Abstract: The timing for displaying an overlay a video is optimized using a system that includes an engagement component that receives information regarding frames of a video at which respective users of a plurality of users stop engaging with the video and an analysis component that calculates user engagement scores for the frames based on the information, wherein the user engagement scores reflect a percentage of the plurality of users that stop engaging with the video at the respective frames, and wherein the higher the percentage the lower the user engagement score. The system further includes an identification component that identifies a subset of the frames associated with relatively lower user engagement scores in comparison to other frames of the video and an overlay component that configures an overlay for displaying over the video near at least one of the frames included in the subset.

Abstract: Methods and systems for simulating experience of attending a live program for viewers of a livestreaming content are disclosed. A livestreaming content viewed by a plurality of viewers on respective electronic devices is tracked to detect an event. Viewer interaction data are received from the respective electronic devices associated with the plurality of viewers in response to a detected event in the livestreaming content. The viewer interaction data includes viewer responses of the plurality of viewers in response to the detected event. The viewer interaction data received from the plurality of viewers are aggregated. A cheering audio is synthesized in relation to the detected event based on the aggregated viewer interaction data. An enriched livestreaming content is generated based on the livestreaming content and the cheering audio. Playback of the enriched livestreaming content is facilitated on the respective electronic devices for the plurality of viewers.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to identify media based on historical data. An example method includes: comparing (a) a pitch shifted fingerprint, (b) a time shifted fingerprint, or (c) a resampled fingerprint to a reference fingerprint; in response to a match between any of (a) the pitch shifted fingerprint, (b) the time shifted fingerprint, or (c) the resampled fingerprint and the reference fingerprint, generating indications of (a) a pitch shift value, (b) a time shift value, or (c) a resample ratio that caused the match; in response to collecting broadcast media for a threshold period of time, processing the one or more indications; and in response to a request for a recommendation for information associated with a query, transmitting the recommendation including one or more frequencies of occurrence of (a) the pitch shift value, (b) the time shift value, or (c) the resample ratio.

Abstract: In one aspect, an example method includes (i) obtaining fingerprint repetition data for a portion of video content, with the fingerprint repetition data including a list of other portions of video content matching the portion of video content and respective reference identifiers for the other portions of video content; (ii) identifying the portion of video content as a program segment rather than an advertisement segment based at least on a number of unique reference identifiers within the list of other portions of video content relative to a total number of reference identifiers within the list of other portions of video content; (iii) determining that the portion of video content corresponds to a program specified in an electronic program guide using a timestamp of the portion of video content; and (iv) storing an indication of the portion of video content in a data file for the program.

Abstract: A time associated with last selection of a content item at a user device may be determined. A time associated with a last user interaction with the user device may be determined. During presentation of the content item at the user device, it may be determined that a supplemental content insertion opportunity is available. A request for a supplemental content item may be sent to a server, wherein the request comprises data indicating the time associated with the last selection of the content item at the user device and data indicating the time associated with the last user interaction with the user device.

Abstract: The disclosure relates to a method for monitoring usage of at least one application executed within an operating system of an electronic device. The method includes at least one iteration of, while a processing condition is not satisfied, determining a video event associated with a current application and storing the video event in a data structure, delivering a set of video events. The method further includes, once the processing condition is satisfied, determining a past time slot, calling an usage statistics API provided on the operating system, delivering a set of application events representative of whether an application has been put in system foreground or system background, retrieving application events that have occurred during the past time slot, retrieving video events that have occurred during the past time slot, chronologically ordering the retrieved application and video events, and delivering a consolidated set of ordered events.

Abstract: Aspects of the subject disclosure may include, for example, receiving network requirements for a new telecommunications network to be developed by a network operator, developing a workflow for design and development of the new telecommunications network based on stored knowledge base data about past telecommunications network projects of the network operator, the stored knowledge base data including information about prior milestone processes of the past telecommunications network projects of the network operator, modifying the information about prior milestone processes based on the network requirements for the new telecommunications network, forming a plurality of milestone processes for the new telecommunications network, and performing respective milestone processes of the plurality of milestone processes to design, develop and test the new telecommunications network, including performing a network ready milestone process of final testing of network functions and a certification that the new telecommunica

Abstract: The present specification provides, amongst other things, a computing resource Intermediation engine and method that can select and generate secondary content based on one or more demographic parameters, secondary content options and engagement ratings for those secondary content options.

Abstract: A monitoring camera video sharing system includes a user terminal for viewing of video data of a monitoring camera, and a server for storing access permission data including first user information and monitoring camera information. The first user information includes an access right to the video data of the monitoring camera, and the monitoring camera information indicates information of the monitoring camera. The server updates the access permission data by adding an access right to the video data of the selected monitoring camera to second user information based on sharing processing of the access permission data to the second user information, and sets the video data of the selected monitoring camera linked to the access permission data and specified by the monitoring camera information such that the video data of the selected monitoring camera is accessible by the user terminal based on the second user information.

Abstract: A first system may receive first data of a user associated with a first identifier and second data of the user associated with a second identifier that exclude user identifying information from a first and second data source, respectively. A key may be received from a second system that associates the first and second identifier with a common identifier generated to preserve user anonymity at the first system. The key may be used to generate aggregated data for the user that includes the first and second data and is associated with the common identifier. The aggregated data may be processed to determine the user as a candidate for an offer set. The common identifier for the user is included in a candidacy list for the offer set, and provided to the second system to facilitate request processing associated with the presentation of the offer set to candidate users.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed for media crediting and, more particularly, methods and apparatus of media device detection for minimally invasive media meters. An example apparatus disclosed herein to detect media devices presenting media includes means for generating a cluster of media locations from a reference population of media locations based on media identifying information received from a presentation of media at an unidentified media device at a first media location, means for determining media devices available at the media locations in the cluster of media locations, means for identifying the unidentified media device based on (1) the media devices available at the media locations in the cluster of media locations and (2) an identity of a media device determined to be available in a majority of media locations in the cluster of media locations.

Abstract: Techniques are described for low-latency streaming of media content using a lossless protocol. For example, a media stream, comprising encoded video data, can be streamed to a plurality of streaming clients via the lossless protocol. The plurality of streaming clients can be monitored to determine whether any of them have fallen behind in streaming the media stream. When a streaming client falls behind, a portion of video data to be streamed to the streaming client can be selectively dropped based on scalability information and/or long term reference (LTR) frame information. The low-latency streaming can be performed without using per-client quality feedback from the plurality of streaming clients. When streaming using a semi-lossy protocol, a plurality of delivery modes can be used, where each delivery mode is for a different type of encoded video data and provides a different level of reliability.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to associate geographic locations with user devices. An example apparatus includes at least one memory, instructions in the apparatus, and processor circuitry to execute the instructions to extract source identifying information from a beacon transmitted from a user device to an audience measurement entity, classify a network that connects the user device to the audience measurement entity based on the source identifying information, query a location service of a provider of a mobile network to retrieve a location associated with the user device, store the location in association with media identifying information, and generate a user profile associating the user device with the retrieved location.

Abstract: Systems and methods for time-shifted prefetching of predicted content for wireless users. Prefetching video data can include retrieving video data and feature data and generating a video candidate set including selected videos of the related video data. The method can further include determining predicted videos using a machine learning algorithm. The predicted videos can be videos of the video candidate set that are likely to be viewed by a user at a future time. The method can further include prefetching the predicted videos by downloading the predicted videos.

Abstract: The MOBILE CONTENT TRACKING PLATFORM METHODS (“Mobile-Track”) transforms TV program schedule listing information and user channel selection via Mobile-Track components, into TV audience viewing data and ad effects data. A method is disclosed, comprising: obtaining TV program schedule listing data; providing the obtained TV program schedule listing data to a user mobile device; receiving an indication of user channel selection; determining a media program associated with the received indication of user channel selection; determining an ad associated with the determined media program; providing a survey question related to the determined ad to the user mobile device; and receiving a user response to the survey question.

Abstract: In some embodiments, an electronic device presents recommended content for a set of users in accordance with some embodiments of the disclosure. The electronic device is associated with a plurality of user accounts, including a first user account, displaying, via a display generation component, a media browsing user interface associated with the first user account. In accordance a determination that the first user account satisfies one or more criteria and in accordance with a determination that a first subset of user accounts of the plurality of user accounts associated with the electronic device has been selected for inclusion in content recommendations, the electronic device is configured to display, in the media browsing user interface, one or more representations of media content recommendations based on one or more characteristics of the first user account and the first subset of user accounts.

Abstract: Recommending audiovisual content items to a subscriber based on locations visited by the subscriber. Location information from a mobile device having GPS functionality is received. The location information includes locations the subscriber has visited with the mobile device. The various locations may be counted, ranked, and grouped. Interests of the subscriber are predicted based on the location information. One or more audiovisual content items associated with the predicted interest are selected. A determination regarding the predicted programming of interest being available is made. Item recommendations are generated for available ones of the selected audiovisual content items. A referral of the one or more item recommendations is presented to the subscriber.

Abstract: Methods, apparatus, and systems to collect audience measurement data are disclosed. An example system includes at least one non-transitory machine readable storage medium including instructions which, when executed, cause a machine to at least: generate behavior data developed during a first time period based on first media data and user data corresponding to one or more users of a household, the user data to include demographic information for the one or more users associated with the household, identify second media data during a second time period different than the first time period, the second media data identified without identification of the one or more users of the household, and associate the demographic information to the second media data based on the behavior data generated during the first time period associated with the one or more users.

Abstract: Methods, systems, and apparatuses for audience metric determination are described herein. An audience segment may be targeted for delivery of content. A clustering algorithm may be used to categorize a quantity of users or devices into subsets based on a propensity to consume, present or output a particular type of content, and a quantity of time to output the particular type of content. A weight may be assigned to each subset based on its relevance to other subsets, such as based on data variance, e.g., on a distance to a midpoint of a specific subset of the subsets. An index parameter may be determined for the datasets, e.g., based on each weight for each subset, and data may be generated that reflects a ranking of content delivery spots for delivery of content to the audience segment.

Abstract: Methods and apparatus to measure audience composition and recruit audience measurement panelists are disclosed. An example media device disclosed herein includes processor circuitry to execute instructions to: detect a first code embedded in a video stream of media presented by the media device, the first code to reference first audience measurement data to identify the media; after detection of the first code, provide a prompt to request input of second audience measurement data; and cause transmission of at least one of the first audience measurement data or the second audience measurement data to a remote data processing facility.

Abstract: A system and method of split rendering for lightfield or immersive media by using an edge-cloud and peer-to-peer based architecture. The system and method include the use of a combination of cloud-based devices and edge-devices to provide distributed processing in connection with the streaming of media, and in particular lightfield or immersive media, to an end user device. The system and method further include the use of multiple cloud and edge devices to provide parallel streaming of a given media package to an end user device.

Abstract: Processor systems to estimate audience sizes and impression counts for different frequency intervals are disclosed. An example processor system includes a memory management unit (MMU) to assign requests from computing devices indicative of accesses to media to a first block of memory and to assign user-identified impression data corresponding to user-identified impressions to a second block of memory. The processor system including an arithmetic logic unit (ALU) to determine multipliers relating a first probability distribution for the user-identified impressions to a second probability distribution for census impressions. The multipliers based on probability constraints defined by weighted probabilities associated with the second probability distribution, where different ones of the probabilities weighted based on estimated populations for corresponding ones of the different demographics.

Abstract: An example apparatus includes at least one memory, instructions, and at least one processor to execute the instructions to generate electronic mobile measurement data based on network communications received from first client devices, select attributes associated with the electronic mobile measurement data to include in a model, generate the model based on the attributes and a first portion of the electronic mobile measurement data, determine a percentage of a second portion of the electronic mobile measurement data that the model correctly associates with corresponding first users of the first client devices, and when the percentage satisfies a threshold determine: (a) when a second user operating a second client device is a primary user, and (b) when the user operating the second client device is a third user, and associate demographic information of the second user with the electronic mobile measurement data to reduce a misattribution error.

Abstract: Mechanisms for inhibiting the transmission of media content based on frequency and exposure measurements are provided.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed for editing tuning data collected via automated content recognition. Examples include determining whether a time conflict exists between first tuning data corresponding to a first tuning event and second tuning data corresponding to a second tuning event. Examples also include that, in response to determining that the time conflict exists, creating a third tuning event based on the first tuning data, the second tuning data, and one or more criteria. Examples also include that modifying at least one of the first tuning event or the second tuning event based on the third tuning event. Examples also include that crediting a media presentation by the presentation device based on edited tuning data, the edited tuning data including the first modified tuning event, the second modified tuning event, and the third tuning event.

Abstract: Methods and systems are proposed herein to address the inefficiencies in the recommendation schemes available. More specifically, the embodiments disclosed herein provide a new recommendation scheme, whereby a user enters various criteria for what he/she would like to be recommended. For example, the system may receive a user preference for content recommendations from a user. The system may retrieve a user profile for the user. The system may compare the user preference to the user profile to determine a criterion for content recommendations for the user. The system may receive a content attribute for content provided by a content provider. The system may match the criterion to the content attribute. The system may, in response to matching the criterion to the content attribute, generate for display a recommendation to the user for the content.

Abstract: Methods and systems are provided for generating a program guide that simultaneously displays future content item identifiers and past content item identifiers. The future content items will be available from a content source in future time slots. The future content item identifiers are sized according to the lengths of corresponding future content items. The past content item that were available from the content source in past time slots. The past content item identifiers are not sized according to the lengths of corresponding past content items.

Abstract: Systems and methods are provided herein for determining reduction of interest in a content series and to increasing the interest upon such determination. This may be accomplished by a device monitoring consumption of a content series to determine a pattern of consumption. The device may identify a change in the pattern of consumption indicative of a reduction of interest and determine a reason for the reduction in interest. Based on the reason for the reduction of interest, the device may provide an operation, such as a spoiler, to increase interest in the content series.

Abstract: Methods and apparatus to detect user attentiveness to portable devices are disclosed. An example portable device includes memory, computer readable instructions, and processor circuitry to execute the computer readable instructions to at least download an exposure measurement application from a first server via a network, and execute the exposure measurement application to detect at least one of a first orientation change of the portable device or a first position change between the portable device and a user, compare the at least one of the first orientation change or the first position change to a plurality of spatial condition change combinations associated with respective likelihoods indicative of user attentiveness related to the portable device to determine user attentiveness data associated with a presentation on the portable device, and cause the user attentiveness data to be transmitted via the network.

Abstract: This application provides a gesture information processing method and apparatus, an electronic device, and a storage medium. The method includes: determining an electromyography signal collection target object in a gesture information usage environment; dividing the electromyography signal sample through a sliding window having a fixed window value and a fixed stride into different electromyography signals of the target object, and denoising the different electromyography signals of the target object; recognizing the denoised different electromyography signals, and determining probabilities of gesture information represented by the different electromyography signals; and weighting the probabilities of the gesture information represented by the different electromyography signals, so as to determine gesture information matching the target object.

Abstract: A flexible approach to segmenting a resource (e.g., a media resource, such as a media segment, or other resource, such as a resource normally fetched or pushed using general file transfer protocols like HTTP) into a plurality of fragments. By employing such an approach, the delay until the resource can be utilized at the client side is reduced. Certain embodiments are provided which apply the flexible segmentation approach to ISOBMFF media segments for video streaming, such as would be used with Live DASH streaming.

Abstract: An electronic apparatus includes a processor. The processor is configured to: obtain information about weighted values respectively assigned to a plurality of items having values that represent content features of a plurality of pieces of content, wherein the weighted values are assigned as higher weight values to items corresponding to the content features preferred by according to genres of content; identify a user content viewed by a user of the electronic apparatus; calculate similarities in values of the plurality of items, respectively, between a plurality of pieces of recommendable content and the user content, respectively; calculate recommendation scores of the plurality of pieces of content, based on the calculated similarities and the weighted values assigned to the items according to the genres; and perform a recommendation operation for one or more content pieces, of which the calculated recommendation score is equal to or higher than a preset ranking.

Abstract: Techniques for generating an interactive and dynamically updated content stream are described herein. Data assets that correspond to video segments provided by a plurality of content providers may be maintained. A video segment may correspond to a first portion of a full video segment for a piece of content from the plurality of content providers. A content stream may be generated that includes video segments provided by the plurality of content providers based at least in part on the data assets. The content stream may be presented via an application. Second input may be received via the application during presentation of the content stream that corresponds to a navigation command for displaying different content. The content stream may be updated to at least one of removing certain video segments or adding new video segments to the content stream based at least in part on the second input.

Abstract: Systems and methods for tuning hyperparameters of a model includes: receiving a multi-criteria tuning work request for tuning hyperparameters of the model of the subscriber to the remote tuning service, wherein the multi-criteria tuning work request includes: a first objective function of the model to be optimized by the remote tuning service; a second objective function to be optimized by the remote tuning service, the second objective function being distinct from the first objective function; computing a joint tuning function based on a combination of the first objective function and the second objective function; executing a tuning operation of the hyperparameters for the model based on a tuning of the joint function; and identifying one or more proposed hyperparameter values based on one or more hyperparameter-based points along a convex Pareto optimal curve.

Abstract: A computer-implemented method of using video program viewer interaction data that has been loaded to a media measurement database as input to a measurement engine which then calculates Linear, DVR, and VOD asset viewing activity at three levels: (a) Video Program, (b) Video Program Airing, (c) Video Program Airing Segment, where each level provides summary metrics for groupings of Demographic, Geographic, and/or Device Characteristic, and also second-by-second viewing metrics, including counting advertising impressions, within the Demographic, Geographic, Device groupings. System also accounts for reduced value of ad viewing when viewer is using trick plays or when viewer delays viewing recorded content. Together these metrics provide detailed information on customer viewing behavior which can be used to drive business decisions for service providers, advertisers and content producers. Additionally, a viewing histogram analysis is produced.

Abstract: Systems and methods are disclosed herein for determining a preference of a user for or against a participant of an event when a media asset relating to the event is being played based on monitoring user reactions to the event. Specifically, the media guidance application may detect a change of at least one of the biometric state (e.g., pulse rate, blood pressure, etc.) and physiological state (e.g., facial expression, gesture, body movement, etc.) with a user. The media guidance application may determine that an event has occurred within the sporting event that triggered the change. The media guidance application may determine that the user favors the first participant over the second participant, if the user reacts positively towards a score for the first participant in the event.

Abstract: A method, systems and apparatuses for preventing a user abandonment from a browser window are disclosed. In an exemplary embodiment, abandonment prevent on a user mobile device comprises pushing a placeholder variable onto a history object of the browser window while the user is on a first website. The user may be directed to the first website from a referrer. User may navigate backwards using a device interface element, such as a back button. When user navigation triggers a popstate, and the history for the browser is null, the user is directed to a preferred website.

Abstract: In one aspect, a method includes (i) accessing broadcast-schedule data associated with a channel; (ii) using the accessed broadcast-schedule data to identify an upcoming content-modification opportunity on the channel; (iii) responsive to identifying the upcoming content-modification opportunity on the channel, identifying a content-presentation device tuned to the channel; and (iv) causing supplemental content to be transmitted to the identified content-presentation device, to facilitate the identified content-presentation device performing a content-modification operation related to the identified upcoming content-modification opportunity on the channel.

Abstract: A display apparatus displaying content corresponding to an installation location and a server providing content considering the installation location of the display apparatus are provided. The display apparatus includes: a display; a communication interface configured to communicate with an external server; and a processes configured to control the communication interface to receive, from the external server, content including an element with a large number of expected viewers identified based on at least one of a number of viewers or an installation location, and control the display to display the received content.

Abstract: Systems, methods, and non-transitory computer-readable media can obtain a plurality of input signals describing user activity associated with a plurality of entities. A respective amount of surge in user activity with each of the plurality of entities can be determined based on the plurality of input signals. The plurality of entities can be ranked based on their surges in user activity. Information describing one or more of the ranked entities can be provided as recommendations to one or more users.

Abstract: In some embodiments, a method receives a message that identifies an asset being viewed on a video service at a timestamp. The message is associated with a profile identifier and a plurality of messages are received for the profile identifier while using the video service. The method determines a channel identifier for a channel in which the asset is offered during the timestamp and determines whether a channel change occurred based on the channel identifier and a prior channel identifier determined from a prior message in the plurality of messages. When the channel change has occurred, the method stores the channel identifier and timestamp for the profile identifier in a data structure, wherein a list of channels is generated by querying the data structure to determine a set of channel identifiers and timestamps for the profile identifier.

Abstract: Techniques where a computing device such as a smart phone can be leveraged to facilitate ratings data capture through a set top box (“STB”) associated with a content subscription service such as satellite television are described. In an exemplary embodiment, STB channel data can be communicated to the computing device via a personal area network (“PAN”) connection such as Bluetooth. The computing device in turn can forward the STB channel data to a remote server over a wide area network (“WAN”).

Abstract: The present disclosure relates to a medical imaging method for enabling magnetic resonance imaging of a subject (318) using a set of imaging parameters of imaging protocols, the method comprising: receiving information related to the subject; using a predefined machine learning model for suggesting at least one imaging protocol for the received information, wherein the imaging protocol comprises at least part of the set of imaging parameters and associated values; providing the imaging protocol.

Abstract: A facility for generating and displaying information regarding the packaging of individual media asset files associated with multiple scheduled presentations of interstitial media assets is provided. Scheduling information is received regarding future scheduled media presentations for each of a plurality of media assets via one or more content channels. Based at least in part on the received scheduling information, one or more package files are generated such that each package file includes a plurality of media asset files, each corresponding to at least one of the future scheduled media presentations, for distribution to a plurality of distinct media receivers. Database information is generated regarding the generating of the one or more package files, such that the database information includes a completion time associated with the generating of each package file.

Abstract: Methods, systems, and media for identifying abusive user accounts based on playlists are provided. In accordance with some embodiments of the disclosed subject matter, a method for identifying abusive content is provided, the method comprising: determining at least one feature associated with a playlist created by a user-generated channel; calculating a playlist score associated with the playlist based on a playlist classifier, wherein the playlist classifier comprises a function that maps the at least one feature to the playlist score; calculating a channel score associated with the user-generated channel based at least on the calculated playlist score; determining that one or more content items associated with the user-generated channel is to be demoted based on the calculated channel score, wherein the one or more content items comprises the playlist; and causing the one or more content items to be demoted.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed for a window-metric rule for return path data (RPD). In some examples, a media monitor includes a count calculator to determine a first count of a number of devices in a first household that reported RPD for a first day. The example media monitor also includes an RPD selector to compare the first count to a current window-metric, the current window-metric associated with a window of time for the first household, the window of time associated with a content provider and having N number of days. The example RPD selector the RPD from the first household for the first day when the first count does not satisfy the current window-metric. The example media monitor also includes a report generator to generate a media exposure report using the RPD from the first household for the first day when the first count satisfies the current window-metric, independent of if the first household has heartbeat data available.

Abstract: Methods, apparatus, systems and articles of manufacture to identify media sources are disclosed. Example apparatus disclosed herein include an exclusivity determiner, a unique asset identifier, and a signature matcher. The exclusivity determiner is to determine whether media is exclusive to a provider. The unique asset identifier is to associate a first signature of the media with the provider if the media is exclusive to the provider. The signature matcher is to identify the provider based on a second signature, the second signature matching the first signature, the second signature extracted from media presented at a media presentation location.

Abstract: A method, set-top box, and non-transitory computer readable medium are disclosed for presentation of audiovisual content with closed captions. The method includes receiving, via an input device interfaced with the electronic device, an instruction requesting a replay of previously viewed video content with closed captioning; sending, to the display device interfaced with the electronic device, one or more thumbnail images of the previously viewed video content to be displayed on the display device; receiving, via the input device interfaced with the electronic device, one of the one or more thumbnail images of the previously viewed video content being selected for replay of the previously viewed video content; and sending, to the display device interfaced with the electronic device, closed captioning with the previously viewed video content starting at a video frame corresponding to the one of the one or more thumbnails of the previously viewed video content selected for replay.

Abstract: Methods, apparatus, and systems to collect audience measurement data are disclosed. Disclosed example systems include first means for processing to collect (i) first media identification data corresponding to a first time period and (ii) second media identification data corresponding to a second time period after the first time period, the first media identification data and the second media identification data associated with a monitored location and second means for processing to develop behavior data based on the first media identification data and people meter data collected during the first time period with a people meter associated with the monitored location, the people meter to be deactivated during the second time period and identify an audience member associated with the second media identification data based on the behavior data and without access to any people meter data corresponding to the second time period.

Abstract: Systems and methods are described for mitigating errors introduced during processing of user input such as voice input. A query may be derived from processed user input. A performance predictor analyzes the query and uses historical data to predict whether the query will return relevant results if executed. If the query's predicted performance is below a threshold, a query rewriter may identify potential alternatives to the query from a library of “known good” queries. Different analyzers may be applied to identify different sets of alternatives, and machine learning models may be applied to rank the outputs of the analyzers. The best-matching alternatives from each analyzer may then be provided as inputs to a further machine learning model, which assesses the probability that each of the identified alternatives reflects the intent of the user. A most likely alternative may then be selected to execute in place of the original query.