Abstract: An optical disc is authenticated by measuring physical attributes of the disc. A challenge is presented to the drive comprising the disc. The challenge includes locations on the disc to be used for authentication. The locations are determined each time the disc is to be authenticated. No restriction is placed on the locations on the medium, and no restriction is placed on the number of locations. Locations on the disc are accessed and an answer to the challenge is calculated in accordance with a physical attribute pertaining to the locations. The answer can include an angle between the locations, the physical separation between the locations, an amount of time elapsed between detection of the locations, an amount of time taken to read data between written between the locations, or a number of rotations occurring between detection of the locations. The answer is analyzed to determine the validity of the disc.

Abstract: A hidden dynamics value in speech is represented by a higher order, discretized dynamic model, which predicts the discretized dynamic variable that changes over time. Parameters are trained for the model. A decoder algorithm is developed for estimating the underlying phonological speech units in sequence that correspond to the observed speech signal using the higher order, discretized dynamic model.

Abstract: Systems and methods for block importance analysis to enhance browsing of web page search results are described. In one aspect, a server analyzes content of a document as a function of multiple block importance criteria. The server assigns a respective block importance level of multiple importance levels to respective block(s) of the analyzed content. The server generates one or more customized documents from block(s) of the content as a function of respective assigned block importance level(s) of the block(s). Each of the one or more customized documents is generated in a particular format of multiple formats to enhance user interaction with the document on a small form factor computing device.

Abstract: Methods for recognizing strings, labeling the strings with a semantic category and providing e-commerce actions based on the category is disclosed. The semantic category may include a type label and other metadata. Recognizer plug-ins perform the recognition of particular strings in an electronic document. The recognizer plug-ins may be packaged with an application program module or they may be written by third parties to recognize particular strings that are of interest. Action plug-ins provide possible actions to be presented to the user based upon the type label associated with the string. Tradenames, trademarks, formal names or types of consumer products may be labeled and actions to buy the products may be presented. The metadata may be used to implement coupon and affiliate programs to reward frequent shoppers or frequent recommenders. Numerous other e-commerce opportunities are presented via the semantic category and the metadata.

Abstract: An interactive entertainment system has a program provider which distributes video content programs to multiple subscribers over a distribution network. Each subscriber has a user interface unit which receives the digital video program and converts it for display on a television, monitor, or other display unit. The user interface unit has a processor and memory. An electronic programming guide (EPG) resides in the memory and is executable on the processor to organize programming information that is descriptive of the programs supplied over the interactive entertainment system. The EPG supports a user interface (UI) which visually correlates programs titles to scheduled viewing times. A hyperlink browser also resides in memory and is executable on the processor. One or more hyperlinks, which reference target resources containing interactive content related to the video programs, are integrated as part of the EPG UI.

Abstract: A speech recognition method provides improved modeling in recognition accuracy using hidden Markov models. During training, the method creates a senone tree for each state of each phoneme encountered in a data set of training words. All output distributions received for a selected state of a selected phoneme in the set of training words are clustered together in a root node of a senone tree. Each node of the tree beginning with the root node is divided into two nodes by asking linguistic questions regarding the phonemes immediately to the left and right of a central phoneme of a triphone. At a predetermined point, the tree creation stops, resulting in leaves representing clustered output distributions known as senones. The senone trees allow all possible triphones to be mapped into a sequence of senones simply by traversing the senone trees associated with the central phoneme of the triphone.