Abstract: The subject technology receives session data from a set of users associated with a retail website. The subject technology generates, using at least one machine learning model, a purchase intent prediction based at least in part on the session data. The subject technology determines a set of sessions that correspond to conversions on the retail website based on the session data. The subject technology generates a combination of data using first data from the purchase intent prediction with second data based on the set of sessions that correspond to conversions. The subject technology generates information related to a missed purchase segment based at least in part on the combination of data. The subject technology provides the information for display on a client device.

Abstract: A monitoring platform may obtain information that identifies a product or service and may collect one or more reviews associated with the product or service from a plurality of sources, wherein each review includes respective review information. The monitoring platform may process the one or more reviews to determine respective additional review information associated with each review of the one or more reviews. The monitoring platform may select, using a machine learning model, a particular review, of the one or more reviews, based on the review information and the additional review information associated with the one or more reviews. The monitoring platform may cause display, on a display of a client device, of a prompt for a response to the particular review and may obtain the response from the client device. The monitoring platform may cause the response to be posted to a source associated with the particular review.

Abstract: A method for extracting a dam emergency event based on a dual attention mechanism is provided. The method includes: performing data preprocessing, building a dependency graph, building a dual attention network, and filling a document-level argument. The performing data preprocessing includes labeling a dam emergency corpus and encoding sentences. Building a dependency graph includes assisting a model to mine a syntactic relation based on a dependency. Building a dual attention network includes weighing and fusing an attention network based on a graph transformer network (GTN) and capturing key semantic information in the sentence. Filling a document-level argument includes filling a document-level argument by detecting a key sentence and ordering a semantic similarity. The method introduces a dependency and overcomes the long-range dependency problem based on the dual attention mechanism, thus achieving high identification accuracy and reducing a lot of labor costs.

Abstract: Disclosed herein are systems and methods capable of establishing a communication session between a customer and an analyst. The contents of the communication session are analyzed to make recommendations of goods and services to the customer. Otherwise, the communication session may be redirected from one channel, for example, chatting to another channel, for example, voice call, to another analyst. The customer's information and the communication session details are retained, and provided to another analyst before the customer is redirected. Such systems, apparatuses, methods, and computer program products use real-time machine learning scoring algorithm to determine which analyst the customers should be transferred to, and thereby saving a lot of time for both the analysts and the customers, and significantly reduce misroutes by eliminating human errors.

Abstract: Systems and methods are disclosed for electronically detecting and determining a type of inspection to recommend for a property and/or one or more characteristics of the property using input from multiple computing devices, such as image capturing devices, mobile devices, external data sources, internal data sources, and/or other data sources. A computing device may receive, via a communication interface, a plurality of images of a property. The computing device may determine, based on the plurality of images of the property, a score indicative of a level of care of the property. The computing device may determine, based on the plurality of images of the property, an estimate of a value of the property. Based on the determined score indicative of the level of care of the property and the determined estimate of the value of the property, the computing device may determine a recommendation for a type of inspection to perform on the property.

Abstract: A waste service management system, method, and computer program product improve waste services by responding to air quality that is outside and inside a waste service vehicle. A controller monitors at least one exterior air quality sensor that is mounted to, and exposed to air exterior to, the waste service vehicle. The controller monitors at least one interior air quality sensor that is mounted to the waste service vehicle and exposed to air interior to a waste hauling body that is one of carried and attached to the waste service vehicle. The controller compares measurement levels of exterior and interior airborne chemical constituents to one or more corresponding air quality thresholds. The controller presents a notification on a user interface device indicating any exceedance of the air quality threshold and an associated origin of one of the exterior air and the interior air.

Abstract: A task-based learning using task-directed prediction network can be provided. Training data can be received. Contextual information associated with a task-based criterion can be received. A machine learning model can be trained using the training data. A loss function computed during training of the machine learning model integrates the task-based criterion, and minimizing the loss function during training iterations includes minimizing the task-based criterion.

Abstract: Methods, apparatuses, and systems are disclosed for analyzing quality control (QC) strategies that are applied to testing processes an analyte in order to meet an acceptable level of probability of patient harm that could result from incorrect test results. The measure of patient harm takes into account severity of patient harm, as well as its occurrence. Methods include calculating, based on the parameters of the QC strategies and the test apparatus, an expected number of incorrect final results E(Nuf) due to a test system failure. The value of E(Nuf) can be used as part of a calculation of a predicted level of probability patient harm. The ratio of the acceptable level of probability of patient harm to the predicted level of probability patient harm can determine the adequacy of the QC strategies.

Abstract: A system and method for real-time geophysical social grouping comprising customer profiles and venue profiles, wherein the profiles comprise expressed and inferred attributes, and a social grouping and recommendation server which utilizes machine learning algorithms on the profiles to generate recommendations for social group pairing, venues, and activities. Attribute matching provides optimized grouping between customers who share certain commonalities while also providing venues a system for locating and attracting ideal customers. Machine learning algorithms may be used to analyze profile attributes and identify patterns of commonality that would not otherwise be recognized. This system allows patrons to meet, dine, and socialize with one or more matched individuals at a venue that satisfies all participants preferences and attributes.

Abstract: Disclosed herein are systems and methods capable of establishing a communication session between a customer and an analyst. The contents of the communication session are analyzed to make recommendations of goods and services to the customer. Otherwise, the communication session may be redirected from one channel, for example, chatting to another channel, for example, voice call, to another analyst. The customer's information and the communication session details are retained, and provided to another analyst before the customer is redirected. Such systems, apparatuses, methods, and computer program products use real-time machine learning scoring algorithm to determine which analyst the customers should be transferred to, and thereby saving a lot of time for both the analysts and the customers, and significantly reduce misroutes by eliminating human errors.

Abstract: Systems and methods are provided for using video/still images captured by continuously recording optical sensors mounted on waste collection vehicles used in in the waste collection, disposal and recycling industry for operational and customer service related purposes. A system is provided for auditing the fill status of a customer waste container by a waste services provider during performance of a waste service activity.

Abstract: A method for parsing information from a plurality of product labels using information technology. The method includes obtaining constituent information with an ingredient data platform from text and graphics found on a portion of a label from a plurality of the product labels and assigning base attributes automatically with the ingredient data platform to each piece of the constituent information on at least one of the product labels. The method includes associating the base attributes assigned by the ingredient data platform with different base attributes in at least one pre-constructed taxonomy data structure handled by the ingredient data platform to establish relationships between the base attributes that were previously assigned with the ingredient data platform and the base attributes from the pre-constructed taxonomy data structure.

Abstract: A maintenance support system for a plurality of maintenance-target apparatus in a plant includes a maintenance information delivery unit that delivers maintenance information Im that indicates an operational status of each maintenance-target apparatus to maintenance personnel every preset time T, wherein the maintenance information delivery unit delivers predictive information If to the maintenance personnel at a specified time Tf between the previous delivery of the maintenance information Im and the next delivery of the maintenance information Im, the predictive information If providing predictions regarding the operational status of each maintenance-target apparatus at the next delivery of the maintenance information Im.

Abstract: A monitoring platform may obtain information that identifies a product or service and may collect one or more reviews associated with the product or service from a plurality of sources, wherein each review includes respective review information. The monitoring platform may process the one or more reviews to determine respective additional review information associated with each review of the one or more reviews. The monitoring platform may select, using a machine learning model, a particular review, of the one or more reviews, based on the review information and the additional review information associated with the one or more reviews. The monitoring platform may cause display, on a display of a client device, of a prompt for a response to the particular review and may obtain the response from the client device. The monitoring platform may cause the response to be posted to a source associated with the particular review.

Abstract: A system and method for cost-based migration planning includes determining cost factors and constraints associated with at least one of migrating and transforming a plurality of components. A cost-based model is constructed using the cost factors and constraints. The cost-based model is applied to groupings of the plurality of components to provide a cost-based determination grouping scheme in accordance with the constraints. The grouping scheme is output to provide migration waves for migrating or transforming the plurality of components.

Abstract: A processor may receive, from one or more electronic control units (ECUs) of a particular vehicle, vehicle-system data including information about a state of one or more vehicle-systems of the particular vehicle. Also, the processor may receive, from one or more data sources other than the one or more ECUs, other input data indicative of a state of the particular vehicle. Based on the vehicle-system data and on the other input data, the processor may determine repair information to be specified in a repair order (RO) associated with the particular vehicle. Further, the processor may generate an RO specifying at least the determined repair information, wherein the generated RO is associated with the particular vehicle. Then, the processor may direct a display device to display at least one visual indication representative of the generated RO.

Abstract: A system and method for temporal feasibility analyses that links a service requirement to a source capable of providing a service. In embodiments, the service requirement is received from a requestor, wherein the service requirement defines a type of service, location, and a request to provide the service within a predetermined time period. A repository of sources is analyzed to determine qualified sources that provide service to a region that includes the service location and have a response status greater than a predefined threshold, where the response status indicates a probability of the source being available to provide the service within the time period. A request is sent to each qualified source to provide the service, and responses are then received from at least one of the qualified sources in response to sending the request.

Abstract: Systems and methods are disclosed for electronically detecting and determining a type of inspection to recommend for a property and/or one or more characteristics of the property using input from multiple computing devices, such as image capturing devices, mobile devices, external data sources, internal data sources, and/or other data sources. A computing device may receive, via a communication interface, a plurality of images of a property. The computing device may determine, based on the plurality of images of the property, a score indicative of a level of care of the property. The computing device may determine, based on the plurality of images of the property, an estimate of a value of the property. Based on the determined score indicative of the level of care of the property and the determined estimate of the value of the property, the computing device may determine a recommendation for a type of inspection to perform on the property.

Abstract: An online system allows third-party systems to provide content for distribution to users. The online system allows trademark owners to search for content items that are likely to violate trademarks. The online system receives proof of the trademark registration from trademark owner's devices. The online system receives a search request for content items that are likely to violate a trademark. The online system ranks the content items that are likely to violate the input trademark based on various factors. One of the factors used for ranking the content items is a score generated by a machine learning based model that indicates a likelihood of violation of trademarks by a content item. The online system controls the rate of distribution of content items of a content provider based on information indicating whether the content provider violates trademarks, for example, by blocking distribution of content items by specific content providers.

Abstract: A system, method, and computer-readable medium are disclosed for performing a recommendation operation, comprising: optimizing a product list to provide an optimized product list for use when generating a recommendation for an account; optimizing a neighbor set to provide an optimized neighbor set for use when generating the recommendation for the account; boosting a self-cosine similarity metric to provide a boosted self-cosine similarity metric, the self-cosine similarity metric corresponding to the account; and, providing a recommendation for the account, the recommendation being based on the optimized product list, the optimized neighbor set and the boosted self-cosine similarity metric.