Streaming machine learning unidirectional models is facilitated by the use of embedding vectors. Processing blocks in the models apply embedding vectors as input. The embedding vectors utilize context of future data (e.g., data that is temporally offset into the future within a data stream) to improve the accuracy of the outputs generated by the processing blocks. The embedding vectors cause a temporal shift between the outputs of the processing blocks and the inputs to which the outputs correspond. This temporal shift enables the processing blocks to apply the embedding vector inputs from processing blocks that are associated with future data.

A compression scheme can be selected for an input data stream based on characteristics of the input data stream. For example, when the input data stream is searched for pattern matches, input stream characteristics used to select a compression scheme can include one or more of: type and size of an input stream, a length of a pattern, a distance from a start of where the pattern is to be inserted to the beginning of where the pattern occurred previously, a gap between two pattern matches (including different or same patterns), standard deviation of a length of a pattern, standard deviation of a distance from a start of where the pattern is to be inserted to the beginning of where the pattern occurred previously, or standard deviation of a gap between two pattern matches. Criteria can be established whereby one or more characteristics are used to select a particular encoding scheme.

A compression scheme can be selected for an input data stream based on characteristics of the input data stream. For example, when the input data stream is searched for pattern matches, input stream characteristics used to select a compression scheme can include one or more of: type and size of an input stream, a length of a pattern, a distance from a start of where the pattern is to be inserted to the beginning of where the pattern occurred previously, a gap between two pattern matches (including different or same patterns), standard deviation of a length of a pattern, standard deviation of a distance from a start of where the pattern is to be inserted to the beginning of where the pattern occurred previously, or standard deviation of a gap between two pattern matches. Criteria can be established whereby one or more characteristics are used to select a particular encoding scheme.

Various aspects of the disclosure are directed to content rendition generation. Sets of disparately-formatted media content are reformatted into corresponding renditions of media content having a common format. The common format includes device-indeterminate ID linking data that links respective portions (e.g., assets or a structural component including the assets) of each rendition with the common format to corresponding portions of the disparately-formatted media content. For each rendition, reformatted assets are generated in which each reformatted is specific to one of a plurality of disparate types of devices, based upon characteristics of the disparate device types. Access to the portions of disparately-formatted media content and/or the assets within the portions of disparately-formatted media content is tracked, based on the linking data.

Various aspects of the disclosure are directed to content rendition generation. Sets of disparately-formatted media content are reformatted into corresponding renditions of media content having a common format. The common format includes device-indeterminate ID linking data that links respective portions (e.g., assets or a structural component including the assets) of each rendition with the common format to corresponding portions of the disparately-formatted media content. For each rendition, reformatted assets are generated in which each reformatted is specific to one of a plurality of disparate types of devices, based upon characteristics of the disparate device types. Access to the portions of disparately-formatted media content and/or the assets within the portions of disparately-formatted media content is tracked, based on the linking data.

In an approach for dynamically generating correctly formatted Extensible Markup Language (XML) based off of a subset of data from a database, a processor receives an input pointing to a subset of data. A processor collects the subset of data from existing sources. A processor modifies the subset of data using wildcarding to generate a modified subset of data. A processor generates a set of XML files from the modified subset of data. A processor creates an immutable record of the subset of data, the modified subset of data, and the set of XML files. A processor outputs the set of XML files.

A system for editing an arrangement comprising information regarding the relationship between two or more source data files comprises a processing unit, a first library comprising one or more source data files, and a user interface. The user interface comprises a display and a user input. The processing unit provides a graphical user interface on the display. The processing unit is configured such that a user can modify an initial arrangement of source data files via the graphical user interface to produce a modified arrangement of source data files.

A method, system and computer device for generating portal pages, including: registering all portal component systems to a portal system portlet library, enabling the portal system to obtain portal portlets from the portlet library, creating a portal page through the portal system, configuring the required portlet system for a new portal page from the portal portlet library, and save name information, page ID identity information and configured portal portlet system information of the portal page into a database; at the same time, the portal system will transfer the newly created portal page information to an IFrame HTML tag contained in the portal system's main page. The IFrame will present the newly created page and load the configured portlets, realizes the partial refresh of a single page of the portal system SPA, save server resources and broadband resources, improve page response speed, and obtain a better user experience.

A method, system and computer device for generating portal pages, including: registering all portal component systems to a portal system portlet library, enabling the portal system to obtain portal portlets from the portlet library, creating a portal page through the portal system, configuring the required portlet system for a new portal page from the portal portlet library, and save name information, page ID identity information and configured portal portlet system information of the portal page into a database; at the same time, the portal system will transfer the newly created portal page information to an IFrame HTML tag contained in the portal system's main page. The IFrame will present the newly created page and load the configured portlets, realizes the partial refresh of a single page of the portal system SPA, save server resources and broadband resources, improve page response speed, and obtain a better user experience.

A method may include obtaining a first byte stream from first document code and a second byte stream from second document code. The first document code has a document type and the second document code has the document type. The method may further include identifying, in the first byte stream, nonvisual noise corresponding to a custom byte code defined in a custom character encoding set. The nonvisual noise is invisible when rendering the first document code. The method may further include replacing, in the first byte stream, the custom byte code with at least one standard byte code defined in a standard character encoding set to obtain modified document code. The second document code uses the standard character encoding set. The method may further include comparing the modified document code with the second document code by comparing the first byte stream with the second byte stream.