Abstract: Methods for automatically evaluating ASR outputs and providing annotations, including corrections, on the transcriptions—in order to improve recognition—may be based on an analysis of sessions of user voice queries, utilizing time-ordered ASR transcriptions of user voice queries (i.e., user utterances). This utterance-based approach may involve extracting both session-level and query-level characteristics from a voice query sessions and identifying patterns of query reformulation in order to detect erroneous transcriptions and automatically determine an appropriate correction. Alternative, or in addition, ASR outputs may be evaluated based on user behavior. The outcomes may be classified as positive or negative. An ASR transcription may be labeled using the description of the outcome. The labeled transcription may be used as training data to train a model to output improved transcriptions of voice queries.

Abstract: Voice inputs may be analyzed based on syntactic properties and respective dependency tree structure (e.g., parse tree) may be generated. Dependency tree structures may be grouped/clustered based on an associated root word/phrase. A visual interface may display the dependency tree structure in a manner that maps each voice input to a proper response (e.g., an action, an operation, a command, etc.).

Abstract: Methods and systems for using contextual data (e.g., closed captioning data/information, linear metadata, etc. . . . ) to recommend content are described. A historical record of programs (e.g., content segments) accessed by a user on different channels and at different times may be used to generate content access information (e.g., content access/history information associated with a user, viewer information, etc. . . . ). The content access information may comprise topics determined from contextual data derived from each of the programs. A correlation between the content access information and each of the topics may be used to recommend content relevant to the user's viewing history.

Abstract: A method, system and computer program product for detecting outliers in a set of data points. In one embodiment, the method comprises partitioning the set of data points into a plurality of bins with each of the data points assigned to a respective one of the bins. A plurality of local lists are formed in parallel identifying points in the bins as outliers, and the local lists are merged into a global list to identify one or more of the points as outliers of the data set. Embodiments of the invention provide an outlier detection system that can parallelize in two levels. The dataset is split into partitions, called bins, and outliers are found in each bin in parallel. The execution of a single bin is also parallelized. Embodiments of the invention can scale to very large datasets by these two modes of parallelism.