Granted Patent B1

US 10,380,653 · App. 12/924,037 · Granted Aug 13, 2019

Inventors: Pete Flint (San Francisco, CA); Jesper Sparre Andersen (San Francisco, CA); Sami Inkinen (San Francisco, CA)

Assignee: Trulia, LLC

G06Q30/0278

Abstract

Claims (46)

1. A valuation system for estimating a valuation of a target property comprising a non-transitory computer-readable medium having instructions stored thereon, which when executed by one or more processors of the valuation system cause the valuation system to:

receive, at the one or more processors, user traffic activity information concerning property information included in the website, wherein the user traffic activity comprises information of a first set of users each of whom has accessed a property web page for a target property;

identify, by the one or more processors, a set of properties other than the target property whose property web page was accessed by at least a threshold number of users in the first set of users;

select, by the one or more processors, a group of comparable properties for the target property from the set of identified properties based on the user-traffic activity of each of the threshold number of users in the first set of users;

retrieve, from a memory coupled to the one or more processors, property data regarding the group of comparable properties;

receive, by the one or more processors, actively provided user-generated data, including user reviews on the website of a neighborhood of the target property from a second set of users, wherein the user reviews concern a set of categories regarding the neighborhood;

receive, by the one or more processors, fact-based data comprising data not based on user activity or user input;

process, by the one or more processors, the property data using a first data model to obtain a first valuation of the target property;

process, by the one or more processors, the user reviews using a second data model to obtain a second valuation of the target property;

subject, by the one or more processors, the fact-based data to a third data model to obtain a third valuation of the target property,

wherein at least one of the first data model, the second data model, or the third data model is configured to process only one among the property data, the user-generated data, and the fact-based data; and

combine, by the one or more processors, the obtained first, second, and third valuations to obtain a collective estimated valuation of the target property; and

display, at an interface, the collective estimated valuation of the target property.

2. The system as recited in claim 1 , wherein the user traffic activity includes user click patterns or browsing habits/history.

3. The system as recited in claim 1 , wherein the valuation system is further configured to receive, by the one or more processors, thumb up/down data regarding the target property, wherein the thumb up/down data comprises a user agreement or disagreement with an estimated value of the target property.

4. The system as recited in claim 1 , wherein combining includes using a weighted average of the output of the one or more data models to obtain an estimate valuation of the target property.

5. The system as recited in claim 1 , wherein the valuation system is further configured to process, by the one or more processors, fact-based data and user-generated data using a data model that is configured to process both fact-based data and user-generated data.

6. The system as recited in claim 1 , wherein the valuation system is further configured to:

receive, by the one or more processors, an indication of weights associated with the first set of users or the second set of users; and

adjust, by the one or more processors, the output of the first data model or the second data model using the weights before the combining.

7. The system as recited in claim 6 , wherein the indication is received through a slider bar on the display that enables the current user to adjust the weights between “all users” and “one user” and wherein for “one user,” the adjusting includes giving a weight only to information related to the target user, and for “all users,” the adjusting includes giving a weight information related to all of the first or second set of users.

8. The system as recited in claim 6 , wherein the indication of weights includes giving a larger weight to a user review of one of the second of users who lives in the neighborhood.

9. The system as recited in claim 1 , wherein the valuation system is further configured to process, by the one or more processors, user-generated data with at least one user-defined data model.

10. The system as recited in claim 1 , wherein the valuation system is further configured to receive, by the one or more processors, user-selected subset of individually selectable contributing factors selected from a plurality of individually selectable contributing factors.

11. The system as recited in claim 10 , wherein combining outputs of the first and second data models is based on the user-selected subset of individually selectable contributing factors.

12. The system as recited in claim 1 , wherein the passively provided user-generated data includes a popularity of the target property based on an amount of traffic the Web information regarding the target property receives and a time on market of the target property.

13. The system as recited in claim 1 , wherein selecting the group of comparable properties using the user traffic activity comprises calculating a similarity between the target property and each of the set of properties.

14. The system as recited in claim 1 , wherein the group of comparable properties is further identified by determining a present value of one of the set of properties.

15. The system as recited in claim 1 , wherein the combining of the outputs of the first and second data models to obtain a collective estimated valuation of the target property comprises using an aggregator, wherein the aggregator uses a perceptron that has been trained to determine the collected estimated valuation of the target property based on the outputs of the first and second data models.

16. A method for estimating a valuation of a target property, comprising:

receiving by one or more processors, user traffic activity information concerning property information included in a real estate website, wherein the user traffic activity comprises information of a first set of users each of whom has accessed a property web page for the target property;

using the user traffic activity to identify, by the one or more processors, a set of properties other than the target property whose property web page was accessed by at least a threshold number of users in the first set of users;

using the user traffic activity to select, by the or more processors, a group of comparable properties for the target property from the set of identified properties based on the user-traffic activity of each of the threshold number of users in the first set of users;

retrieving, from a memory coupled to the one or more processors, property data regarding the group of comparable properties;

processing, using the one or more processors, the property data regarding the group of comparable properties using a data model to obtain a valuation of the target property; and

causing the obtained valuation of the target property to be displayed at at least one interface coupled to the one or more processors.

17. A computer program product for estimating a valuation of a target property, the computer program product being embodied in a non-transitory computer readable medium and comprising computer instructions for:

receiving, by one or more processors, user traffic activity information concerning property information included in a real estate website, wherein the user traffic activity comprises information of a first set of users each of whom has accessed a property web page for the target property;

identifying, by the one or more processors, a set of properties other than the target property whose property web page was accessed by at least a threshold number of users in the first set of users;

selecting, by the one or more processors, a group of comparable properties for the target property from the set of identified properties based on the user-traffic activity of each of the threshold number of users in the first set of users;

retrieving, from a memory coupled to the one or more processors, property data regarding the group of comparable properties;

receiving, by the one or more processors, actively provided user-generated data, including user reviews on the website of a neighborhood of the target property from a second set of users, wherein the user reviews concern a set of categories regarding the neighborhood;

processing, by the one or more processors, the property data using a first data model to obtain a first valuation of the target property;

processing, by the one or more processors, the user reviews using a second data model to obtain a second valuation of the target property;

combining, by the one or more processors, the obtained first and second valuations to obtain a collective estimated valuation of the target property; and

causing the collective estimated valuation of the target property to be displayed at at least one interface coupled to one or more processors.

Assignments (8)

Cited By (3)

US 12,271,967 US 12,455,895 US 12,682,288