Abstract: The present disclosure provides for crop yield estimation by identifying, via image processing, a field in which a crop is grown; identifying a plurality of regions within the field; identifying, by processing growth metrics via a model, a plurality of data collection points in the plurality of regions, wherein a given data collection point of the plurality of data collection points within a given region of the plurality of regions is identified by multivariate analysis as representative of growing conditions in the given region; receiving in-field data linked to the data collection points of the plurality; and predicting a yield for the crop in the field based on the in-field data.

Abstract: A method, computer system, and a computer program product for generating instructions to highlight software change may be provided. In one embodiment, the technique comprises obtaining information about a requested modification to an original code of a software program and classifying it based on the type of change requested by the modification. The unit tests available are then identified. At least one of the identified unit tests are selected and customized based on classification of the type of code modification requested. Using the at least one selected and customized unit test, the differences between the original and modified code may be identified. One or more test execution stories are then generated related to the modification of the code, to highlight the changes. Test execution stories are further analyzed to provide any additional missing information.

Abstract: In an approach to composing a machine learning model for complex data sources, a computer receives data and associated metadata corresponding to a machine learning task from a user. A computer determines a task context and a problem domain. A computer identifies the machine learning task. A computer evaluates a match between the problem domain and one or more pre-compiled models. A computer selects at least two of the one or more pre-compiled models. A computer generates one or more multimodal model combinations with the selected at least two of the one or more pre-compiled models. A computer executes the multimodal model combinations with the data and associated metadata. A computer displays the results of the executed one or more multimodal model combinations to the user. A computer determines whether a level of error associated with the results is acceptable to the user based on a response from the user.

Abstract: Aspects of the invention include receiving a first instructional video describing a first method of using a function of a first software application; determining each step of the first method described in the first instructional video; determining, by the processor, a second method of using a first function of a second software application, wherein the first function of the first software application is substantially the same as the first function of the second software application; and creating a second instructional video describing each step of the second method.

Abstract: Described are techniques for application migration. The techniques include migrating an application to a target cloud infrastructure and generating a cost-aware code dependency graph during execution of the application on the target cloud infrastructure. The techniques further include modifying the application by removing source code corresponding to unused nodes according to the cost-aware code dependency graph and replacing identified source code of a high-cost subgraph of the cost-aware code dependency graph with calls to a generated microservice configured to provide functionality similar to the identified source code. The techniques further include implementing the modified application on one or more virtual machines of the target cloud infrastructure.

Abstract: A mobile device may generate a plurality of signals based on sensor data received from a plurality of sensors associated with the mobile device. Based on a logic combination of the plurality of signals, the mobile device may detect its status indicating one of different scenarios. Based on the detected mobile device status and a configured policy, the mobile device may determine a type of deletion action for deleting data on the mobile device. The mobile device may perform the determined type of deletion action.

Abstract: Chatbot logs can be received. The chatbot logs can be clustered or classified into clusters of potential intents. Trending topics can be detected by analyzing at least social media content. Semantic similarities between the potential intents and the trending topics can be computed. Based on the computed semantic similarities, at least one of the potential intents to at least one of the trending topics can be correlated. Training of the chatbot using the chatbot logs associated with the correlated potential intent can be triggered.

Abstract: One or more processors may detect that an extreme event that occurred in a first region. One or more processors may determine that the extreme event creates a drift in a probability distribution that is an output of a model associated with the first region. One or more processors may retrain the model associated with the first region using data associated with the extreme event. One or more processors may identify a second region similar to the first region according to a defined correlation threshold. One or more processors may generate data representing potential extreme scenarios for the second region based on the detected extreme event in the first region. One or more processors may retrain a model associated with the second region based on the generated data representing potential extreme scenarios.

Abstract: A topic for building a new intent on which to train a chatbot can be received. A database of chatbot training data can be searched for a candidate intent having meta-knowledge similar to the received topic. Utterances associated with the candidate intent can be extracted. The received topic and the extracted utterances can be input to a trained machine learning model. The trained machine learning model generates example utterances for the new intent. The new intent with the generated example utterances can be used as training data for training the chatbot.

Abstract: A method of repairing crashed applications includes detecting a crash in an application operating in a host computing device. The application is migrated to a remote computer server. The remote computer server provisions computing resources to the application, while the application is resident in the remote computer server. Resumed operation of the application is executed, using the provisioned computing resources, in the remote computer server. Execution results are generated from the application, in the remote computer server. The generated execution results are migrated from the application back to the host computing device.

Abstract: Chatbot logs can be received. The chatbot logs can be clustered or classified into clusters of potential intents. Trending topics can be detected by analyzing at least social media content. Semantic similarities between the potential intents and the trending topics can be computed. Based on the computed semantic similarities, at least one of the potential intents to at least one of the trending topics can be correlated. Training of the chatbot using the chatbot logs associated with the correlated potential intent can be triggered.

Abstract: Systems and methods for provisioning a customized software stack are disclosed. A user inquiry for a technical issue is received from an inquiring user and at least one response to the user inquiry is received. A customized software stack is established for a virtual computing environment based on one or more characteristics similar to one or more user characteristics of a user computing environment of the inquiring user and the at least one response.

Abstract: An embodiment includes executing a querying process that searches for candidate code libraries to replace a current library in a software application. The embodiment also includes receiving a search result from the querying process, wherein the search result includes a set of candidate code libraries. The embodiment also includes identifying a top candidate code library as a highest ranking code library of the set of candidate code libraries based on predetermined metrics. The embodiment also includes generating a revised software application from the original software application, the generating comprising replacing the current library in the original software application with the top candidate code library.

Abstract: A method for automatically generating enhancements to audio-visual (AV) content, a content manager, and a computer program product. One embodiment may comprise, by a processing unit, analyzing data about consumer interactions to generate consumer insights about the original AV content, automatically associating the consumer insights with a segment of original AV content, automatically generating content for the segment responsive to the consumer insights, and injecting the generated content into the original AV content to create modified AV content. The embodiment may further comprise, by a network interface, receiving data about consumer interactions with original AV content and automatically transmitting the modified AV content.

Abstract: An approach to identifying architectures of machine learning models meeting a user defined constraint. The approach can receive input associated with evaluating machine learning models from a user. The approach can determine acceptable architectural templates to evaluate the machine learning models based on the input and determine a list of architectures and metrics based on a calculation of maximum neural network sizes of the acceptable architectural templates not exceeding the constraint. The approach can send the list of architectures and metrics to the user for selection.

Abstract: Aspects of the invention include a computer-implemented method, including performing simulations of a form of cyber-attack based on different input parameters to determine a respective time to perform each cyber-attack on a plurality of features of a sensor-based device. Additionally, performing simulations of a plurality of mitigating processes for each cyber-attack based on different input parameters to determine a respective time to perform each mitigating process. An associated risk level of each cyber-attack is determined based at least in part on the simulations. A mitigation process is selected based at least in part on the associated risk levels.

Abstract: Described are techniques for application migration. The techniques include migrating an application to a target cloud infrastructure and generating a cost-aware code dependency graph during execution of the application on the target cloud infrastructure. The techniques further include modifying the application by removing source code corresponding to unused nodes according to the cost-aware code dependency graph and replacing identified source code of a high-cost subgraph of the cost-aware code dependency graph with calls to a generated microservice configured to provide functionality similar to the identified source code. The techniques further include implementing the modified application on one or more virtual machines of the target cloud infrastructure.

Abstract: A microservice and the recipe for that microservice are automatically created by modifying a local environment in an integrated development environment executing on a computing system to construct a desired microservice, recording commands entered while modifying the local environment, computing a list of changes from the recorded commands that change the local environment and compiling the list of changes into a recipe comprising commands and dependencies sufficient to assemble an operating system and software files that are sufficient to instantiate the desired microservice.

Abstract: Aspects of the invention include a computer-implemented method, which includes retrieving an installation protocol for a sensor-based device. Receiving negative feedback data representing negative feedback about directions for a step in the installation protocol. Performing an installation analysis of an installation of the sensor-based device. Amending the directions, based at least in part on the installation analysis.

Abstract: Aspects of the invention include receiving a first instructional video describing a first method of using a function of a first software application; determining each step of the first method described in the first instructional video; determining, by the processor, a second method of using a first function of a second software application, wherein the first function of the first software application is substantially the same as the first function of the second software application; and creating a second instructional video describing each step of the second method.