Abstract: A video information extraction system includes a memory to store a video; a textual information extraction module to obtain information about terms, entities, relations, and events from a ground truth caption corresponding to the video; and a video captioning module including an encoder (i) to receive the information about the terms, entities, relations, and events from the textual information extraction module, and (ii) to extract video features from the video; and a decoder to generate a text caption based on the extracted video features.

Abstract: In accordance with various embodiments of the disclosed subject matter, a method and framework configured for modeling a pattern of life (POL) by processing both categorical data and non-categorical data (e.g., numeric, spatial etc.), conducting pattern of life estimation (POLE), and detecting anomalous data in a multi-dimensional data set in a substantially simultaneous manner by comparing statistical PoL results.

Abstract: Disclosed is a method and apparatus for transforming data in vector form. Each vector is composed of a set of attributes that are either boolean or have been mapped to boolean form. The vectors may or may not fall into categories assigned by a subject matter expert (SME). If categories exist, the categorical labels divide the vectors into subsets. The first transformation calculates a prior probability for each attribute based on the links between attributes in each subset of the vectors. The second transformation computes a new numeric value for each attribute based on the links between attributes in each subset of the vectors. The third transformation operates on vectors that have not been categorized. Based on the automatic selection of categories from the attributes, this transformation computes a new numeric value for each attribute based on the links between attributes in each subset of the vectors.

Abstract: A data sharing and indexing system composed of multiple local indexing systems residing on network storage devices. Each (instantiation of the groupware) local storage device runs software for advanced content-based indexing, and each local indexing system forms a node in a peer-to-peer network. The indexing software performs topic-based categorization by means of a higher-order path analysis algorithm, which mimics human intuition by considering both high- and low-order links between data elements. The indexes generated by the software are automatically partitioned into topic indexes. The topical similarity of indexes to each other and to a pre-established set of topic indices is measured using a cross-training algorithm. The peer-to-peer network is implemented by a novel mesh-based, self-healing protocol, providing specialized means for sharing data and topic indexes. The software leverages the index sharing technology to provide content- and metadata-based searching features.

Abstract: An apparatus consisting of a networked data storage device and a software system for automatically indexing the data on the storage device by topic, and for sharing both the data and the topic indexes on the network. These may be used to provide the user of the apparatus with functionality including, but not limited to, content-based searching, sharing and automatic recommendations. The storage device stores its data by means of a file system. The indexing portion of the software performs topic-based categorization by means of a higher-order path analysis algorithm, which mimics human intuition by considering both high- and low-order links between data elements. The indexes generated by the software arc automatically partitioned by topic. The indexing software operates continuously and transparently, updating the index as data is added to or removed from the storage device.

Abstract: A method and system for constructing a social network using advanced content-based indexing, data and index sharing, and a peer-to-peer network in conjunction with a web site interface. The system uses indexing techniques to identify and share common interests among users of the network, and integrated peer-to-peer software facilitates data sharing based on these interests. Users who wish to join the social network download client software from the network's web site. When installed, the software indexes data on the user's local storage devices by considering possible high- and low-order links between data elements. The indexing software generates an index of user data that is partitioned into separate topic indexes. This index represents a cross-section of the user's interests. The user is then able to join various “friend” groups on the network, and to select by topic which portions of the index are shared with which friend groups.

Abstract: Disclosed is a method and apparatus for transforming data in vector form. Each vector is composed of a set of attributes that are either boolean or have been mapped to boolean form. The vectors may or may not fall into categories assigned by a subject matter expert (SME). If categories exist, the categorical labels divide the vectors into subsets. The first transformation calculates a prior probability for each attribute based on the links between attributes in each subset of the vectors. The second transformation computes a new numeric value for each attribute based on the links between attributes in each subset of the vectors. The third transformation operates on vectors that have not been categorized. Based on the automatic selection of categories from the attributes, this transformation computes a new numeric value for each attribute based on the links between attributes in each subset of the vectors.

Abstract: A data sharing and indexing system (for the purpose of forming an imarketplace), which is composed of multiple local indexing systems residing on network storage devices. Each (instantiation of the groupware) local storage device runs software for advanced content-based indexing, and each local indexing system forms a node in a peer-to-peer network. The indexing software performs topic-based categorization by means of a higher-order path analysis algorithm, which mimics human intuition by considering both high- and low-order links between data elements. The indexes generated by the software are automatically partitioned into topic indexes. The topical similarity of indexes to each other and to a pre-established set of topic indices is measured using a cross-training algorithm. The peer-to-peer network is implemented by a novel mesh-based, self-healing protocol, providing specialized means for sharing data and topic indexes.

Abstract: An apparatus consisting of a networked data storage device and a software system for automatically indexing the data on the storage device by topic, and for sharing both the data and the topic indexes on the network. These may be used to provide the user of the apparatus with functionality including, but not limited to, content-based searching, sharing and automatic recommendations. The storage device stores its data by means of a file system. The software executes either on co-processing hardware within the storage device enclosure or on some other device on the network. The indexing portion of the software performs topic-based categorization by means of a higher-order path analysis algorithm, which mimics human intuition by considering both high- and low-order links between data elements. The indexes generated by the software are automatically partitioned by topic. The indexing software operates continuously and transparently, updating the index as data is added to or removed from the storage device.

Abstract: A method and system for constructing a social network using advanced content-based indexing, data and index sharing, and a peer-to-peer network in conjunction with a web site interface. The system uses novel indexing techniques to identify and share common interests among users of the network, and integrated peer-to-peer software facilitates data sharing based on these interests. Users who wish to join the social network download client software from the network's web site. When installed, the software indexes data on the user's local storage devices by considering possible high- and low-order links between data elements, mimicking human intuition. The indexing software generates an index of user data that is partitioned or into separate topic indexes. This index represents a cross-section of the user's interests. The user is then able to join various “friend” groups on the network, and to select by topic which portions of the index are shared with which friend groups.