Abstract: A targeting rule set of association rules may be created. A plurality of association rules may be selected. A plurality of candidate rule sets may be constructed based on the plurality of association rules. The plurality of candidate rule sets may be evaluated to produce metrics. One of the plurality of candidate rule sets may be assigned as the targeting rule set based on the metrics. A custom list of user may be formed using association rules. A user may be determined to belong to a segment by applying one or more rules of a targeting rule set to user attribute data. A custom list of cookies to show advertising may be formed using combinations of association rules.

Abstract: The present invention addresses two ubiquitous and pressing problems of modern data analytics technology. Many modern pattern recognition technologies produce models with excellent predictivity but (a) they are “black boxes”, that is they are opaque to the user; (b) they are too large, and/or expensive to execute in less powerful computing platforms. The invention “opens up” a black box model by converting it to a compact and understandable model that is functionally equivalent. The invention also converts a predictive model into a functionally equivalent model into a form that can be implemented and deployed more easily or efficiently in practice. The benefits include: model understandability and defensibility of modeling. A particularly interesting application is that of understanding the decision making of humans, comparison of the behavior of a human or computerized decision process against another and use to enhance education and guideline compliance/adherence detection and improvement.

Abstract: Described is a system and method for determining a classification of an application that includes initiating a stress test on the application, the stress test including a predetermined number of stress events, wherein the stress events are based on a network impairment. A response by the application to each stress event is identified and the application is classified as a function of the response into one of a first classification and a second classification, the first classification indicative of a normal application and the second classification indicative of an undesired application. If, the application is in the second classification, a network response procedure is executed.

Abstract: Systems and methods described herein utilize supervised machine learning to generate a figure-of-speech prediction model for classify content words in a running text as either being figurative (e.g., as a metaphor, simile, etc.) or non-figurative (i.e., literal). The prediction model may extract and analyze any number of features in making its prediction, including a topic model feature, unigram feature, part-of-speech feature, concreteness feature, concreteness difference feature, literal context feature, non-literal context feature, and off-topic feature, each of which are described in detail herein. Since uses of figure of speech in writings may signal content sophistication, the figure-of-speech prediction model allows scoring engines to further take into consideration a text's use of figure of speech when generating a score.

Abstract: An apparatus and method for mobile prediction is disclosed. In an embodiment, a predictive component creates graphs of user behaviors and receives subgraphs derived from a global graph. By combining a subset of the user behavior graph and the subgraph, a user behavior is predicted. Other embodiments are described and claimed.

Abstract: Disclosed is a method of predicting user's position. This method comprises, creating information on a plurality of location clusters by processing a plurality of position data for a user with a probability based clustering algorithm; receiving a current position data of the user and determining a first location cluster to which the current data is mapped among the plurality of location clusters; and creating second information related to a probability that the user moves from the first location cluster to a second location cluster among the plurality of location clusters. The position data is a data tuple including latitude, longitude, and time. For all the plurality of location clusters, the information includes a determined representative position value of each of the location clusters.

Abstract: This disclosure relates to systems and methods for improved knowledge mining. In one embodiment, a method is disclosed, which comprises filtering aggregated data encoded according to multiple data formats, using a combination of sliding-window and boundary-based filtration techniques. Machine learning and natural language processing are applied to the filtered data to generate a business ontology. Also, using a prediction analysis, one or more recommended classification techniques are automatically identified. The filtered data is clustered into an automatically determined number of categories based on the automatically recommended one or more classification techniques. The one or more classification techniques may utilize iterative feedback between a supervised learning technique and an unsupervised learning technique.

Abstract: Disclosed is a system and method for processing account registration by identifying account candidates attempting to open an account as abusive. That is, the present disclosure discusses identifying, and challenging and marking abusive account registration. The present disclosure takes into account users' behaviors on a network and the impact to the cost and/or revenue of the network. The present disclosure is proactive as it allows for actions to be taken at the earliest possible time in the registration process before an account is created. This prevents abusive activity from taking place within the network and effecting services and privileges available to legitimate users. Additionally, the effects of the disclosed systems and methods minimize the negative impacts of abusive activity on normal user accounts.

Abstract: A computing device determines an SVDD to identify an outlier in a dataset. First and second sets of observation vectors of a predefined sample size are randomly selected from a training dataset. First and second optimal values are computed using the first and second observation vectors to define a first set of support vectors and a second set of support vectors. A third optimal value is computed using the first set of support vectors updated to include the second set of support vectors to define a third set of support vectors. Whether or not a stop condition is satisfied is determined by comparing a computed value to a stop criterion. When the stop condition is not satisfied, the first set of support vectors is defined as the third set of support vectors, and operations are repeated until the stop condition is satisfied. The third set of support vectors is output.

Abstract: A liquid state machine pulse domain neural processor circuit comprising an asynchronous input filter circuit provided for, at any given time, receiving a series of analog input signals and generating in response a set of time-encoded values that depend on the series of analog input signals received at said given time and before said given time; and an asynchronous trainable readout map circuit for transforming at least a portion of said set of time encoded values into output signals.

Abstract: A clothing search system provides a clothing search to users using a component-based image search. Retailer catalogs are analyzed to determine clothing components within clothing images. Features associated with the components are determined. When a user requests a clothing search, the clothing search system selects clothing based on the components and features requested by the user. The user may also provide an image to the clothing search system. The clothing search system determines components and features of the image and identifies clothing with matching components.

Abstract: A method of predicting food items consumed by a user of a food-logging application is disclosed. Loggings of consumptions of food items are received. A predictive model is generated based on the received loggings. The predictive model generates a prediction of one or more additional food items that a target user will consume or is likely to have consumed (e.g., at a particular time). The prediction is generated based on an application of the predictive model to one or more data items (e.g., data items streaming into the system in real time from the target user or other users that are relevant to food consumptions by the target user). The prediction of the consumption of the one or more additional food items by the user may then be communicated for presentation to the target user in a user interface.

Abstract: Technologies are disclosed herein for generating and utilizing machine-learning generated classifiers configured to identify document relationships. Manually-generated data is captured that indicates if documents in a document corpus have a relationship with one another, such as duplicates or variations. A determination may then be made as to whether a classifier is to be generated based on the duplicate decision data. If a classifier is to be generated, machine learning may be performed using training documents from the document corpus and the duplicate decision data to generate a classifier. The machine-learning generated classifier may then be utilized in a production environment to determine whether a new document is a duplicate of documents in the document corpus and/or to identify other relationships between documents in the document corpus, such as documents that are similar or are variations of one another.

Abstract: A computer-implemented method includes receiving, by a computing device, input activations and determining, by a controller of the computing device, whether each of the input activations has either a zero value or a non-zero value. The method further includes storing, in a memory bank of the computing device, at least one of the input activations. Storing the at least one input activation includes generating an index comprising one or more memory address locations that have input activation values that are non-zero values. The method still further includes providing, by the controller and from the memory bank, at least one input activation onto a data bus that is accessible by one or more units of a computational array. The activations are provided, at least in part, from a memory address location associated with the index.

Abstract: The invention relates to forming a prediction using an experience matrix, a matrix based on sparse vectors such as random index vectors. At least a part of a first experience matrix and at least a part of at least a second experience matrix are caused to be combined (1410) to obtain a combined experience matrix. The experience matrices comprise sparse vectors or essentially similar vectors in nature, and said experience matrices comprise information of at least one system, for example contexts of a system. At least a part of at least one sparse vector of the combined experience matrix is accessed to form a prediction output (1420), and a system is controlled (1430) in response to said prediction output.

Abstract: A method and apparatus used for general purpose problem solving using entanglement properties of holography. Intelligent point-based entities having spatial and other electromagnetic properties called DROPLETS [Data-Representative-Object-Particle(s)-Liking-EnTanglement] are generated as delegate objects—avatars—connected to data sources representing situations, event or other problems. A DROPLET's properties are controlled by changes in input data, self-state, feedback, and/or changes of other DROPLETS. Coherent rays are introduced and interact with DROPLETS, generating an INTELLIGENCE WAVEFRONT. Interference patterns are recorded and converted to binary machine codes of a near-infinite set, instructing where to store human/machine-readable content within a plurality of associative memories. Said content includes waveforms, harmonics, codes, data, and other holograms, which are dispersed and stored wholistically throughout using spread spectrum techniques.

Abstract: A computer-implementable method for providing cognitive insights comprising: receiving streams of data from a plurality of data sources; processing streams of data from a plurality of data sources via a plurality of agents, the processing the streams of data from the plurality of data sources via the plurality of agents performing a respective plurality of cognitive operations on the streams of data; and, providing cognitive insights based upon the performing the respective plurality of cognitive operations on the streams of data from the plurality of data sources.

Abstract: A data architecture for use within a cognitive information processing system environment comprising: a plurality of data sources, the plurality of data sources comprising a public data source and a private data source, the public data source comprising publicly available travel information, the private data source comprising privately managed, company specific travel information; and, a cognitive data management module, the cognitive data management module accessing information from the plurality of data sources and providing the information to an inference and learning system.

Abstract: Techniques disclosed herein describe inferring user interests based on metadata of a plurality of multimedia objects captured by a plurality of users. An analysis tool receives, for each of the users, metadata describing each multimedia object in the plurality of objects associated with that user. Each multimedia object includes one or more attributes imputed to that object based on the metadata. The analysis tool identifies one or more concepts from the one or more attributes. Each concept includes at least a first attribute that co-occurs with a second attribute imputed to a first multimedia object. The analysis tool associates a first one of the plurality of users with at least one of the concepts based on the attributes imputed to multimedia objects associated with the first one of the plurality of users.

Abstract: Systems and methods for or training as sequence tagger, such as conditional random field model. More specifically, the systems and methods train a sequence tagger utilizing partially labeled data from crowd-sourced data for a specific application and partially labeled data from search logs. Further, the systems and methods disclosed herein train a sequence tagger utilizing only partially labeled by utilizing a constrained lattice where each input value within the constrained lattice can have multiple candidate tags with confidence scores. Accordingly, the systems and methods provide for a more accurate sequence tagging system, a more reliable sequence tagging system, and a more efficient sequence tagging system in comparison to sequence taggers trained utilizing at least some fully-labeled training data.