Abstract: Techniques for text evaluation are provided. A curated dataset comprising a plurality of textual documents is received. A tree of classifiers is trained, based on the curated dataset, to identify use cases. A feature graph model is generated, based on the curated dataset, to determine textual similarity. A new document is received, and a plurality of requirements is extracted from the new document. For each requirement, one or more vector scores are generated by evaluating the requirement using the tree of classifiers, one or more feature scores are generated by evaluating the requirement using the feature graph model, and one or more use cases are identified for the new textual document based on the one or more vector scores and the one or more feature scores. An implementation is generated for the new document based on the one or more use cases.

Abstract: A system and method for managing the provisioning of electricity to an electric vehicle includes identifying an electric vehicle when the electric vehicle connects to a charging point; determining whether the electric vehicle is associated with a smart contract; validating the smart contract associated with the electric vehicle; creating a charging transaction for charging the electric vehicle; and applying the smart contract associated with the electric vehicle against the charging transaction. The electric vehicle electricity contract management system and method may use a consensus system such as a distributed ledger system or blockchain.

Abstract: A computer system message generated by an application programming interface (API) or addressed to the API can be received. A first data sensitivity score for at least a first of a plurality of data elements in a payload of the computer system message and at least second data sensitivity score for at least a second of the plurality of data elements in the payload of the computer system message can be determined. Based on the first data sensitivity score and at least the second data sensitivity score, a differential security can be applied to the computer system message. Applying the differential security can include masking the first of the plurality of data elements and not masking the second of the plurality of data elements. The computer system message can be electronically communicated the to a destination to which the message is addressed.

Abstract: A power-distribution routing system of a power-transmission company receives a request to route electrical power to users through a power-grid infrastructure during a specified future period of time. The system retrieves time-stamped blockchain data that identifies past fluctuations in energy demand, service agreements between energy companies, and energy-production and demand-fulfilment histories of energy-generating sources like power plants. The system also retrieves extrinsic contextual and socioeconomic data from online sources and various business applications. An artificially intelligent cognitive framework uses a sliding-frame mechanism to infer patterns in the rate of change of user demand during past time periods similar to the period specified by the request. The system ranks each source by its demonstrated ability to satisfy the patterns of demand in consideration of the contextual data.

Abstract: Techniques for text evaluation are provided. A curated dataset comprising a plurality of textual documents is received. A tree of classifiers is trained, based on the curated dataset, to identify use cases. A feature graph model is generated, based on the curated dataset, to determine textual similarity. A new document is received, and a plurality of requirements is extracted from the new document. For each requirement, one or more vector scores are generated by evaluating the requirement using the tree of classifiers, one or more feature scores are generated by evaluating the requirement using the feature graph model, and one or more use cases are identified for the new textual document based on the one or more vector scores and the one or more feature scores. An implementation is generated for the new document based on the one or more use cases.

Abstract: A power-distribution routing system of a power-transmission company receives a request to route electrical power to users through a power-grid infrastructure during a specified future period of time. The system retrieves time-stamped blockchain data that identifies past fluctuations in energy demand, service agreements between energy companies, and energy-production and demand-fulfilment histories of energy-generating sources like power plants. The system also retrieves extrinsic contextual and socioeconomic data from online sources and various business applications. An artificially intelligent cognitive framework uses a sliding-frame mechanism to infer patterns in the rate of change of user demand during past time periods similar to the period specified by the request. The system ranks each source by its demonstrated ability to satisfy the patterns of demand in consideration of the contextual data.

Abstract: A computer-implemented method includes: receiving, by a computing device, information regarding an application programming interface (API); classifying, by the computing device, the API using natural language classification on to a baseline taxonomy; extracting, by the computing device, information regarding features of the API based on the classifying to the baseline taxonomy; performing, by the computing device, an assessment on the API relative to other APIs within a same class as the API based on the extracting the information regarding the API features and the classifying the API; and updating, by the computing device, the API based on the performing the assessment.

Abstract: A method, system and computer program product for determining API pricing. Consumption parameters are identified using a supervised learning model. The API consumption parameters refer to any parameters that can be used to describe an API (functionality or otherwise) and can be used to compare other comparable APIs in similar domains provided by other providers. Furthermore, reference pricing is determined using machine learning using the identified API consumption parameters. Additionally, the API price is determined dynamically using the identified API consumption parameters and the determined reference pricing. An API pricing score is then derived for the API price using the supervised learning model. The API price is selected as the suggested price for the API in response to the API pricing score exceeding a threshold value. In this manner, an API price is established that reflects the true value of the API assessed by the API consumer.

Abstract: A method (and structure and computer product) of recognizing and tracking an object includes receiving image data from at least two camera sensors and analyzing the image data to determine attributes of an object detected as present in the image data. A first decision tree (DT) is developed to classify the object, based on the attributes determined from the image data. A second DT is developed to refine the first DT, wherein the second DT is used to refine the first DT by determining which attributes are most significant in the classification of the object.

Abstract: A method, system and computer program product are disclosed for creating a dynamic heterogeneous IoT network from existing IoT network devices to identify and implement a dynamic response to an event. In an embodiment, the method comprises identifying an occurrence of a specified event; and creating a heterogeneous IoT network to respond to the event, including selecting a master IoT device and identifying one or more participatory IoT devices for the created network from multiple IoT devices from one or more existing IoT networks. The master IoT device is in bidirectional communications with the participatory IoT devices. The master device identifies a response to the specified event, translates the response into commands, and sends the commands to the participatory IoT devices to implement the response. The master IoT device may pass control to another master device in another heterogeneous IoT network if the response is not accomplished.

Abstract: A method, system and computer program product for determining API pricing. Consumption parameters are identified using a supervised learning model. The API consumption parameters refer to any parameters that can be used to describe an API (functionality or otherwise) and can be used to compare other comparable APIs in similar domains provided by other providers. Furthermore, reference pricing is determined using machine learning using the identified API consumption parameters. Additionally, the API price is determined dynamically using the identified API consumption parameters and the determined reference pricing. An API pricing score is then derived for the API price using the supervised learning model. The API price is selected as the suggested price for the API in response to the API pricing score exceeding a threshold value. In this manner, an API price is established that reflects the true value of the API assessed by the API consumer.

Abstract: A computer system message generated by an application programming interface (API) or addressed to the API can be received. A first data sensitivity score for at least a first of a plurality of data elements in a payload of the computer system message and at least second data sensitivity score for at least a second of the plurality of data elements in the payload of the computer system message can be determined. Based on the first data sensitivity score and at least the second data sensitivity score, a differential security can be applied to the computer system message. Applying the differential security can include masking the first of the plurality of data elements and not masking the second of the plurality of data elements. The computer system message can be electronically communicated the to a destination to which the message is addressed.

Abstract: A computer-implemented method includes: receiving, by a computing device, information regarding a range in which an Internet-of-Things (IoT) network is to be implemented; determining, by the computing device, respective detection scores for a plurality of IoT devices for each of a plurality of proposed congregation of IoT devices; determining, by the computing device, a minimum number of the plurality of IoT devices to cover the range by incorporating the detection scores into a covariance model; and outputting, by the computing device, information identifying the minimum number of the plurality of IoT devices for designing the IoT network.

Abstract: A method (and structure and computer product) of recognizing and tracking an object includes receiving image data from at least two camera sensors and analyzing the image data to determine attributes of an object detected as present in the image data. A first decision tree (DT) is developed to classify the object, based on the attributes determined from the image data. A second DT is developed to refine the first DT, wherein the second DT is used to refine the first DT by determining which attributes are most significant in the classification of the object.

Abstract: A system and method for managing the provisioning of electricity to an electric vehicle includes identifying an electric vehicle when the electric vehicle connects to a charging point; determining whether the electric vehicle is associated with a smart contract; validating the smart contract associated with the electric vehicle; creating a charging transaction for charging the electric vehicle; and applying the smart contract associated with the electric vehicle against the charging transaction. The electric vehicle electricity contract management system and method may use a consensus system such as a distributed ledger system or blockchain.

Abstract: A computer-implemented method includes: receiving, by a computing device, information regarding an application programming interface (API); classifying, by the computing device, the API using natural language classification on to a baseline taxonomy; extracting, by the computing device, information regarding features of the API based on the classifying to the baseline taxonomy; performing, by the computing device, an assessment on the API relative to other APIs within a same class as the API based on the extracting the information regarding the API features and the classifying the API; and updating, by the computing device, the API based on the performing the assessment.

Abstract: Optimizing the allocation of farmland between different crops is provided. First and second Deep Boltzmann machines (DBMs) are built, wherein the hidden layers of the DBMs are split into a plurality of neural networks, each neural network modeling a different timeframe of crop growth. A plurality of factors related to crop growth are fed into the first DBM, which is trained to produce a first multi-class output of predicted maximum crop yields within a specified overall timeframe. The first multi-class output is fed into the second DBM, which is trained to produce a second multi-class output of predicted crop yields. The second multi-class output is fed into a decision support system that generates a recommended allocation of the farmland among different crops during different timeframes to maximize total yield.

Abstract: A computer-implemented method includes: receiving, by a computing device, information regarding a range in which an Internet-of-Things (IoT) network is to be implemented; determining, by the computing device, respective detection scores for a plurality of IoT devices for each of a plurality of proposed congregation of IoT devices; determining, by the computing device, a minimum number of the plurality of IoT devices to cover the range by incorporating the detection scores into a covariance model; and outputting, by the computing device, information identifying the minimum number of the plurality of IoT devices for designing the IoT network.

Abstract: Methods and systems for generating recommendation data to address behaviors exhibited by an entity are described. A processor may construct a finite automaton based on entity data associated with the entity. Each state of the finite automaton may represent a sentiment, and the finite automaton may accept a language representing a set of behaviors. The processor may receive a request comprising an input behavior string. The processor may apply the input behavior string on the finite automata to determine an output string. The processor may identify at least one neural model mapped to the output string, where the identified neural model comprises logic that facilitates interpretation of a cause of the behaviors among the input behavior string. The processor may generate the recommendation data using the identified neural model, where the recommendation data comprises a recommendation to address the behaviors among the input behavior string.

Abstract: Attribute-based application programming interface (API) comparative benchmarking is provided. In response to determining that a target API maps to an existing API classification based on attributes of the target API, a weighted average of benchmark confidence scores of other APIs in a same class as the target API is determined. A benchmark confidence score is determined for the target API based on feedback, reviews, and ratings. The benchmark confidence score of the target API is compared with the weighted average of benchmark scores. An attribute-based API classification mapping is updated based on the comparison. Pricing for the target API is determined based on a weighted average of API pricing across the other APIs in the same class as the target API.