valuing homes discussed above, a new approach to valuing
`homes that was responsive to owner input, as well as having
`a high level of accuracy, and being inexpensive and conve-
`nient, would have significant utility.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a block diagram showing some ofthe components
`typically incorporated in at least some of the computer sys-
`tems and other devices on which the facility executes.
`FIG. 2 is a flow diagram showing steps typically performed
`by the facility to automatically determine current values for
`homes in a geographic area.
`FIG. 3 is a table diagram showing sample contents of a
`recent sales table.
`
`2
`
`FIG. 4A is a flow diagram showing steps typically per-
`formed by the facility in order to construct a tree.
`FIG. 4B is a flow diagram showing steps typically per-
`formed by the facility in order to determine whether and how
`to split a node of a tree.
`FIG. 5 is a table diagram showing sample contents of a
`basis table containing the basis information selected for the
`tree.
`
`FIG. 6 is a tree diagram showing a root node corresponding
`to the basis table 500.
`
`FIG. 7 is a tree diagram showing a completed version ofthe
`sample tree.
`FIG. 8 is a flow diagram showing steps typically performed
`by the facility in order to score a tree.
`FIG. 9 is a table diagram showing sample results for scor-
`ing a tree.
`FIG. 10 is a display diagram showing detailed information
`about an individual home.
`
`FIG. 11 is a display diagram showing a map identifying a
`number of homes in the same geographic area.
`FIG. 12 is a display diagram showing a display typically
`presented by the facility containing the attributes of a particu-
`lar home.
`
`FIG. 13 is a display diagram showing a display typically
`presented by the facility to identify possible comparable sales
`on a map.
`FIG. 14 is a flow diagram showing steps typically per-
`formed by the facility in order to tailor a valuation of a subject
`home based on information provided by a user such as the
`home’ s owner.
`
`FIG. 15 is a display diagram showing a sample display
`typically presented by the facility to display an initial valua-
`tion of the subject home and solicit updated home attributes
`from the user.
`
`FIG. 16 is a display diagram showing a typical display
`presented by the facility to permit the user to describe
`improvements made to the subject home.
`FIG. 17 is a display diagram showing a sample display
`typically presented by the facility to enable the user to
`describe otheraspects ofthe subject home that affect its value.
`FIG. 18 is a display diagram showing a sample display
`presented by the facility in order to enable the user to identify
`comps regarded by the owner as similar to the subject home.
`FIGS. 19A-19F show a sample display typically presented
`by the facility in order to present an overall revised value for
`the subject home.
`FIG. 20 is a table diagram showing sample contents of
`recent sales information used to construct a linear regression
`valuation model that is based on the attributes whose values
`
`are available for the user to update in the first step of the
`process of generating a tailored valuation.
`
`DETAILED DESCRIPTION
`
`Overview
`
`A software facility for automatically determining a current
`value for a home or other property that is tailored to input
`from its owner or another user (“the facility”) is described.
`While the following discussion liberally employs the word
`“home” to refer to the property being valued in other nearby
`properties, those skilled in the art will appreciate that the
`facility may be straightforwardly applied to properties of
`other types. Similarly, while a wide Variety of users may use
`the facility, including the owner, an agent or other person
`representing the owner, a prospective buyer, an agent or other
`person representing prospective buyer, or another third party.
`
`

`
`US 7,970,674 B2
`
`3
`
`4
`
`5
`
`10
`
`20
`
`25
`
`30
`
`40
`
`In some embodiments, the facility uses a web site to receive
`information from a user and display to the user a refined
`valuation for the home that is based upon the information
`provided by the user. In some embodiments, the information
`provided by the user may include additional, corrected, and/
`or updated attributes of the home relative to the attributes
`known by the facil'ty, such as attributes retrieved by the
`facility from a public or private database of home attributes;
`information about improvements to the home; information
`about other factors li <ely to affect the value ofthe home, such
`as well-kept grounds, historical significance, ground water
`issues, etc.; and ir formation identifying, among recent,
`nearby sales of comparable homes (“comps”), those that the
`user regards as the Ir ost similar to the subject home. In some
`embodiments, the facility displays the results of refining its
`valuation in a manner that makes clear how the valuation was
`
`affected by the diffe *ent information provided by the user.
`By enabling an user to refine a valuation ofhis or her home
`based upon informaton about the home known to the user, the
`facility in many cases makes the valuation more accurate than
`would otherwise be possible, and/or helps the user to more
`fully accept the valuation as appropriate.
`Home Valuation
`
`the facility constructs andjor
`In some embodiments,
`applies housing price models each constituting a forest of
`classification trees. In some such embodiments, the facility
`uses a data table that identifies, for each ofa number ofhomes
`recently sold in the geographic region to which the forest
`corresponds, attributes of the home and its selling price. For
`each of the trees comprising the forest, the facility randomly
`selects a fraction of homes identified in the table, as well as a
`fraction of the attributes identified in the table. The facility
`uses the selected attributes of the selected homes, together
`with the selling prices of the selected homes, to construct a
`classification tree in which each non-leaf node represents a
`basis for differentiating selected homes based upon one ofthe
`selected attributes. For example, where number of bedrooms
`is a selected attribute, a non-leaf node may represent the test
`“number ofbedrooms§4.” This node defines 2 subtrees in the
`
`tree: one representing the selected homes having 4 or fewer
`bedrooms, the other representing the selected homes having 5
`or more bedrooms. Each leaf node ofthe tree represents all of
`the selected homes having attributes matching the ranges of
`attribute values corresponding to the path from the tree’ s root
`node to the leaf node. The facility assigns each leaf node a
`value corresponding to the mean of the selling prices of the
`selected homes represented by the leaf node.
`In some areas of the country, home selling prices are not
`public records, and may be difficult or impossible to obtain.
`Accordingly, in some embodiments, the facility estimates the
`selling price of a home in such an area based upon loan values
`associated with its sale and an estimated loan-to-value ratio.
`
`In order to weight the trees ofthe forest, the facility further
`scores the usefulness of each tree by applying the tree to
`homes in the table other than the homes that were selected to
`
`construct the tree, and, for each such home, comparing the
`value indicated for the home by the classification tree (i .e., the
`value of the root leaf node into which the tree classifies the
`
`home) to its selling price. The closer the values indicated by
`the tree to the selling prices, the higher the score for the tree.
`In most cases, it is possible to determine the attributes of a
`home to be valued. For example, they can often be obtained
`from existing tax or sales records maintained by local gov-
`emments. Altematively, a home’s attributes may be inputted
`by a person familiar with them, such as the owner, a listing
`agent, or a person that derives the information fron1 the owner
`or listing agent. In order to determine a value for a home
`
`whose attributes are known, the facility applies all ofthe trees
`ofthe forest to the home, so that each tree indicates a value for
`the home. The facility then calculates an average of these
`values, each weighted by the score for its tree, to obtain a
`value for the home. In various embodiments, the facility
`presents this value to the owner of the l1ome, a prospective
`buyer of the hon1e, a real estate agent, or another person
`interested in the value of the home or the value of a group of
`homes including the home.
`In some embodiments, the facility applies its model to the
`attributes of a large percentage ofhomes in a geographic area
`to obtain and convey an average home value for the homes in
`that area. In some embodiments, the facility periodically
`determines an average home value for the homes in a geo-
`graphic area, and uses them as a basis for determining and
`conveying a home value index for the geographic area.
`Because the approach employed by the facility to deter-
`mine the value of a home does not rely on the home having
`recently been sold, it ca11 be used to accurately value virtually
`any home whose attributes are known or ca11 be determined.
`Further, because this approach does not require the services of
`a professional appraiser, it can typically determine a home’s
`value quickly and inexpensively, in a manner generally free
`from subjective bias.
`FIG. 1 is a block diagram showing some ofthe components
`typically incorporated in at least son1e of the computer sys-
`tems and other devices on which the facility executes. These
`computer systems and devices 100 may include one or more
`central processing units (“CPUS”) 101 for executing com-
`puter programs; a computer memory 102 for storing pro-
`grams and data—including data structures, database tables,
`other data tables, etc.—while they are being used; a persistent
`storage device 103, such as a hard drive, for persistently
`storing programs and data; a computer-readable media drive
`104, such as a CD-ROM drive, for reading programs a11d data
`stored on a computer-readable medium; and a network con-
`nection 105 for connecting the computer system to other
`computer systems, such as via the Internet, to exchange pro-
`grams and/or data—including data structures. In various
`embodiments, the facility can be accessed by any suitable
`user interface including Web services calls to suitable APIs.
`While computer systems configured as described above are
`typically used to support the operation of the facility, one of
`ordinary skill in the art will appreciate that the facility may be
`implemented using devices of various types and configura-
`tions, and having various components.
`FIG. 2 is a flow diagram showing steps typically performed
`by the facility to automatically determine current values for
`homes in a geographic area. The facility may perform these
`steps for one or more geographic areas of one or more differ-
`ent granularities, including neighborhood, city, county, state,
`country, etc. These steps may be performed periodically for
`each geographic area, such as daily. In step 201, the facility
`selects recent sales occurring in the geographic area. The
`facility may use sales data obtained from avariety ofpublic or
`private sources.
`FIG. 3 is a table diagram showing sample contents of a
`recent sales table. The recent sales table 300 is made up of
`rows 301-315, each representing a home sale that occurred in
`a recent period of time, such as the preceding 60 days. Each
`row is divided into the following columns: ar1 identifier col-
`umn 321 containing a11 identifier for the sale; an address
`colurrm 322 containing the address of the soldhome; a square
`foot column 323 containing the floor area of the l1ome; a
`bedrooms column 324 containing the number ofbedrooms in
`the home; a bathrooms column 325 containing the number of
`bathrooms in the home; a floors column 326 containing the
`
`

`
`US 7,970,674 B2
`
`5
`
`6
`
`number of floors in the home; a view column 327 indicating
`whether the home has a view; a year column 328 showing the
`year in which the house was constructed; a selling price
`column 329 containing the selling price at whicl1 the home
`was sold; and a date column 330 showing the date on which 5
`the home was sold. For example, row 301 indicates that sale
`number 1 of the home at 111 Mai11 St., Hendricks, Ill. 62012
`having a floor area of 1850 square feet, 4 bedrooms, 2 bath-
`rooms, 2 floors, no view, built in 1953, was for$132,500, and
`occurred on Jan. 3, 2005. While the contents of recent sales
`table 300 were included to pose a comprehensible example,
`those skilled in the art will appreciate that the facility can use
`arece11t sales table having columns corresponding to different
`and’or a larger number ofattributes, as well as a larger number
`of rows. Attributes that may be used include, for example,
`construction materials, cooling technology, structure type,
`fireplace type, parking structure, driveway, heating technol-
`ogy, swimming pool type, roofing material, occupancy type,
`home design type, view type, view quality, lot size and dimen-
`sions, number of rooms, number of stories, school district,
`longitude and latitude, neighborhood or subdivision,
`tax
`assessment, attic and other storage, etc. For a variety of rea-
`sons, certain values may be omitted from the recent sales
`table. In some embodiments, the facility imputes missing
`values using the median value i11 the same colunm for con-
`tinuous variables, or the mode (i.e., most frequent) value for
`categorical values.
`While FIG. 3 and each of the table diagrams discussed
`below show a table whose contents and organization are
`designed to make them more comprehensible by a human
`reader, those skilled in the art will appreciate that actual data
`structures used by the facility to store this information may
`differ from the table shown, in that they, for example, may be
`organized in a different manner; may contain more or less
`information than shown; may be compressed andjor
`encrypted; etc.
`Returning to FIG. 2, in steps 202-205, the facility con-
`structs and scores a number oftrees, such as 100. This number
`is configurable, with larger numbers typically yielding better
`results but requiring the application of greater computing
`resources. In step 203, the facility constructs a tree. In some
`embodiments, the facility constructs and applies random for-
`est valuation models using an R mathematical software pack-
`age available at http://cran.r-project.org/ and described at
`http://www.maths .lth.se/help/R/ .R/library/randomForest’
`html/randon1Forest.htn1l. Step 203 is discussed in greater
`detail below in connection with FIG. 4. In step 204, the
`facility scores the tree constructed in step 203. Step 204 is
`discussed in greater detail below in connection with FIG. 8.
`In steps 206-207, the facility uses the forest of trees con-
`structed and scored in steps 202-205 to process requests for
`home valuations. Such requests may be individually issued
`by users, or issued by a program, such as a program that
`automatically requests valuations for all homes in the geo-
`graphic area at a standard frequency, such as daily, or a
`program that requests valuations for all of the homes occur-
`ring on a particular map in response to a request fron1 a user
`to retrieve the map. In step 206, the facility receives a request
`for valuation identifying the home to be valued. In step 207,
`the facility applies the trees constructed in step 203, weighted
`by the scores generated for them in step 204, to the attributes
`in the home identified i11 the received request in order to
`obtain a valuation for the home identified in the request. After
`step 207, the facility continues in step 206 to receive the next
`request.
`Those skilled in the art will appreciate that the steps shown
`in FIG. 2 and in each of the flow diagrams discussed below
`
`may be altered in a variety of ways. For example, the order of