RANKING QUERY RESULTS USING BIOMETRIC PARAMETERS

Abstract

Methods, systems, and apparatus, including computer program products, for providing query results using biometric parameters. One of the methods includes providing a search result in response to receiving a search query. If one or more of biometric parameters of a user indicate likely negative engagement by the user with the first search result, an additional search result is obtained and provided in response to the search query.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit under 35 U.S.C. §119(e) of the filing date of U.S. Provisional Patent Application No. 61/654,746, filed on Jun. 1, 2012, entitled “Ranking Query Results Using Biometric Parameters.” This application also claims the benefit under 35 U.S.C. §119(e) of the filing date of U.S. Provisional Patent Application No. 61/654,518, filed on Jun. 1, 2012, entitled “General Purpose Question and Answer Handling System.” This application also claims the benefit under 35 U.S.C. §119(e) of the filing date of U.S. Provisional Patent Application No. 61/654,437, filed on Jun. 1, 2012, entitled “Automatically Providing A Follow-Up Answer to a Question Based on User Feedback.” This application also claims the benefit under 35 U.S.C. §119(e) of the filing date of U.S. Provisional Patent Application No. 61/654,512, filed on Jun. 1, 2012, entitled “Automatically Training a Dialog System Using User Feedback.” The entirety of the foregoing applications is herein incorporated by reference.

BACKGROUND

This specification relates to ranking query results.

Internet search engines and personal assistant devices provide information from resources (e.g., Web pages, images, text documents, multimedia content) in response to user queries. A query result can include, for example, a snippet of information that answers the user query. The usefulness of a search engine or personal assistant device can depend on its ability to provide satisfactory query results.

SUMMARY

This specification describes technologies relating to modifying the scoring and ranking of query results using biometric indicators of user satisfaction or negative engagement with a search result.

In general, one innovative aspect of the subject matter described in this specification can be embodied in methods that include the actions of providing a search result in response to receiving a search query; receiving one or more biometric parameters of a user that are obtained after the search result is presented to the user; determining, one or more computers, that one or more of the biometric parameters indicate likely negative engagement by the user with the first search result; in response to determining that one or more of the biometric parameters indicate likely negative engagement with the search result, obtaining an additional search result in response to the search query; and providing the additional search result. Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods. A system of one or more computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination of them installed on the system that in operation causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.

The foregoing and other embodiments can each optionally include one or more of the following features, alone or in combination. The actions include providing a search result to a user; receiving one or more biometric parameters of the user and a satisfaction value; and training a ranking model using the biometric parameters and the satisfaction value. Determining that one or more biometric parameters indicate likely negative engagement by the user with the first search result comprises detecting an increased body temperature, detecting pupil dilation, detecting eye twitching, detecting facial flushing, detecting a decreased blink rate, or detecting an increased heart rate. Obtaining the additional search result comprises obtaining an additional search result having a satisfaction score that satisfies a threshold. The actions include providing the additional search result to a second user; receiving one or more biometric parameters from the second user; generating satisfaction score for the additional search result from a trained ranking model using the one or more biometric parameters from the second user; and ranking the additional search result among one or more other search results by respective satisfaction scores. Obtaining the additional search result comprises obtaining a next highest search result for the search query; computing a satisfaction score for the next highest search result; and determining that the satisfaction score for the next highest search result satisfies a threshold. The one or more computers comprise a dialogue engine.

Particular embodiments of the subject matter described in this specification can be implemented so as to realize one or more of the following advantages. The performance of existing search engines and personal assistant devices can be improved. For example, query results can be re-ranked based on previous user feedback by considering biometric data in the ranking Alternatively, or additionally, user satisfaction to a presented query result can be determined from biometric data, and a new query result can be presented based on the determined satisfaction. Query results can be personalized based on the biometric data for a specific user or a population of users.

The details of one or more embodiments of the subject matter described in this specification are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages of the invention will become apparent from the description, the drawings, and the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates an example search system for providing query results responsive to submitted queries.

FIGS. 2A and 2B illustrate example graphical user interfaces displaying a query result provided in response to a query.

FIG. 3 illustrates an example method for obtaining an alternative search result based on satisfaction scores

FIG. 4 illustrates an example method for generating an aggregate satisfaction score for a query result for a query.

FIG. 5 illustrates an example method for training a machine learning system to output a user satisfaction score given an input of user biometric parameters.

DETAILED DESCRIPTION

FIG. 1 illustrates an example search system 130 for providing query results responsive to submitted queries, as can be implemented in an internet, an intranet, or another client and server environment. The search system 130 is an example of an information retrieval system in which the systems, components, and techniques described below can be implemented.

A user 102 can interact with the search system 130 through a client device 110. For example, the client device 110 can be a computer coupled to the search system 130 through a network 120, which can be, for example, a local area network (LAN), a wide area network (WAN), e.g., the Internet, or any other appropriate network for communicating data. In some implementations, the search system 130 and the client device 110 can be one machine. For example, a user can install a desktop search application on the client device 110. The client device 110 will generally include a random access memory (RAM) and a processor.

The search system 130 can be implemented as, for example, computer programs installed on one or more computers in one or more locations that are coupled to each other through a network. The search system 130 includes a natural language query processing engine 140, a dialogue engine 150, and a search engine 160.

A user 102 can submit a query 112 to the search system 130 by a variety of methods. For example, user 102 can submit the query by speaking the query 112. An audio input device associated with client device 110 will detect the query 112 and transmit the query 112 through network 120 to search system 130. Other methods of submitting queries to search system 130 can also be performed. For example, user 102 can interact with a user interface associated with client device 110 to submit the query 112. The user interface can be any appropriate input device (e.g., keyboard, mouse, touch display device) that allows user 102 to enter query 112 into a graphical user interface of search system 130.

When the query 112 is received, the search system 130 can determine how to appropriately process and respond to the query 112. For example, in the case that the query 112 is an audio query, a natural language query processing engine 140 that includes one or more speech recognition modules that transform an audio query into one or more terms.

A dialogue engine 150 can evaluate the terms of the query and determine an appropriate way of responding to the query. For example, the dialogue engine 150 can determine that an appropriate response to a query is to provide one or more search results. The dialogue engine 150 can also determine that an appropriate response to a query is to provide driving directions.

To generate search results for the query, the dialogue engine 150 can provide the query to a search engine 160. When the query 112 is received by the search engine 160, the search engine 160 identifies resources that satisfy the query 110. The search engine 160 may also identify a particular “snippet” or section of one or more of the resources that are relevant to the query. The search engine 160 will generally include an indexing engine 162 that indexes resources (e.g., web pages, images, or news articles on the Internet) found in a collection of resources, an index database that stores the indexed resources, and a ranking engine 164 (or other software) to rank the resources that match the query 112.

The ranking engine 134 ranks query results that are responsive to the query by determining one or more signals for the query result and the query pair, determining a score for each query result, and then ranking the query results based on the received scores. Examples of signals include signals indicating relevance of the resource to the query and signals indicating the quality of the resource.

In response to the received query 112, the search system 130 transmits a first search result 114 over the network 120 back to the client device 110. For example, the search system 130 can generate a search results page that identifies the first search result 114 and which can be rendered in a user interface of the client device 110.

The client device 110 also includes a biometric parameters module that collects biometric parameters from the user 102 in order to measure the satisfaction or engagement of the user 102 with the first search result 114. The biometric parameters module 106 can collect body temperature, pupil dilation, eye twitch, facial flushing, heart rate, facial features, or other types of biometric parameters from the user 102 after the user is presented with the first search result 114.

The client device 110 can then transmit the collected biometric parameters 116 over the network 120 to the search system 130 so that the search system 130 can evaluate the biometric parameters 114 to measure user engagement with the first search result, as will be described in more detail below.

If the biometric parameters 116 indicate likely negative engagement with the first search result 114, the search system 130 can automatically obtain a second search 118 and transmit the second search result 118 over the network 120 to the client device 110. In some implementations, the search system 130 obtains a second search result 118 with which other users have not had likely negative engagement, based on their respective previously received biometric parameters.

In response to receiving the second search result 118, the client device 110 can present the second search result 118 on the client device 110 instead of the first search result 114.

FIGS. 2A and 2B illustrate example graphical user interfaces displaying a query result provided in response to a query.

In FIG. 2A, the query “Where should I eat in San Francisco?” 202 is submitted to a search system through a graphical user interface of user device 220, and in response, the search system causes query result 204 to be presented in the graphical user interface of the user device 220.

The search system can gather biometric feedback from a user for the query 202 and query result 204, e.g. from a biometric parameters module of a client device. For example, a camera on mobile device 220 can capture the facial features of a user 208 and can provide the captured facial features to the search system. The biometric parameters gathered by the search system can be anonymized.

The search system can evaluate the received biometric parameters. For example, if the user had a bad experience at a restaurant associated with search result 204, seeing the name of the restaurant in search result 204 may cause the user 208 to exhibit a frown or a scowl, which can indicate likely negative engagement with search result 204.

The system can use the biometric parameters to rerank search results obtained in response to the search query 202 and instead provide the alternative search result 206 to the user device 220 for presentation to the user 208.

The user device 220 can again capture biometric parameters from the user 208, e.g. by using a camera, in connection with the user 208 viewing the alternative search result 206. The user device 220 can then provide the biometric parameters to the search system for analysis. If the biometric parameters do not indicate likely negative engagement with the alternative search result 206, the search system can determine that the user is satisfied with the alternative search result 206. Thus, using the biometric parameters alone as input from the user the search system can identify an alternative query without further input from the user 208.

While FIGS. 2A-2B show a visual presentation of query results, presenting query results can include various forms of presentation including, for example, causing query results to be presented on a display device, causing sounds corresponding to the query results to be presented, or causing haptic feedback corresponding to the query results to be presented. Other methods of presenting query results are possible.

While FIGS. 2A-2B illustrate an example where a user submits queries through a text box in the graphical user interface of a search engine, the user can submit queries through various interfaces, including, for example, by speaking the query.

FIG. 3 illustrates an example method for obtaining an alternative search result based on satisfaction scores. For convenience, the example method 300 will be described in reference to a system that performs method 300. The system can be, for example, the search system 114 described above with reference to FIG. 1.

The system receives a query (302), for example, as described above with reference to FIG. 1. The system provides a search result in response to receiving the query (304).

The system receives biometric parameters after the search result is presented to the user (306). For example, a user device can use a biometric parameters module to obtain one or more biometric parameters of a user in connection with the user being presented with the obtained search result.

The system determines that one or more biometric parameters indicate likely negative engagement with the search result (308). For example, the system can use the received biometric parameters as input to a trained model. As output, the model can generate a score indicating that the input biometric parameters indicate likely negative engagement with the search result. In some implementations, the system builds the model using biometric parameters obtained from other users, as will be described in more detail below with reference to FIG. 5.

The system obtains an additional search result in response to the search query (310). Because the user's biometric parameters indicated likely negative engagement with the search result, the system can in response obtain a different query result that is more likely to result in positive engagement by the user.

For example, the system can obtain a query result that has a high aggregate satisfaction score computed from individual satisfaction scores for other users, described in more detail below with reference to FIG. 4.

In some implementations, the system can modify a ranking of search results identified in response to the search query by using aggregate satisfaction scores for the search results. In some implementations, the system modifies the initial query result scores by scaling each score by an aggregate satisfaction value for the query result. For example, the system can divide or multiply an the initial query result score by the aggregate satisfaction score, or by a value derived from the aggregate satisfaction score. In some other implementations, the system can rank the search results identified in response to the search query by aggregate satisfaction scores. Other modifications can also be used. For example, the system can add the aggregate satisfaction score to the initial score.

The search system can then select a query result for presentation based on the modified ranking For example, the system can select the highest-scoring, or top n highest-scoring query results, where n is a positive integer. The system can then cause one or more query results from the modified ranking to be presented on a user device or cause sound describing the one or more query results to be presented through a speaker associated with the user device.

FIG. 4 illustrates an example method for generating an aggregate satisfaction score for a query result for a query. For convenience, the example method 400 will be described in reference to a system that performs method 400. The system can be, for example, the search system 114 described above with reference to FIG. 1.

The system provides a query result that satisfies a particular query to a user (402), for example, as described above with reference to FIG. 1.

The system obtains data describing one or more biometric parameters of a user to whom the search result is presented (404). The data can be anonymized. The system can obtain the biometric parameters in various ways. For example, in some implementations, the system obtains the biometric parameters from video or image data received from an image or video input device, e.g., a camera configured to obtain video data. In other implementations, the system obtains the biometric parameters from infrared data received from an infrared input device, e.g., a camera configured to obtained infrared data. The data describing the one or more biometric parameters can be obtained from video and infrared data using conventional techniques. The data can describe at least one of the of user's body temperature, heart rate, and facial features. The facial features can describe at least one of the user's pupil dilation, eye twitch, facial flushing, blink rate, and facial expression.

In some implementations, the data describing the one or more biometric parameters of the user is obtained in response to presenting the first query result to the user. In some implementations, the data is obtained periodically before and after presenting the first query results.

The system determines a user satisfaction score the query result (406). The user satisfaction score indicates how satisfied the user is with the query result in response to the query and can represent likely positive or negative engagement with the query result. In some implementations, the system provides the data describing one or more biometric parameters for a given presentation of the query result for the query as input to a model trained using machine learning techniques, for example, as described below with reference to FIG. 5.

In some implementations, the system uses multiple trained models to generate the user satisfaction score, where each model is trained to output a score for the biometric parameters according to a different aspect of user satisfaction, for example, as described below with reference to FIG. 5. The system can then combine the distinct scores from the multiple models to generate an overall user satisfaction score. For example, the system can use a first model that scores the biometric parameters based on quality of emotion, e.g., ranging from content to discontent, and a second model that scores data describing biometric parameters based on strength of emotion, e.g., ranging from weak to strong. In some implementations, the trained models are classifiers that classify the biometric parameters into one of multiple possible classifications, e.g. content or discontent. The resulting scores from the multiple models can be combined into an overall user satisfaction score according to a combination function. For example, two scores from two different models can be added or multiplied together. Alternatively, two scores can be combined by a third trained model, e.g. using linear regression. For example, the third model can be trained to combine scores from the quality of emotion model and the strength of emotion model. In some implementations, the overall satisfaction score is an average of the individual scores from the individual models. Other combinations, for example, a summation, of the individual scores can alternatively be used.

The system generates an aggregate satisfaction score for the query result (408). The system can use user satisfaction scores generated for a query result and query pair to determine an aggregate satisfaction value for the query result. The aggregate satisfaction value can represent the overall positive or negative engagement of users with the query result when provided in response to the query. In some implementations, the aggregate satisfaction score is generated from all user satisfaction scores determined by the system. Alternatively, the aggregate satisfaction score is generated from a subset of user satisfaction scores determined by the system. For example, the aggregate satisfaction score can be generated from user satisfaction scores determined over a period of time. Other methods of generating aggregate satisfaction scores from a subset of user satisfaction values are possible. For example, the aggregate satisfaction score can be generated from a number of the most recently determined aggregate satisfaction scores.

After generating an aggregate satisfaction score for a query result and a query pair, the system can use the aggregate satisfaction score to rank the query result among one or more other query results obtained for the received query. For example, as described above with reference to FIG. 3, the system can use the aggregate satisfaction score to provide a query result that will likely result in positive engagement with a user. For example, if a user's biometric parameters indicate likely negative engagement with a first query result, the system can obtain a different query result that has a high aggregate satisfaction score, which can be a query result that the system determines is likely to be satisfying to the user.

FIG. 5 illustrates an example method for training a machine learning system to output a user satisfaction score given an input of user biometric parameters. For convenience, the example method 500 will be described in reference to a system that performs method 500. The system can be, for example, the search system 130 described above with reference to FIG. 1.

The system receives query result and query pairs and user biometric parameters for each query result and query pair, and a user satisfaction value for each query result and query pair (502-506). For example, the data can be collected by presenting queries and query results to one or more human raters. The system can measure the biometric parameters of the human rater for each query and query result presented to the human rater and ask the human rater to also provide a satisfaction value indicating the rater's satisfaction with the query result, which can reflect, for example, positive or negative engagement with the query result. The biometric parameters can include, for example, body temperature, heart rate, and facial features. The facial features can include the user's pupil dilation, eye twitch, facial flushing, blink rate, and facial expression. Other biometric parameters can also be used. In some implementations, the system asks users to provide satisfaction values ranging in value from −1 to 1, 0 to 1, or 1 to 10. Various ranges can alternatively be used.

The system trains a model using the received data (508), for example, using conventional machine learning techniques. The system can train the model to use the biometric parameters to detect when a user is likely to be dissatisfied with a query result. For example, the system can detect a rise in or a heightened body temperature, which can indicate likely negative engagement with a query result. The system can also detect pupil dilation, which can indicate interest and likely positive engagement with the query result. The system can also detect eye twitching, which can indicate stress and likely negative engagement with the query result. The system can also detect facial flushing, which can indicate anger or annoyance and likely negative engagement with the query result. The system can also detect an increased blink rate, which can indicate excitement and likely positive engagement with the query result. The system can also detect an increase in heart rate, which can indicate stress and likely negative engagement with a query result.

The system can transform the received biometric parameters into one or more feature vectors for each query result and query pair. Each feature vector for a query result and query pair can represent biometric parameters of a user obtained after the user is presented with a query result provided in response to a query. A machine learning algorithm analyzes a collection of training data that includes feature vectors and a satisfaction value for each query result and query pair.

The model trained by the system takes as input a user's biometric parameters and outputs a user satisfaction score that represents the satisfaction of the user with one or more query results provided in response to a query. In some implementations, multiple classifiers can be trained using method 500 to score biometric parameters according to different aspects of user satisfaction. For example, classifiers can be trained to output scores based on quality of emotion, e.g., ranging from content to discontent, and strength of emotion, e.g. ranging from weak to strong.

Embodiments of the subject matter and the operations described in this specification can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions, encoded on a computer storage medium for execution by, or to control the operation of, data processing apparatus. The computer storage medium can be, or be included in, a computer-readable storage device, a computer-readable storage substrate, a random or serial access memory array or device, or a combination of one or more of them. The computer storage medium can also be, or be included in, one or more separate physical components or media (e.g., multiple CDs, disks, or other storage devices).

The operations described in this specification can be implemented as operations performed by a data processing apparatus on data stored on one or more computer-readable storage devices or received from other sources.

The term “data processing apparatus” encompasses all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, a system on a chip, or multiple ones, or combinations, of the foregoing. The apparatus can include special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit). The apparatus can also include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, a cross-platform runtime environment, a virtual machine, or a combination of one or more of them. The apparatus and execution environment can realize various different computing model infrastructures, such as web services, distributed computing and grid computing infrastructures.

A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.

The processes and logic flows described in this specification can be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit).

Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a processor for performing actions in accordance with instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks. However, a computer need not have such devices. Moreover, a computer can be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a portable storage device (e.g., a universal serial bus (USB) flash drive), to name just a few. Devices suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

To provide for interaction with a user, embodiments of the subject matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. In addition, a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user's client device in response to requests received from the web browser.

Embodiments of the subject matter described in this specification can be implemented in a computing system that includes a back-end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front-end component, e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (“LAN”) and a wide area network (“WAN”), an inter-network (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks).

The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. In some embodiments, a server transmits data (e.g., an HTML page) to a client device (e.g., for purposes of displaying data to and receiving user input from a user interacting with the client device). Data generated at the client device (e.g., a result of the user interaction) can be received from the client device at the server.

While this specification contains many specific implementation details, these should not be construed as limitations on the scope of the invention or of what may be claimed, but rather as descriptions of features specific to particular embodiments of the invention. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

Thus, particular embodiments of the invention have been described. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain implementations, multitasking and parallel processing may be advantageous.

Claims

1. A computer-implemented method comprising:

receiving a search query that was submitted by a user;

obtaining a set of search results that are identified as responsive to the search query, each search result having a respective query score;

providing, for output to the user, a particular query result;

obtaining one or more particular biometric parameters of the user that were obtained after the particular query result was provided for output to the user;

providing (i) at least one of the one or more particular biometric parameters, (ii) the search query, and (iii) the particular query result, to a predictive model that is trained to predict satisfaction scores using training data that includes, for each of multiple search query-query result pairs, (i) one or more biometric parameters of a human rater to whom the query result was provided for output in response to the search query, and (ii) a satisfaction score assigned to the search query-query result pair by the human rater;

receiving, from the predictive model, a particular satisfaction score that corresponds to the at least one of the one or more particular biometric parameters;

aggregating the query score associated with the particular query result with the particular satisfaction score; and

re-ranking the particular query result, among the set of query results, based on the aggregated score.

2. The method of claim 1, comprising:

training the predictive model using the at least one of the one or more particular biometric parameters and the particular satisfaction score.

3. The method of claim 1, wherein the at least one of the one or more particular biometric parameters comprises:

data indicating body temperature,

data indicating pupil dilation,

data indicating eye twitching,

data indicating facial flushing,

data indicating blink rate, or

data indicating heart rate.

4. The method of claim 1, comprising:

selecting an additional query result from among the re-ranked set of query results, wherein selecting the additional query result from among the re-ranked set of query results comprises determining that the additional query result has an aggregated score that satisfies a threshold.

5. The method of claim 1, comprising:

providing, to a second user, an additional query result selected from among the re-ranked set of query results;

obtaining one or more particular biometric parameters of the second user;

providing at least one of the one or more of the particular biometric parameters of the second user to a predictive model; and

receiving, from the predictive model, a second user satisfaction score that corresponds to the at least one of the one or more particular biometric parameters of the second user, wherein aggregating the query score associated with the particular query result with the particular satisfaction score comprises aggregating the query score associated with the particular query result with the particular satisfaction score and with the second user satisfaction score.

6. (canceled)

7. The method of claim 1, wherein providing the one or more of the particular biometric parameters to the predictive model comprises providing, by a dialogue engine, the one or more of the particular biometric parameters to the predictive model.

8. A system comprising:

one or more computers that include one or more processors and one or more storage devices storing instructions that are operable, when executed by the one or more processors, to cause the one or more computers to perform operations comprising: receiving a search query that was submitted by a user;

obtaining a set of search results that are identified as responsive to the search query, each search result having a respective query score;

providing, for output to the user, a particular search query result;

obtaining one or more particular biometric parameters of the user that were obtained after the particular query result was provided for output to the user;

providing (i) at least one of the one or more particular biometric parameters, (ii) the search query, and (iii) the particular query result, to a predictive model that is trained to predict satisfaction scores using training data that includes, for each of multiple search query-query result pairs, (i) one or more biometric parameters of a human rater to whom the query result was provided for output in response to the search query, and (ii) a satisfaction score assigned to the search query-query result pair by the human rater;

receiving, from the predictive model, a particular satisfaction score that corresponds to the at least one of the one or more particular biometric parameters;

aggregating the query score associated with the particular query result with the particular satisfaction score; and

re-ranking the particular query result, among the set of query results, based on the aggregated score.

9. The system of claim 8, wherein the operations comprise:

training the predictive model using the at least one of the one or more particular biometric parameters and the particular satisfaction score.

10. The system of claim 8, wherein the at least one of the one or more particular biometric parameters comprises:

data indicating body temperature,

data indicating pupil dilation,

data indicating eye twitching,

data indicating facial flushing,

data indicating blink rate, or

data indicating heart rate.

11. The system of claim 8, wherein the operations further comprise:

selecting an additional query result from among the re-ranked set of query results, wherein selecting the additional query result from among the re-ranked set of query results comprises determining that the additional query result has an aggregated score that satisfies a threshold.

12. The system of claim 8, wherein the operations comprise:

providing, to a second user, an additional query result selected from among the re-ranked set of query results;

obtaining one or more particular biometric parameters of the second user;

providing at least one of the one or more of the particular biometric parameters of the second user to a predictive model; and

receiving, from the predictive model, a second user satisfaction score that corresponds to the at least one of the one or more particular biometric parameters of the second user, wherein aggregating the query score associated with the particular query result with the particular satisfaction score comprises aggregating the query score associated with the particular query result with the particular satisfaction score and with the second user satisfaction score.

13. (canceled)

14. The system of claim 8, wherein providing the one or more of the particular biometric parameters to the predictive model comprises providing, by a dialogue engine, the one or more of the particular biometric parameters to the predictive model.

15. A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

receiving a search query that was submitted by a user;

obtaining a set of search results that are identified as responsive to the search query, each search result having a respective query score;

providing, for output to the user, a particular query result;

obtaining one or more particular biometric parameters of the user that were obtained after the particular query result was provided for output to the user;

providing (i) at least one of the one or more particular biometric parameters, (ii) the search query, and (iii) the particular query result, to a predictive model that is trained to predict satisfaction scores using training data that includes, for each of multiple search query-query result pairs, (i) one or more biometric parameters of a human rater to whom the query result was provided for output in response to the search query, and (ii) a satisfaction score assigned to the search query-query result pair by the human rater;

receiving, from the predictive model, a particular satisfaction score that corresponds to the at least one of the one or more particular biometric parameters;

aggregating the query score associated with the particular query result with the particular satisfaction score; and

re-ranking the particular query result, among the set of query results, based on the aggregated score.

16. The computer-readable medium of claim 15, wherein the operations comprise:

training the predictive model using the at least one of the one or more particular biometric parameters and the particular satisfaction score.

17. The computer-readable medium of claim 15, wherein the at least one of the one or more particular biometric parameters comprises:

data indicating body temperature,

data indicating pupil dilation,

data indicating eye twitching,

data indicating facial flushing,

data indicating blink rate, or

data indicating heart rate.

18. The computer-readable medium of claim 15, wherein the operations further comprise:

selecting an additional query result from among the re-ranked set of query results, wherein selecting the additional query result from among the re-ranked set of query results comprises determining that the additional query result has an aggregated score that satisfies a threshold.

19. The computer-readable medium of claim 15, wherein the operations comprise:

providing, to a second user, an additional query result selected from among the re-ranked set of query results;

obtaining one or more particular biometric parameters of the second user;

providing at least one of the one or more of the particular biometric parameters of the second user to a predictive model; and

receiving, from the predictive model, a second user satisfaction score that corresponds to the at least one of the one or more particular biometric parameters of the second user,

wherein aggregating the query score associated with the particular search query result with the particular satisfaction score comprises aggregating the query score associated with the particular query result with the particular satisfaction score and with the second user satisfaction score.

20. (canceled)

21. The computer-readable medium of claim 15, wherein providing the one or more of the particular biometric parameters to the predictive model comprises providing, by a dialogue engine, the one or more of the particular biometric parameters to the predictive model.

22. The method of claim 1, wherein aggregating the query score associated with the particular query result with the particular satisfaction score comprises:

scaling the query score by the particular satisfaction score.

23. The system of claim 8, wherein aggregating the query score associated with the particular query result with the particular satisfaction score comprises:

scaling the query score by the particular satisfaction score.

24. The method of claim 1, comprising:

selecting an additional query result from among the re-ranked set of query results; and

providing the additional query result.