Abstract: An adaptive machine learning system for predicting file controls includes a memory, an interface, and a processor. The memory stores a plurality of controls for incoming files and the interface receives a first file and a second file. The first file has a first property and the second file has a second property. The processor determines a type for each of the first property and the second property, wherein the type of each property is related to a first file control. The processor also determines that the first property and the second property each satisfy the first file control. If the value of the first property and the second property are above a first threshold, the processor changes a value of the first control for incoming files.

Abstract: An adaptive machine learning system for predicting file controls includes a memory, an interface, and a processor. The memory stores a plurality of controls for incoming files and the interface receives a first file and a second file. The first file has a first property and the second file has a second property. The processor determines a type for each of the first property and the second property, wherein the type of each property is related to a first file control. The processor also determines that the first property and the second property each satisfy the first file control. If the value of the first property and the second property are above a first threshold, the processor changes a value of the first control for incoming files.

Abstract: Methods, systems, and computer-readable media for detecting errors within a system by using behavior profiles are presented. At a first time, user requests may be received and serviced. The serviced user requests may be logged. Based on the logged user requests, profiles may be determined. At a second time, user requests may be received and serviced. The serviced user requests may be logged. The logged serviced user requests may be compared to the profiles determined at a first time. For example, the determined profiles may include an error rate for serviced user requests. At the second time, an error rate for the logged serviced user requests may be compared to an error rate included the determined profiles. Serviced users requests may be flagged based on the comparison.

Abstract: Methods, systems, and computer-readable media for detecting errors within a system by using behavior profiles are presented. At a first time, user requests may be received and serviced. The serviced user requests may be logged. Based on the logged user requests, profiles may be determined. At a second time, user requests may be received and serviced. The serviced user requests may be logged. The logged serviced user requests may be compared to the profiles determined at a first time. For example, the determined profiles may include an error rate for serviced user requests. At the second time, an error rate for the logged serviced user requests may be compared to an error rate included the determined profiles. Serviced users requests may be flagged based on the comparison.

Abstract: Methods, systems, and computer-readable media for detecting errors within a system by using behavior profiles are presented. At a first time, user requests may be received and serviced. The serviced user requests may be logged. Based on the logged user requests, profiles may be determined. At a second time, user requests may be received and serviced. The serviced user requests may be logged. The logged serviced user requests may be compared to the profiles determined at a first time. For example, the determined profiles may include an error rate for serviced user requests. At the second time, an error rate for the logged serviced user requests may be compared to an error rate included the determined profiles. Serviced users requests may be flagged based on the comparison.

Abstract: Methods, systems, and computer-readable media for detecting errors within a system by using behavior profiles are presented. At a first time, user requests may be received and serviced. The serviced user requests may be logged. Based on the logged user requests, profiles may be determined. At a second time, user requests may be received and serviced. The serviced user requests may be logged. The logged serviced user requests may be compared to the profiles determined at a first time. For example, the determined profiles may include an error rate for serviced user requests. At the second time, an error rate for the logged serviced user requests may be compared to an error rate included the determined profiles. Serviced users requests may be flagged based on the comparison.

Abstract: Methods, systems, and computer-readable media for deploying a software module in a dynamic configuration framework are presented. A system may be running a software service, such as a software service that abstracts or transforms requests such that the requests may be serviced by a web service. The system may receive a request to deploy a new software module. In response to the request, the system may retrieve a binary file from a database. The binary file may comprise, for example, a Java Archive (.jar) file. A real-time class loader may then be accessed, where the real-time class loader may be configured to deploy the retrieved binary file. The software module may then be deployed by the real-time class loader using the retrieved binary file. The deployment may be achieved without interrupting the software service being run on the system.

Abstract: A lightweight cigarette filter including a filter element containing filter material having a total Denier of less than 35,000 or a bulk density of less than 0.12 gcm.sup.-3 in which part of the volume of the filter is occupied by a space filler impervious to smoke flow therethrough and the bulk density of which is less than the bulk density of the equivalent amount of filter material which it replaces.