Abstract: Communications where one or more of the participants are at least partially anonymized. For example, various embodiments allow for the hiding of facial features, the distortion of voice characteristics, or both, while still allowing substantive communication.

Abstract: A data processing system implements detecting that a first user device associated with a first user has added a first sponsored content item to a host electronic document by adding a first reference to the first sponsored content item to the host electronic document. The first sponsored content item is stored separately in a memory of a cloud-based service from the host electronic document. The data processing system further implements determining that the first user has permission to share the first sponsored content item with other users of the cloud-based service, generating sponsor information to associate the sponsored content item with the host electronic document to permit users having access to the host electronic document to access the sponsored content item, and storing the sponsor information in a sponsored access datastore.

Abstract: Examples described herein relate to a security management system to secure a container ecosystem. In some examples, the security management system may protect one or more entities such as container management applications, container images, containers, and/or executable applications within the containers. The security management system may make use of digital cryptography to generate digital signatures corresponding to one or more of these entities and verify them during the execution so that any compromised entities can be blocked from execution and the container ecosystem may be safeguarded from any malicious network attacks.

Abstract: Examples described herein relate to a security management system to secure a container ecosystem. In some examples, the security management system may protect one or more entities such as container management applications, container images, containers, and/or executable applications within the containers. The security management system may make use of digital cryptography to generate digital signatures corresponding to one or more of these entities and verify them during the execution so that any compromised entities can be blocked from execution and the container ecosystem may be safeguarded from any malicious network attacks.

Abstract: Techniques are disclosed relating to application verification. In various embodiments, a computing device includes a secure circuit configured to maintain a plurality of cryptographic keys of the computing device. In such an embodiment, the computing device receives, from an application, a request for an attestation usable to confirm an integrity of the application, instructs the secure circuit to use one of the plurality of cryptographic keys to supply the attestation for the application, and provides the attestation to a remote computing system in communication with the application. In some embodiments, the secure circuit is configured to verify received metadata pertaining to the identity of the application and use the cryptographic key to generate the attestation indicative of the identity of the application.

Abstract: Examples described herein relate to a security management system to secure a container ecosystem. In some examples, the security management system may protect one or more entities such as container management applications, container images, containers, and/or executable applications within the containers. The security management system may make use of digital cryptography to generate digital signatures corresponding to one or more of these entities and verify them during the execution so that any compromised entities can be blocked from execution and the container ecosystem may be safeguarded from any malicious network attacks.

Abstract: A method for tamper protection in cryptographic calculations is provided. A cryptographic calculation includes a plurality of normal rounds and a plurality of redundant rounds. The method includes obtaining a first variable x and a second variable y using a random number generator; dividing the normal rounds into a first normal section and a second normal section, and dividing the redundant rounds into a first redundant section and a second redundant section according to the first variable x and the second variable y; executing the first normal section and the first redundant section in sequence using a clock signal; in response to completion of the first redundant section and a first calculation result of the first normal section and a second calculation result of the first redundant section being the same, executing the second normal section and the second redundant section in sequence to complete the cryptographic calculation.

Abstract: Systems and methods for managing telemetry services in a heterogeneous computing platform are described. In some embodiments, an Information Handling System (IHS) may include a heterogeneous computing platform having a plurality of devices and a memory coupled to the heterogeneous computing platform, where the memory includes a plurality of sets of firmware instructions, where each set of firmware instructions, upon execution by a respective device, enables the respective device to provide a corresponding firmware service, and where at least one of the plurality of devices operates as an orchestrator configured to: receive a policy from an Information Technology Decision Maker (ITDM) or Original Equipment Manufacturer (OEM) and collect telemetry from at least one of the plurality of devices based, at least in part, upon the policy.

Abstract: A computer includes a processor including a plurality of registers, a memory, and a storage medium. A processor of a computer is configured to execute an encryption process of generating encrypted user data including a plurality of encrypted data blocks using the plurality of registers, and add a DIF including CRC to the encrypted data blocks and store the result in a storage medium. The encryption process includes repeatedly executing a first process of reading partial data from a predetermined number of the data blocks and storing the partial data in a first register, a second process of storing encrypted partial data obtained by encrypting the partial data stored in the first register in a second register, and a third process of executing an operation for calculating CRC using the encrypted partial data stored in the second register and storing a result of the operation in a third register.

Abstract: A system for providing forensic tracing of memory device content erasure and tampering is disclosed. The system uses a special command that enables forensic tracing in a secure memory device. Once the forensic tracing is enabled, firmware of the memory device tracks the data stored on the memory device. The command specifies whether the tracking and tracing is for the entire memory device or for a region of the memory device. The firmware confirms that the forensic tracing is enabled, and a target protection region is defined. Once an authenticated command for an operation to access, modify, or erase data of the memory device is received from a host, the system enables the operation to proceed. The system creates a trace of the operation and the metadata of the target region that is modified within a secure memory region of the memory device that is not addressable by the host device.

Abstract: A distributed CAD system includes a server computing device comprising a reference CAD object database and a CAD file database. The reference CAD object database comprises a plurality of reference CAD objects, the CAD file database comprises one or more master CAD files, and each of the one or more master CAD files comprises master object information and master identification information. The distributed CAD system includes a plurality of client computing devices. Each of the plurality of client computing devices comprises a client database, and each client database comprises one or more local CAD files and local object information associated with each of the one or more local CAD files. The server computing device is configured to determine whether an input received by a given client computing device from among the plurality of client computing devices satisfies one of a file creation condition and a file modification condition.

Abstract: A distributed CAD system includes a server computing device comprising a reference CAD object database and a workstation database. The reference CAD object database comprises a plurality of reference CAD objects, and the workstation database includes a plurality of workstation entries. The distributed CAD system includes a plurality of client computing devices. Each client computing device from among the plurality of client computing devices comprises a client database, and the client database comprises one or more local CAD files and one or more local properties associated with each of the one or more local CAD files. Each client computing device is configured to: select a given local CAD file, identify a given workstation entry associated with the given local CAD file, selectively modify a set of one or more master properties defined by the given workstation entry and a set of the location-based information defined by the given workstation entry.

Abstract: A system for verifying insurance coverage is described. A central system receives an address of a property into a central system including information about insurance products for a premises. The central system retrieves an informal two-dimensional rendering of the premises from at least one external data store. The central system analyzes said informal two-dimensional rendering. The central system determines features found on the premises depicted by the two-dimensional rendering, wherein said features comprise walls, doors, windows, and appliances. The central system automatically generates a three-dimensional model of the premises on demand using only the two-dimensional rendering of the premises. The central system verifies insurance coverage for the property using the three-dimensional model.

Abstract: An amusement park attraction design system may include an object token having a first tracker coupled to a first projection surface, a visualization tool having a second tracker, and an image sensor that generates location data based on the first tracker and the second tracker. The amusement park attraction design system may also include a controller communicatively coupled to the image sensor and a projector. The controller may receive the location data from the image sensor to determine the object token and the visualization tool based on the first tracker and the second tracker, respectively. The controller may also receive image content based on an interaction between the object token and the visualization tool and send an indication of the image content to be projected. The projector may receive the indication of the image content and project the image content onto the first projection surface.

Abstract: A computer-implemented method for generating a 3D model representing a factory. The method includes obtaining a point cloud from a scan of the factory and fitting the point cloud with linear CAD extrusions. Such a method is an improved solution for generating a 3D model representing a factory.

Abstract: Systems and methods for modeling capabilities within system models of industrial systems. One system includes an electronic processor configured to receive a request for a system model of an industrial system associated with a set of components. The electronic processor is also configured to access system data associated with the industrial system. The electronic processor is also configured to generate, based on at least the system data, a system model such that the system model includes a set of capabilities associated with the industrial system, wherein each capability included in the set of capabilities is associated with a function of the industrial system. The electronic processor is also configured to enable transmission of the system model for display.

Abstract: The present disclosure relates to a system and method for optimizing building design, construction and operation by tracking post-construction operational data, along with design and construction data. The system addresses the energy performance gap associated with modem construction.

Abstract: Garments in their natural form are represented by meshes, where vertices (entities) are connected (related) to each other through mesh edges. Earlier methods largely ignored this relational nature of garment data while modeling garments and networks. Present disclosure provides a particle-based garment system and method that learn to simulate template garments on the target arbitrary body poses by representing physical state of garment vertices as particles, expressed as nodes in a graph, and dynamics (velocities of garment vertices) is computed through a learned message-passing. The system and method exploit this relational nature of garment data and network implemented to enforce strong relational inductive bias in garment dynamics thereby accurately simulating garments on the target body pose conditioned on body motion and fabric type at any resolution without modification even for loose garments, unlike existing state-of-the-art (SOTA) methods.

Abstract: Provided are a method, a device and electronic equipment for detecting a stability of a generator set in a power system. The method includes following steps: determining a target matrix by using Liapunov's direct method based on a state space matrix corresponding to power management unit (PMU) data; obtaining a state deviation corresponding to the PMU data and an oscillation curve corresponding to the state deviation in a case of determining the power system being in a stable state according to the target matrix; determining extreme point values corresponding to the oscillation curve at each oscillation; and evaluating a working state of the generator set in the power system according to the extreme point values.

Abstract: By a tracking device, a tracking method, or a computer-readable non-transitory storage medium storing a tracking program, a state value of a mobile object is estimated to track the mobile object, the observation value of the mobile object observed at an observation time is acquired, a prediction state value is acquired, a true value of the state value at the observation time is estimated by nonlinear filtering.

Abstract: Apparatus operative via executable software to quantify building metrics using artificial intelligence analysis of a design plan. A design plan is represented using multiple dynamic components. Each dynamic component may include a parameter changeable via the user interactive interface. The dynamic components may be arranged in a user interactive interface to form a first set of boundaries, including a respective length and area, and defining at least a portion of a first unit. AI is used to determine a longest path of egress and a supportable occupancy load that may be used to determine compliance with parameters of a given code. The AI may assess whether a building described in the design plans complies with a relevant code set forth by an authority having jurisdiction. Codes may include, for example, codes enforcing fire safety and the Americans with Disabilities Act.

Abstract: A method for modeling an identification model of a tire capacity, including: obtaining tire test data, wherein the tire test data comprises a tire angular velocity, a wheel effective radius, a tire slip angle, a wheel center velocity, a tire longitudinal force, a tire lateral force and a tire vertical force; obtaining a total slip ratio and a normalized tire force according to the tire test data; obtaining a tire capacity corresponding to the total slip ratio and the normalized tire force according to the tire test data; and performing training using the total slip ratio, the normalized tire force, and the tire capacity through a machine learning algorithm to complete the modeling of the identification model of the tire capacity.

Abstract: A method of generating a circuit design parameter meeting a target specification, the method including generating a first probability distribution of a first specification using a first model provided a first parameter where the first model is configured to infer a correlation between the first parameter and the first probability distribution, generating a second parameter using a second model provided the first probability distribution, and updating the first model based on the second parameter.

Abstract: Machine assisted systems and methods for enhancing the resolution of an IC thermal profile from a system analysis are described. These systems and methods can use a neural network based predictor, that has been trained to determine a temperature rise across an entire IC. The training of the predictor can include generating a representation of two or more templates identifying different portions of an integrated circuit (IC), each template associated with location parameters to position the template in the IC; performing thermal simulations for each respective template of the IC, each thermal simulation determining an output based on a power pattern of tiles of the respective template, the output indicating a change in temperature of a center tile of the respective template relative to a base temperature of the integrated circuit; and training a neural network.

Abstract: In various examples, scenarios may be defined using a declarative description—e.g., defining a behavior of interest—that the present system may convert into a procedural description for generating one or more instances and/or variations of a scenario for testing an autonomous or semi-autonomous machine in a virtual environment. The system may execute observers or evaluators for testing the performance and accuracy of the machine and may compute coverage of various elements based on the generated virtual scenarios, and may feed the results back to the system to generate additional instances and/or variations where the coverage or accuracy is below a desired level. As a result, the system may include an end-to-end framework for generating scenarios in virtual environments, testing and validating the scenarios themselves, and/or testing and validating the underlying autonomous or semi-autonomous systems of the machine—all based on a declarative description.

Abstract: Technologies for producing sampling and analysis plans for evaluating airborne chemical hazards are disclosed. These technologies obtain scenario data indicative of emission of an airborne chemical at a site with a human population; simulate, as a function of the scenario data, emission of the chemical at the site and transport and dispersion of a plume of the chemical across the site in space and time; assess an extent to which concentrations of the chemical in the airborne plume encounter areas containing people, as a function of time and projected human activity and movement across the site; determine, based on the simulation and assessment, recommended sampler device locations to obtain a target quality and target quantity of samples of the chemical for use in a sampling and analysis plan to determine an effect of the chemical on the human population at the site; and present the recommended sampler device locations.

Abstract: A simulation method predicts a behavior of a curable composition in a process of bringing droplets of the curable composition arranged on first and second members into contact with each other, and forming a film of the curable composition on the first member. The method includes inputting a physical property value of a gas between the first and second members, inputting a movement profile of the second member with respect to the first member when bringing the droplets of the curable composition on the first and second members into contact with each other, obtaining a pressure of the gas between the first and second members based on the physical property value and the input movement profile, and predicting, based on the pressure, an amount of a residual gas confined among the droplets by the contact between the droplets and the second member.

Abstract: The present disclosure describes systems and methods for adjusting a logic network. The method includes adding, to the logic network, a first redundant node and determining a first adjustment to a first node of the logic network within a transitive fanin of the first redundant node. The method also includes making the first adjustment to the first node based on determining that a first gain based on the first adjustment satisfies a threshold.

Abstract: Provided is a method for generating an initial guess for simulating a semiconductor device. The method includes; (a) setting 1D direction for a 3D semiconductor device to be simulated and 2D cross sections perpendicular to the 1D direction; (b) setting a charge intensity-gate voltage model for the 2D cross sections; (c) numerically solving an electron or hole continuity equation along the 1D direction to detect integrated electron charge density and quasi-Fermi potential for the 2D cross sections; and (d) calculating a physical quantity for each position of the 2D cross sections by applying the integrated electron charge density and the quasi-Fermi potential for the 2D cross sections to the charge density-gate voltage model, and providing the calculated physical quantity as the initial guess for the semiconductor device simulation. Therefore, the initial solution according to the present invention can accelerate the speed of the semiconductor device simulation.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed for methods and apparatus to generate I/O timing constraint predictions for system-on-chips using machine learning. An apparatus is disclosed herein comprising at least one memory, machine readable instructions, and processor circuitry to at least one of instantiate or execute the machine readable instructions to extract delay features from a first system-on-chip design and a second system-on-chip design, apply a machine learning model to the extracted delay features to determine an estimated delay for the first system-on-chip design, compare the estimated delay for the first system-on-chip design to the second system-on-chip design to determine an error of the first system-on-chip design, and update the second system-on-chip design based on the error determined for the first system-on-chip design.

Abstract: A “weak” undriven state is defined as a signal state, distinguished from conventional unknown and high impedance states, and methods of representing this “weak” undriven state in circuit modelling and power aware digital/mixed-signal simulations for comprehensive and complete RTL-level design verification. The conventional unknown state refers to a circuit element that is powered but has an unknown value, a circuit element that is not powered, or a circuit element having an undriven, floating signal. The unknown state is modified, and the “weak” undriven state refers to a circuit element that is not powered and has an unknown value. The “weak” undriven state can have an electrically high impedance to known supply or ground when no other circuit element is active. The “weak” undriven state distinction is particularly useful to model and verify circuit designs known to be resilient to “weak” undriven states, using event driven logic circuit simulators.

Abstract: Various techniques are disclosed for automatically generating sub-cells for a non-final layout of an analog integrated circuit. Device specifications and partition information for the analog integrated circuit is received. Based on the device specifications and the partition information, first cut locations for a first set of cuts to be made along a first direction of a non-final layout of the analog integrated circuit and second cut locations for a second set of cuts to be made along a second direction in the non-final layout are determined. The first set of cuts are made in the non-final layout at the cut locations to produce a temporary layout. The second set of cuts are made in the temporary layout at the cut locations to produce a plurality of sub-cells.

Abstract: An information processing apparatus has a first drawing function and a second drawing function different from the first drawing function, draws a screen using the first drawing function and the second drawing function, acquires a setting of a character decoration used to draw a predetermined font by the second drawing function, and uses the acquired setting of the character decoration when drawing a character of the predetermined font using the first drawing function.

Abstract: The freshness of one or more terms in a documentation, indicative of a currency of the one or more terms is computed. Each term includes one or more constituent words, and the terms are visually marked as current or out-of-date based on the computed freshness. Upon marking a term as out-of-date, a latest term for the out-of-date term is retrieved or a most possible latest term for the out-of-date term is predicted.

Abstract: Aspects of the present disclosure relate to creating a long note from one or more short notes made in a source document. A long note is a separate document created by the system to capture each of the one or more short notes and related source content from the source document and compiles them in one place for independent review and modification. The long note captures the connection between the short note and the source document with source content such as the related text, images, or other source content within the source document to which the short note is directed. The long note may be created independently of a user request, such that the entry of a short note into the source document initiates the creation of the long note.

Abstract: Implementations described herein relate to causing emoji(s) that are associated with a given emotion class expressed by a spoken utterance to be visually rendered for presentation to a user at a display of a client device of the user. Processor(s) of the client device may receive audio data that captures the spoken utterance, process the audio data to generate textual data that is predicted to correspond to the spoken utterance, and cause a transcription of the textual data to be visually rendered for presentation to the user via the display. Further, the processor(s) may determine, based on processing the textual data, whether the spoken utterance expresses a given emotion class. In response to determining that the spoken utterance expresses the given emotion class, the processor(s) may cause emoji(s) that are stored in association with the given emotion class to be visually rendered for presentation to the user via the display.

Abstract: An electronic device with a display: concurrently displays a first electronic document and a second electronic document, the first electronic document including a displayed content object; detects a first user input that starts on the content object; in response to a determination that the first user input corresponds to a first command, creates and displays a copy of the content object in the second electronic document; and, in response to a determination that the first user input corresponds to a second command, distinct from the first command: selects the content object; displays an animation of the content object separating from the first electronic document; and initiates display of a plurality of icons proximate to the separated content object, the plurality of icons including at least one icon to initiate display of a user interface for sending the content object to one or more other users.

Abstract: Implementations are disclosed for automatically generating computer code that implements a machine learning-based processing pipeline based on multiple different modalities of input. In various implementations, one or more annotations created on a demonstration digital image to annotate one or more visual features depicted in the demonstration digital image may be processed to generate annotation embedding(s). Natural language input describing one or more operations to be performed based on the one or more annotations also may be processed to generate one or more logic embeddings. The annotation embedding(s) and the logic embedding(s) may be processed using a language model to generate, and store in non-transitory computer-readable memory, target computer code. The target computer code may implement a machine learning-based processing pipeline that performs the one or more operations based on the one or more annotations.

Abstract: A system and method for completing a character of a text of a digital document on a computing device, the computing device comprising a processor, a memory, and at least one non-transitory computer readable medium for recognizing input under control of the processor, the at least one non-transitory computer readable medium is configured to cause display (S900) of at least one typeset character of the text on a display interface of the computing device; detect a handwritten input stroke (S902) performed on the digital document in the vicinity of a typeset character; identify an first typeset character (S904) if the typeset character belongs to a list of base characters according to a language model; retrieve a predefined character version (S906) of the first typeset character from the memory; generate a hybrid character (S908) by replacing the initial typeset character by the predefined character; generate a list of character candidates (S910) with associated probabilities of recognition of the hybrid characte

Abstract: The disclosed technology includes multi-way triggering of automatic replication elements selectively within and across documents, cellular atomization of spreadsheet cells and charts while retaining their desired formula, function and content properties, combining the selective multi-way replication with the cellular atomization, and employing a library capability to easily reuse automatically coordinating elements and atomized spreadsheet or tabular cells and charts.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing a machine learning task on a network input to generate a network output. In one aspect, one of the systems includes a neural network configured to perform the machine learning task, the neural network comprising an N-grammer layer and an output neural network, the N-grammer layer configured to: at each of one or more heads: receive a sequence of input embeddings; generate a discrete latent representation of the sequence of input embeddings by using a learned product quantization codebook; generate a plurality of n-gram indices from the discrete latent representation; and generate a latent n-gram representation of the sequence of input embeddings; and generate a sequence of output embeddings, and the output neural network configured to: receive the sequence of output embeddings; and process the sequence of output embeddings to generate the network output.

Abstract: Persistent storage contains an original document and a plurality of summaries of the original document produced by summarization models. One or more processors: provide, to an entity extractor, the original document; receive, from the entity extractor, a list of entities within the original document; provide, to a query generator, the original document and the list of entities; receive, from the query generator, a set of queries answerable by the original document; provide, to a query answerer, the set of queries, the original document, and the plurality of summaries; receive, from the query answerer and for the set of queries, a set of document answers corresponding to the original document and sets of summary answers corresponding to the plurality of summaries; provide, to an answer matcher, the set of document answers and the sets of summary answers; and receive, from the answer matcher, scores for the plurality of summaries.

Abstract: Described herein are systems and methods for performing literature retrieval and related machine learning model training methods. An example computer-implemented method of training a machine learning model configured for literature retrieval I includes receiving a plurality of full-text articles; extracting, from the plurality of full-text articles, a plurality of positive sentence-citation pairs, each positive sentence-citation pair comprising a respective citing sentence and at least one cited article that is associated with the respective citing sentence; creating a labeled dataset comprising the plurality of positive sentence-citation pairs; and training a machine learning model using the labeled dataset.

Abstract: One example method includes receiving a rule-set, including a combination of rules, that was determined to occur in a set of ground truth documents, applying the rule-set to a new document that was not included in the set of ground truth documents, determining whether or not a rule in the rule-set succeeded or failed when applied to a word in the new document, and when the rule is determined to have failed, identifying the failed rule, identifying a confidence level in the determination that the rule failed, and when the confidence level is below a threshold confidence level, identifying the word, to which the failed rule was applied, as a candidate for verification by a human.

Abstract: A storage device stores failures and lessons. A processor accepts input of text serving as a question sentence in a multiple-choice quiz, computes an evaluation value for a consistency between the text and each of lessons stored in the storage device, and determines one lesson to serve as a correct option in the multiple-choice quiz and at least one lesson to serve as an incorrect option in the multiple-choice quiz, based on the evaluation value for each of the plurality of lessons computed in the computing operation.

Abstract: The present application discloses a method of training a sentiment preference recognition model for comment information, a recognition method and device thereof, which belong to the field of natural language processing technology. The method of training a sentiment preference recognition model for comment information includes training and optimizing, based on each piece of comment information, and corresponding original text template and mirrored text template thereof, a sentiment preference recognition model in the original channel and the mirror channel, so that the sentiment preference recognition model is used to output a sentiment preference result for the two comparison objects in the comment information in terms of the attribute.

Abstract: A method and apparatus for generating text. The method comprises: receiving original text and an element tag (201); generating an encoded text feature vector, and an encoded sentence feature vector of each sentence in the original text; executing decoding steps comprising: determining a word attention weight of each word in the original text at the current moment on the basis of a hidden state vector of the decoder at the current moment and the encoded text feature vector (2031); determining a sentence attention weight of each sentence in the original text; determining a normalized word attention weight of each word in the original text at the current moment; and estimating a target word at the current moment; and generating target text on the basis of the target word output by the decoder at each moment (204).

Abstract: Methods, apparatuses, and systems are presented for generating natural language models using a novel system architecture for feature extraction. A method for extracting features for natural language processing comprises: accessing one or more tokens generated from a document to be processed; receiving one or more feature types defined by user; receiving selection of one or more feature types from a plurality of system-defined and user-defined feature types, wherein each feature type comprises one or more rules for generating features; receiving one or more parameters for the selected feature types, wherein the one or more rules for generating features are defined at least in part by the parameters; generating features associated with the document to be processed based on the selected feature types and the received parameters; and outputting the generated features in a format common among all feature types.

Abstract: A text chain generation method includes selecting a to-be-matched phrase chain from a phrase chain set to match the initial phrase chain and determining the largest common subsequence between the to-be-matched phrase chain and the initial phrase chain; updating the initial phrase chain by adding a word from the to-be-matched phrase chain and other than the largest common subsequence into the initial phrase chain; using the updated initial phrase chain as a new initial phrase chain and repeating the previous steps until traversing all phrase chains in the phrase chain set to obtain an updated phrase chain; and connecting a left node located in each branch of the updated phrase chain and not connected to any node to a preset common start node and connecting a right node located in each branch of the updated phrase chain and not connected to any node to a preset common end node.

Abstract: Disclosed are systems and methods for receiving a plurality of comments at a particular phase of a transaction with a member of a networked system, classifying one or more of the plurality of comments into one of a set of predetermined sentiment classifications, applying a trained machine learning system to select a category from a set of predefined categories for each of the one or more comments, applying a natural language processing module to generate a sub-category for each of the one or more comments, associating the generated sub-categories with their respective categories for the one or more comments, and generating a display of the determined categories for the particular transaction with the generated sub-categories, each generated sub-category being graphically connected to their respective categories.