Abstract: Techniques are described for collecting customer data through a software application in order to optimize registration form to maximize registration success rates, based on a user engagement score related to user's interest, needs, registration behavior patterns, and source of user. In some examples, a method includes evaluating user data indicating interactions of a user with a software application. The method further includes determining a user engagement score for the user based at least in part on the user data. The method further includes determining, based at least in part on the user engagement score, a number and a type of one or more registration fields. The method further includes outputting the one or more registration fields to a display device based at least in part on a triggering event.

Abstract: A computer-implemented method, computer system, and computer program product for threat management. A set of features used by a machine learning model is collected by the computer system to determine a threat type for an access attempt when the access attempt is detected. A cluster is determined, by the machine learning model in the computer system, for the access attempt using the set of features, wherein the machine learning model implements clustering to determine the cluster for the access attempt, and wherein the cluster for the access attempt corresponds to the threat type for the access attempt. A set of actions is performed by the machine learning model in the computer system based on the threat type determined for the access attempt.

Abstract: Provided are a method, system, and computer program product for representing text, in which a text is received and analyzed by utilizing a pre-trained embedding model and a feature vector model, wherein selected words in the text have corresponding weights. Operations whose parameters include weights of a feature vector and an embedding are performed to generate a weighted embedding data structure. A summation is performed of all corresponding columns of a plurality of rows of the weighted embedding data structure to generate a data structure that represents the text. The data structure that represents the text is utilized to generate at least one of a classification metadata for the text and a summarization of the text.

Abstract: A computer-implemented method, computer system, and computer program product for threat management. A set of features used by a machine learning model is collected by the computer system to determine a threat type for an access attempt when the access attempt is detected. A cluster is determined, by the machine learning model in the computer system, for the access attempt using the set of features, wherein the machine learning model implements clustering to determine the cluster for the access attempt, and wherein the cluster for the access attempt corresponds to the threat type for the access attempt. A set of actions is performed by the machine learning model in the computer system based on the threat type determined for the access attempt.

Abstract: Provided are a method, system, and computer program product for representing text, in which a text is received and analyzed by utilizing a pre-trained embedding model and a feature vector model, wherein selected words in the text have corresponding weights. Operations whose parameters include weights of a feature vector and an embedding are performed to generate a weighted embedding data structure. A summation is performed of all corresponding columns of a plurality of rows of the weighted embedding data structure to generate a data structure that represents the text. The data structure that represents the text is utilized to generate at least one of a classification metadata for the text and a summarization of the text.

Abstract: Customizing computing resource allocation based on machine learning is provided. A plurality of current users of a service are evaluated based on attributes of each current user. A user value score is generated for each current user based on the attributes. The plurality of current users is clustered into a plurality of user groups based on the user value score. A likelihood of each prospective user becoming a paying customer is predicted based on collected behavior data corresponding to each prospective user. A likelihood of an application corresponding to the service and associated with a particular user will misbehave is predicted based on resource usage and features of the application. Resource allocation to the application corresponding to the service and associated with the particular user is customized based the likelihood that the application will misbehave and whether the particular user is a current user or a prospective user.

Abstract: Training data identifying a plurality of pairs is received. Each pair identifies one or more separate code snippets known to resolve a respective issue of a plurality of issues. For each pair of the plurality of pairs, a respective issue representation of core content of the respective issue and a linear expression of the one or more separate code snippets in a respective code representation is constructed and a model to correlate the respective code representation as resolving the respective issue representation is trained. The model is queried with a new issue and a selected one of the one or more separate code snippets. The model returns a classification indicating whether the selected one of the one or more separate code snippets is likely to resolve the new issue.

Abstract: A service receives training data identifying multiple pairs, each pair identifying one or more separate code snippets known to resolve a respective issue of multiple issues. For each pair, the service constructs a respective issue representation of core content of the respective issue, constructs a linear expression of the one or more separate code snippets in a respective code representation, and trains a model to correlate the respective code representation as resolving the respective issue representation. The service queries the model with a new issue and a selected one of the one or more separate code snippets, wherein the model returns a classification indicating whether the selected one of the one or more separate code snippets is likely to resolve the new issue.

Abstract: Embodiments relate to a computer program product, a computer system and a method using artificial intelligence for dynamically determining sub-category ratings from content commentary of reviews of an online review forum, such as, for example, venue reviews of a crowd-source review forum. In particular embodiments, the computer program product, the computer system, and the method apply the artificial intelligence for dynamically determining sub-category ratings from content commentary of reviews of an online review forum based on personal characteristic data of an entity (e.g., user) profile or entity online history.

Abstract: A method for identifying a user of a website as a potential customer and presenting the potential customer with enhanced website content includes: identifying that a user has landed on the website; identifying a user traffic-source; determining if the identified traffic-source matches a traffic-source in a list of customer-based reference traffic-sources; identifying a specific user behavior pattern on the website; determining if the identified specific user behavior pattern matches a behavior pattern from a list of customer-based reference behavior patterns; identifying the user as a potential customer if the identified traffic-source matches a traffic-source in a list of reference traffic sources and the identified specific user behavior pattern matches a behavior pattern from a list of reference behavior patterns; and displaying enhanced content tailored to the potential customer based upon the specific user behavior pattern or excluding display of enhanced content to the user if not a potential customer.

Abstract: Customizing computing resource allocation based on machine learning is provided. A plurality of current users of a service are evaluated based on attributes of each current user. A user value score is generated for each current user based on the attributes. The plurality of current users is clustered into a plurality of user groups based on the user value score. A likelihood of each prospective user becoming a paying customer is predicted based on collected behavior data corresponding to each prospective user. A likelihood of an application corresponding to the service and associated with a particular user will misbehave is predicted based on resource usage and features of the application. Resource allocation to the application corresponding to the service and associated with the particular user is customized based the likelihood that the application will misbehave and whether the particular user is a current user or a prospective user.

Abstract: An artificial intelligence system monitors and records usage of a set of mobile devices that have been “paired” or grouped together. The system, over time, accumulates a history of times and locations associated with events that occur with respect to the devices. From this history, the system develops a usage model for the devices. The system, if it detects current usage diverges significantly from an expected usage, performs responsive actions which may include: notifying a user, via a paired device, that the mobile device may be lost or misplaced; command the “lost” mobile device to enter a reduced power state, to prolong battery life; and/or command the “lost” device to emit audible and/or visible signals intended to facilitate locating the device. In some embodiments, the system predicts an imminent loss of the device before it becomes lost, based on current usage vs. expected usage embodied in the model.

Abstract: An approach is provided for determining a product value score. Features of a product are identified. Ranking data for each identified feature is collected. The ranking data includes popularity, stability, availability, and memory usage of each feature. Additional data for each feature are determined. The additional data includes an importance, a conversion rate, and a marketing impact of each feature. Based on the ranking data and the additional data, normalized data is generated according to characteristics of the features. Based on a data science prediction analysis which uses the identified features, weights of the features are determined. Based on initial results of the data science prediction analysis, a key product is selected from a set of products. Based on the key product and the weights, product value scores of products in the set of products are determined.

Abstract: A mechanism is provided in a data processing system to implement a chat topic identification system for identifying chat topics in a multi-user messaging platform. The chat topic identification system receives a plurality of chat messages submitted to a communication channel in the multi-user messaging platform, performs cognitive content language analysis, cognitive personality analysis, cognitive tone analysis, social data analysis, and conversation feature analysis. The chat topic identification system stores results in a user profile associated with the user to form a user profile. A new conversation prediction engine determines a conversation separation score representing a likelihood that a new conversation is being started in the communication channel. A new conversation separation recommendation engine generates a conversation separation recommendation based on the conversation separation score.

Abstract: Techniques are described for collecting customer data through a software application in order to optimize registration form to maximize registration success rates, based on a user engagement score related to user's interest, needs, registration behavior patterns, and source of user. In some examples, a method includes evaluating user data indicating interactions of a user with a software application. The method further includes determining a user engagement score for the user based at least in part on the user data. The method further includes determining, based at least in part on the user engagement score, a number and a type of one or more registration fields. The method further includes outputting the one or more registration fields to a display device based at least in part on a triggering event.

Abstract: Techniques are described for collecting customer data through a software application in order to optimize registration form to maximize registration success rates, based on a user engagement score related to user's interest, needs, registration behavior patterns, and source of user. In some examples, a method includes evaluating user data indicating interactions of a user with a software application. The method further includes determining a user engagement score for the user based at least in part on the user data. The method further includes determining, based at least in part on the user engagement score, a number and a type of one or more registration fields. The method further includes outputting the one or more registration fields to a display device based at least in part on a triggering event.

Abstract: Techniques provide include analyzing a first node, associated with a first version of an service, to extract first configuration metadata; analyzing the first node, associated with a second version, subsequent to the first version, of the aservice, to extract second configuration metadata; determining, based upon a comparison of the first configuration metadata and the second configuration metadata, that the first node has not been modified between the first version and the second version; generating a first image of the node for distribution in conjunction with the service in response to the determining that the particular node has not been modified; and caching the first image for use in conjunction with the service.

Abstract: An approach is provided for determining a product value score. Features of a product are identified. Ranking data for each identified feature is collected. The ranking data includes popularity, stability, availability, and memory usage of each feature. Additional data for each feature are determined. The additional data includes an importance, a conversion rate, and a marketing impact of each feature. Based on the ranking data and the additional data, normalized data is generated according to characteristics of the features. Based on a data science prediction analysis which uses the identified features, weights of the features are determined. Based on initial results of the data science prediction analysis, a key product is selected from a set of products. Based on the key product and the weights, product value scores of products in the set of products are determined.

Abstract: A cognitive churn-analysis system of a customer-engagement management platform predicts when and why a customer is likely to reduce a current level of service of a current service offering. The system enhances results of a conventional statistical churn analysis by using cognitive methods to infer a customer's personality traits and evolving sentiment toward the service offering. By processing the results of the statistical analysis and the two cognitive analyses, the system predicts a likelihood that a customer will reduce service at a particular time by performing a churning activity. After further determining a likely reason why the customer would choose to churn, the cognitive system directs the customer-engagement management platform to educe the likelihood that the customer would make such a decision by engaging the customer with remedial actions at an optimal time.

Abstract: A mechanism is provided in a data processing system to implement a chat topic identification system for identifying chat topics in a multi-user messaging platform. The chat topic identification system receives a plurality of chat messages submitted to a communication channel in the multi-user messaging platform. The chat topic identification system performs cognitive content language analysis, cognitive personality analysis, cognitive tone analysis, social data analysis, and conversation feature analysis. The chat topic identification system stores results of the cognitive content language analysis, the cognitive personality analysis, the cognitive tone analysis, and the social data analysis in a user profile associated with the user to form an updated user profile. A new conversation prediction engine determines a conversation separation score representing a likelihood that a new conversation is being started in the communication channel.