Abstract: A method for generating inference graphs over content to answer input inquiries. First, independent factors are produced from the inquiry, and these factors are converted to questions. The questions are then input to a probabilistic question answering system (PQA) that discovers relations which are used to iteratively expand an inference graph starting from the factors and ending with possible answers. A probabilistic reasoning system is used to infer the confidence in each answer by, for example, propagating confidences across relations and nodes in the inference graph as it is expanded. The inference graph generator system can be used to simultaneously bi-directionally generate forward and backward inference graphs that uses a depth controller component to limit the generation of both paths if they do not meet. Otherwise, a joiner process forces the discovery of relations that join the answers to factors in the inquiry.

Abstract: A method of generating secondary questions in a question-answer system. Missing information is identified from a corpus of data using a computerized device. The missing information comprises any information that improves confidence scores for candidate answers to a question. The computerized device automatically generates a plurality of hypotheses concerning the missing information. The computerized device automatically generates at least one secondary question based on each of the plurality of hypotheses. The hypotheses are ranked based on relative utility to determine an order in which the computerized device outputs the at least one secondary question to external sources to obtain responses.

Abstract: A method of generating secondary questions in a question-answer system. Missing information is identified from a corpus of data using a computerized device. The missing information comprises any information that improves confidence scores for candidate answers to a question. The computerized device automatically generates a plurality of hypotheses concerning the missing information. The computerized device automatically generates at least one secondary question based on each of the plurality of hypotheses. The hypotheses are ranked based on relative utility to determine an order in which the computerized device outputs the at least one secondary question to external sources to obtain responses.

Abstract: A method for generating inference graphs over content to answer input inquiries. First, independent factors are produced from the inquiry, and these factors are converted to questions. The questions are then input to a probabilistic question answering system (PQA) that discovers relations which are used to iteratively expand an inference graph starting from the factors and ending with possible answers. A probabilistic reasoning system is used to infer the confidence in each answer by, for example, propagating confidences across relations and nodes in the inference graph as it is expanded. The inference graph generator system can be used to simultaneously bi-directionally generate forward and backward inference graphs that uses a depth controller component to limit the generation of both paths if they do not meet. Otherwise, a joiner process forces the discovery of relations that join the answers to factors in the inquiry.

Abstract: A method, system and computer program product for generating inference graphs over content to answer input inquiries. First, independent factors are produced from the inquiry, and these factors are converted to questions. The questions are then input to a probabilistic question answering system (PQA) that discovers relations which are used to iteratively expand an inference graph starting from the factors and ending with possible answers. A probabilistic reasoning system is used to infer the confidence in each answer by, for example, propagating confidences across relations and nodes in the inference graph as it is expanded. The inference graph generator system can be used to simultaneously bi-directionally generate forward and backward inference graphs that uses a depth controller component to limit the generation of both paths if they do not meet. Otherwise, a joiner process forces the discovery of relations that may join the answers to factors in the inquiry.

Abstract: A method for automatic detection of user preferences for alternate user interface model includes operating a digital device with an active user interface model and receiving one or more input signals from a user of the digital device. The method includes comparing the one or more input signals with one or more latent user interface models and determining if one of the latent user interface models has a higher likelihood given the one or more input signals than the active user interface models. The method also includes responsively substituting the latent user interface with the highest likelihood given the one or more input signals for the active user interface model.

Abstract: According to an aspect, a cyclical directed graphical model that includes a plurality of nodes connected by directed edges is accessed. Each directed edge has an edge strength indicating a degree of relation between the nodes, and each node has a value and a state of either active or inactive. A first node that has at least one directed incoming edge from a second node having a state of active is selected. At least a subset of the nodes in the cyclical directed graphical model are traversed in a reverse direction relative to the directed edges. The traversing is performed along a path starting from the second node and ending at a third node having a state of active. If the path does not include the first node, an inference about a value of the first node is proposed based on a value of the second node.

Abstract: According to an aspect, a cyclical directed graphical model that includes a plurality of nodes connected by directed edges is accessed. Each directed edge has an edge strength indicating a degree of relation between the nodes, and each node has a value and a state of either active or inactive. A first node that has at least one directed incoming edge from a second node having a state of active is selected. At least a subset of the nodes in the cyclical directed graphical model are traversed in a reverse direction relative to the directed edges. The traversing is performed along a path starting from the second node and ending at a third node having a state of active. If the path does not include the first node, an inference about a value of the first node is proposed based on a value of the second node.

Abstract: According to an aspect, influence filtering in a graphical model includes accessing the graphical model in a data store. The graphical model includes a plurality of nodes connected by edges having edge strengths representing a degree of relation between the nodes. A target relation strength for a pair of nodes in the graphical model is received. An edge strength of an edge in a direct path between the pair of nodes is determined by traversing, in the graphical model, one or more paths other than the direct path between the pair of nodes. The determining also includes estimating a cumulative strength of the traversed paths, and calculating the edge strength for the edge in the direct path based on the cumulative strength of the traversed paths and the target relation strength. The calculated edge strength is assigned to the edge in the direct path between the two nodes.

Abstract: According to an aspect, influence filtering in a graphical model includes accessing the graphical model in a data store. The graphical model includes a plurality of nodes connected by edges having edge strengths representing a degree of relation between the nodes. A target relation strength for a pair of nodes in the graphical model is received. An edge strength of an edge in a direct path between the pair of nodes is determined by traversing, in the graphical model, one or more paths other than the direct path between the pair of nodes. The determining also includes estimating a cumulative strength of the traversed paths, and calculating the edge strength for the edge in the direct path based on the cumulative strength of the traversed paths and the target relation strength. The calculated edge strength is assigned to the edge in the direct path between the two nodes.

Abstract: A method for automatic detection of user preferences for alternate user interface model includes operating a digital device with an active user interface model and receiving one or more input signals from a user of the digital device. The method includes comparing the one or more input signals with one or more latent user interface models and determining if one of the latent user interface models has a higher likelihood given the one or more input signals than the active user interface models. The method also includes responsively substituting the latent user interface with the highest likelihood given the one or more input signals for the active user interface model.

Abstract: A method for automatic detection of user preferences for alternate user interface model includes operating a digital device with an active user interface model and receiving one or more input signals from a user of the digital device. The method includes comparing the one or more input signals with one or more latent user interface models and determining if one of the latent user interface models has a higher likelihood given the one or more input signals than the active user interface models. The method also includes responsively substituting the latent user interface with the highest likelihood given the one or more input signals for the active user interface model.

Abstract: According to an aspect, influence filtering in a graphical model includes accessing the graphical model in a data store. The graphical model includes a plurality of nodes connected by edges having edge strengths representing a degree of relation between the nodes. A target relation strength for a pair of nodes in the graphical model is received. An edge strength of an edge in a direct path between the pair of nodes is determined by traversing, in the graphical model, one or more paths other than the direct path between the pair of nodes. The determining also includes estimating a cumulative strength of the traversed paths, and calculating the edge strength for the edge in the direct path based on the cumulative strength of the traversed paths and the target relation strength. The calculated edge strength is assigned to the edge in the direct path between the two nodes.

Abstract: According to an aspect, a cyclical directed graphical model that includes a plurality of nodes connected by directed edges is accessed. Each directed edge has an edge strength indicating a degree of relation between the nodes, and each node has a value and a state of either active or inactive. A first node that has at least one directed incoming edge from a second node having a state of active is selected. At least a subset of the nodes in the cyclical directed graphical model are traversed in a reverse direction relative to the directed edges. The traversing is performed along a path starting from the second node and ending at a third node having a state of active. If the path does not include the first node, an inference about a value of the first node is proposed based on a value of the second node.

Abstract: According to an aspect, influence filtering in a graphical model includes accessing the graphical model in a data store. The graphical model includes a plurality of nodes connected by edges having edge strengths representing a degree of relation between the nodes. A target relation strength for a pair of nodes in the graphical model is received. An edge strength of an edge in a direct path between the pair of nodes is determined by traversing, in the graphical model, one or more paths other than the direct path between the pair of nodes. The determining also includes estimating a cumulative strength of the traversed paths, and calculating the edge strength for the edge in the direct path based on the cumulative strength of the traversed paths and the target relation strength. The calculated edge strength is assigned to the edge in the direct path between the two nodes.

Abstract: According to an aspect, a cyclical directed graphical model that includes a plurality of nodes connected by directed edges is accessed. Each directed edge has an edge strength indicating a degree of relation between the nodes, and each node has a value and a state of either active or inactive. A first node that has at least one directed incoming edge from a second node having a state of active is selected. At least a subset of the nodes in the cyclical directed graphical model are traversed in a reverse direction relative to the directed edges. The traversing is performed along a path starting from the second node and ending at a third node having a state of active. If the path does not include the first node, an inference about a value of the first node is proposed based on a value of the second node.

Abstract: A method, system and computer program product for generating inference graphs over content to answer input inquiries. First, independent factors are produced from the inquiry, and these factors are converted to questions. The questions are then input to a probabilistic question answering system (PQA) that discovers relations which are used to iteratively expand an inference graph starting from the factors and ending with possible answers. A probabilistic reasoning system is used to infer the confidence in each answer by, for example, propagating confidences across relations and nodes in the inference graph as it is expanded. The inference graph generator system can be used to simultaneously bi-directionally generate forward and backward inference graphs that uses a depth controller component to limit the generation of both paths if they do not meet. Otherwise, a joiner process forces the discovery of relations that may join the answers to factors in the inquiry.

Abstract: A method, system and computer program product for generating inference graphs over content to answer input inquiries. First, independent factors are produced from the inquiry, and these factors are converted to questions. The questions are then input to a probabilistic question answering system (PQA) that discovers relations which are used to iteratively expand an inference graph starting from the factors and ending with possible answers. A probabilistic reasoning system is used to infer the confidence in each answer by, for example, propagating confidences across relations and nodes in the inference graph as it is expanded. The inference graph generator system can be used to simultaneously bi-directionally generate forward and backward inference graphs that uses a depth controller component to limit the generation of both paths if they do not meet. Otherwise, a joiner process forces the discovery of relations that may join the answers to factors in the inquiry.

Abstract: A method of generating secondary questions in a question-answer system. Missing information is identified from a corpus of data using a computerized device. The missing information comprises any information that improves confidence scores for candidate answers to a question. The computerized device automatically generates a plurality of hypotheses concerning the missing information. The computerized device automatically generates at least one secondary question based on each of the plurality of hypotheses. The hypotheses are ranked based on relative utility to determine an order in which the computerized device outputs the at least one secondary question to external sources to obtain responses.

Abstract: A method of generating secondary questions in a question-answer system. Missing information is identified from a corpus of data using a computerized device. The missing information comprises any information that improves confidence scores for candidate answers to a question. The computerized device automatically generates a plurality of hypotheses concerning the missing information. The computerized device automatically generates at least one secondary question based on each of the plurality of hypotheses. The hypotheses are ranked based on relative utility to determine an order in which the computerized device outputs the at least one secondary question to external sources to obtain responses.