摘要:
A search system generates an ordered list of destination targets including identified individuals and data objects, based on applying a user context to an information context and to a connection context. The user context identifies a search context (i.e., the information the user is seeking) relative to an identity of the user, and the information context identifies destinations accessible by the user, including data objects and communications-based actions (e.g., voice over IP telephony calls, instant messaging session transcripts, etc.). The connection context identifies relationships between chosen destinations throughout the system, and the strength of those relationships determined from artifacts of the chosen destinations: the connection context is updated based on artifact detection. The portion of the connection context that is relevant to the user context is used to order the portion of the information context that is relevant to the user context, resulting in the ordered list of destination targets.
摘要:
Techniques for presenting a key to a human across a human-machine interface include receiving key data that indicates a key to be input to a target process. The key is reversibly converted to a first human-language statement. The human language statement is designed so that multiple variations in the human-language statement due to common human errors can be correctly reversed to regenerate the key data. The common human-language statement is presented to a human user for use in the target process. After receiving input data that indicates a second human language statement based on human memory of the first human language statement, the second human language statement is corrected for common human language errors. The corrected second human language statement is then converted to second key data and sent to the target process. Thus multiple human inputs are accepted and correctly converted to the key for the target process.
摘要:
In one embodiment, a method includes obtaining a first document that includes at least a first section, and displaying the first document on a display screen. The method also includes determining when the first section has been consumed. Such a determination includes a determination of whether the first section has been displayed on the display screen. If the first section is determined to have been displayed on the display screen, the method also includes providing an indication that the first section has been consumed.
摘要:
A solution is provided for controlling universal serial bus (USB) messages between a plurality of host computers and a USB device. First and second USB servers may communicate with first and second host computers, respectively. A USB client may then communicate with the USB device. A first control path between the USB client and the first USB server and a second control path between the USB client and the second USB server may be established. Then a first data transfer path may be established between the USB client and the first USB server, the first data transfer path enabling the sending of data between the USB client and the first USB server over a data network. Then a busy message may be sent from the USB client to the second USB server over the second control path when the first data transfer path is established with the first USB server.
摘要:
Techniques for presenting identification of a remote network entity to a human user include determining one or more vectors of presentation elements that are readily distinguishable and memorable to a user. Message data is received from a remote process on the network and includes a network identity that uniquely identifies an associated remote entity. The network identity is input to a secure function to produce a secure number. A secure function produces dissimilar output values for similar input values. A particular presentation element is selected from a vector based only on the secure number and presented to the user. Even when the network identity is similar to that of a trusted entity, the particular presentation element is expected to be dissimilar to one associated with the trusted entity.
摘要:
Techniques for associating a physical location with a network interface on an intermediate network node include receiving a data packet from a device connected to a network. Based on data in the data packet, a particular network interface is determined on a first hop intermediate network node. An identity for an entity that originated the data packet is determined based on data in the data packet. A possible physical location for the entity is determined based on the identity. The identity is associated with the possible physical location in an independent data source that is accessible over the network and different from the data packet. The particular network interface is associated with a highest confidence location for the device based on the possible physical location of the entity and any former locations associated with the particular network interface.
摘要:
Coding quantized transform coefficients as occur in image compression includes, for a subset of coefficient amplitude values coding by a first method corresponding to each amplitude of the subset, events of a run of zero or more zero-valued coefficients ending in the single non-zero amplitude, coding by a second method events of a run of zero or more zero-valued coefficients ending in any single non-zero amplitude not in the subset, and coding by a third method the amplitude values of the concatenation of the non-zero amplitude not in the subset. The coding methods produce variable length codewords. The method is suitable for series that have scattered non-zero-valued coefficients.
摘要:
A method of reducing the channel selection transition delay from a first media data channel to a second media data channel includes the operations of accumulating an independent media data frame and any related dependent media data frames based on a first decoder decoding at least a first portion of a second channel to form a second channel current media state, receiving a channel change request from a second decoder to change decoding to the second media channel, identifying an insertion position for a current independent frame relative to a current second channel frame set, and generating the current independent frame based on the second channel current media state corresponding to the insertion position. The dependent media data frames are used to modify the second channel current media state based on the independent media data frame.
摘要:
A portable device includes a display that displays information from an application. A touchpad is oriented perpendicularly to the display. The touchpad detects one or more contact inputs. The one or more contact inputs associate with one or more virtual keys that the application interprets to display the information on the display. An inertial device detects one or more motion inputs. The one or more motion inputs associate with one or more virtual keys that the application interprets to display the information on the display.
摘要:
A method and a carrier medium instructing a processor to execute a method. The method includes accepting a compressed video stream compressed by a compression method that uses macroblocks of possibly different macroblock types for each frame predictive encoding frames. The method further includes identifying a sequence of frames that are substantially visually indistinguishable and that include a predictively encoded frame and the at least one reference frame of the predictively encoded frame. The method further includes creating an ordered set of macroblocks, with the macroblock types of different macroblocks in the ordered set used to encode digital watermark information. The method further includes replacing the predictively encoded frame of the identified sequence with a newly-encoded predictively encoded replacement frame that contains the created ordered set of macroblocks, including the encoded watermarking information.