Abstract: Natural language query translation may be provided. A statistical model may be trained to detect domains according to a plurality of query click log data. Upon receiving a natural language query, the statistical model may be used to translate the natural language query into an action. The action may then be performed and at least one result associated with performing the action may be provided.

Abstract: A technique is described for generating a knowledge graph that links names associated with a first subject matter category (C1) (such as brands) with names associated with a second subject matter category (C2) (such as products). In one implementation, the technique relies on two similarly-constituted processing pipelines, a first processing pipeline for processing the C1 names, and a second processing pipeline for processing the C2 names. Each processing pipeline includes three main stages, including a name-generation stage, a verification stage, and an augmentation stage. The generation stage uses a voting strategy to form an initial set of seed names. The verification stage removes noisy seed names. And the augmentation stage expands each verified name to include related terms. A final edge-forming stage identifies relationships between the expanded C1 names and the expanded C2 names using a voting strategy.

Abstract: A technique is described for generating a knowledge graph that links names associated with a first subject matter category (C1) (such as brands) with names associated with a second subject matter category (C2) (such as products). In one implementation, the technique relies on two similarly-constituted processing pipelines, a first processing pipeline for processing the C1 names, and a second processing pipeline for processing the C2 names. Each processing pipeline includes three main stages, including a name-generation stage, a verification stage, and an augmentation stage. The generation stage uses a voting strategy to form an initial set of seed names. The verification stage removes noisy seed names. And the augmentation stage expands each verified name to include related terms. A final edge-forming stage identifies relationships between the expanded C1 names and the expanded C2 names using a voting strategy.

Abstract: Examples of the present disclosure describe systems and methods for suggested content generation. In an example, a publisher may provide content to a user during a browsing session of the user. A domain associated with the browsing session may be identified, such that suggested content relating to the browsing session may be provided for display in addition to the provided content. The suggested content may omit content provided by the publisher, so as to avoid providing redundant content or reducing user interest in the content provided by the publisher. As a result, the user may interact with the suggested content when progressing through a browsing session rather than navigating away from the publisher. This may enable the publisher to gain insight into the browsing session of the user while also improving the browsing session of the user by making relevant content more easily accessible to the user.

Abstract: Natural language query translation may be provided. A statistical model may be trained to detect domains according to a plurality of query click log data. Upon receiving a natural language query, the statistical model may be used to translate the natural language query into an action. The action may then be performed and at least one result associated with performing the action may be provided.

Abstract: Natural language query translation may be provided. A statistical model may be trained to detect domains according to a plurality of query click log data. Upon receiving a natural language query, the statistical model may be used to translate the natural language query into an action. The action may then be performed and at least one result associated with performing the action may be provided.

Abstract: Click logs are automatically mined to assist in discovering candidate variations for named entities. The named entities may be obtained from one or more sources and include an initial list of named entities. A search may be performed within one or more search engines to determine common phrases that are used to identify the named entity in addition to the named entity initially included in the named entity list. Click logs associated with results of past searches are automatically mined to discover what phrases determined from the searches are candidate variations for the named entity. The candidate variations are scored to assist in determining the variations to include within an understanding model. The variations may also be used when delivering responses and displayed output in the SLU system. For example, instead of using the listed named entity, a popular and/or shortened name may be used by the system.

Abstract: Natural language query translation may be provided. A statistical model may be trained to detect domains according to a plurality of query click log data. Upon receiving a natural language query, the statistical model may be used to translate the natural language query into an action. The action may then be performed and at least one result associated with performing the action may be provided.

Abstract: Click logs are automatically mined to assist in discovering candidate variations for named entities. The named entities may be obtained from one or more sources and include an initial list of named entities. A search may be performed within one or more search engines to determine common phrases that are used to identify the named entity in addition to the named entity initially included in the named entity list. Click logs associated with results of past searches are automatically mined to discover what phrases determined from the searches are candidate variations for the named entity. The candidate variations are scored to assist in determining the variations to include within an understanding model. The variations may also be used when delivering responses and displayed output in the SLU system. For example, instead of using the listed named entity, a popular and/or shortened name may be used by the system.

Abstract: Natural language query translation may be provided. A statistical model may be trained to detect domains according to a plurality of query click log data. Upon receiving a natural language query, the statistical model may be used to translate the natural language query into an action. The action may then be performed and at least one result associated with performing the action may be provided.

Abstract: Within the field of computing, many scenarios involve queries formulated by users resulting in query results presented by a device. The user may request to adjust the query, but many devices can only process requests specified in a well-structured manner, such as a set of recognized keywords, specific verbal commands, or a specific manual gesture. The user thus communicates the adjustment request in the constraints of the device, even if the query is specified in a natural language. Presented herein are techniques for enabling users to specify query adjustments with natural action input (e.g., natural-language speech, vocal inflection, and natural manual gestures). The device may be configured to evaluate the natural action input, identify the user's intended query adjustments, generate an adjusted query, and present an adjusted query result, thus enabling the user to interact with the device in a similar manner as communicating with an individual.

Abstract: A system for determining predictive models associated with online advertising can include a communications interface, a processor, and a display. The communications interface can be configured to receive a partial dataset. The partial dataset may include user information. The processor can be communicatively coupled to the communications interface and configured to identify the partial dataset. The processor can also be configured to determine a first predictive model corresponding to at least part of the partial dataset and a second predictive model by combining a probability distribution with the first predictive model. The display can be communicatively coupled to the processor and configured to display the second predictive model.

Abstract: A method for combining multiple probability of click models in an online advertising system into a combined predictive model, the method commencing by receiving a feature set slice (e.g. corresponding to demographics or taxonomies or clusters), and using the sliced data for training multiple slice-wise predictive models. The trained slice-wise predictive models are combined by overlaying a weighted distribution model over the trained slice-wise predictive models. The combined predictive model then is used in predicting the probability of a click given a query-advertisement pair in online advertising. The method can flexibly receive slice specifications, and can overlay any one or more of a variety of distribution models, such as a linear combination or a log-linear combination.

Abstract: A machine-learning method for estimating probability of a click event in online advertising systems by computing and comparing an aggregated predictive model (a global model) and one or more data-wise sliced predictive models (local models). The method comprises receiving training data having a plurality of features stored in a feature set and constructing a global predictive model that estimates the probability of a click event for the processed feature set. Then, partitioning the global predictive model into one or more data-wise sliced training sets for training a local model from each of the data-wise slices, and then determining whether a particular local model estimates probability of click event for the feature set better than the global model. A given feature set may be collected from historical data, and may comprise a feature vector for a plurality of query-advertisement pairs and a corresponding indicator that represents a click on the advertisement.

Abstract: Training for a statistical dialog manager may be provided. A plurality of log data associated with an intent may be received, and at least one step associated with completing the intent according to the plurality of log data may be identified. An understanding model associated with the intent may be created, including a plurality of queries mapped to the intent. In response to receiving a natural language query from a user that is associated with the intent a response to the user may be provided according to the understanding model.

Abstract: Sponsored search advertising utilizes a click probability as one factor in selecting and ranking advertisements that are displayed with search results. The probability of click may also be referred to as a predicted click-through rate (“CTR”) that may be multiplied by an advertiser's bid for a particular advertisement to rank the display of advertisements. An accurate prediction of the click probability improves the potential revenue that is generated by advertisements in a pay per click system. Other advertising systems may benefit from an accurate and reliable estimate for an advertisement's probability of click in different environments and scenarios.

Abstract: A machine-learning method for estimating probability of a click event in online advertising systems by computing and comparing an aggregated predictive model (a global model) and one or more data-wise sliced predictive models (local models). The method comprises receiving training data having a plurality of features stored in a feature set and constructing a global predictive model that estimates the probability of a click event for the processed feature set. Then, partitioning the global predictive model into one or more data-wise sliced training sets for training a local model from each of the data-wise slices, and then determining whether a particular local model estimates probability of click event for the feature set better than the global model. A given feature set may be collected from historical data, and may comprise a feature vector for a plurality of query-advertisement pairs and a corresponding indicator that represents a click on the advertisement.

Abstract: A method for combining multiple probability of click models in an online advertising system into a combined predictive model, the method commencing by receiving a feature set slice (e.g. corresponding to demographics or taxonomies or clusters), and using the sliced data for training multiple slice-wise predictive models. The trained slice-wise predictive models are combined by overlaying a weighted distribution model over the trained slice-wise predictive models. The combined predictive model then is used in predicting the probability of a click given a query-advertisement pair in online advertising. The method can flexibly receive slice specifications, and can overlay any one or more of a variety of distribution models, such as a linear combination or a log-linear combination.

Abstract: Sponsored search advertising utilizes a click probability as one factor in selecting and ranking advertisements that are displayed with search results. The probability of click may also be referred to as a predicted click-through rate (“CTR”) that may be multiplied by an advertiser's bid for a particular advertisement to rank the display of advertisements. An accurate prediction of the click probability improves the potential revenue that is generated by advertisements in a pay per click system. Other advertising systems may benefit from an accurate and reliable estimate for an advertisement's probability of click in different environments and scenarios.

Abstract: An improved system and method for identifying context-dependent term importance of queries is provided. A query term importance model is learned using supervised learning of context-dependent term importance for queries and is then applied for advertisement prediction using term importance weights of query terms as query features. For instance, a query term importance model for query rewriting may predict rewritten queries that match a query with term importance weights assigned as query features. Or a query term importance model for advertisement prediction may predict relevant advertisements for a query with term importance weights assigned as query features. In an embodiment, a sponsored advertisement selection engine selects sponsored advertisements scored by a query term importance engine that applies a query term importance model using term importance weights as query features and inverse document frequency weights as advertisement features to assign a relevance score.