Abstract: A plurality of HyperText Markup Language (HTML) strings corresponding to a dataset of document object model (DOM) nodes is tokenized according to a tokenization scheme to produce a dictionary of tokens that occur in the dataset. A condensed dictionary of tokens in produced by removing low-value tokens from the vocabulary of tokens. An information matrix is computed based on the condensed dictionary of tokens, the information matrix being a set of values, a value of the set of values corresponding to a frequency of co-occurrence of a pair of tokens. A library of word vectors is derived from the information matrix. A feature vector of an HTML element is generated based at least in part on the library of word vectors, and a classification for the HTML element is obtained from a machine learning model as a result of inputting the feature vector into the machine learning model.

Abstract: A first interface having a first object, a second object, and a third object is executed, where the third object is associated with a first option and a second option. The first option of the third object is selected. The second object is engaged a first time. The first object is engaged to execute a second interface in a first state, the second interface in the first state including a first data item associated with the first option. The second option of the third object is selected. The second object is engaged a second time. The second interface is re-executed to produce a second state. Integration code is generated as a result of a verification. The integration code is provided to a device, thereby causing the device to select one of the first option or the second option of the third object of the first interface.

Abstract: A baseline request produced from an annotated node of a document object model (DOM) tree and a label assigned to the annotated node are obtained. The label is assigned to a set of neighboring nodes of the DOM that perform a same function by recursively causing the system to, for each neighboring node to the annotated node in the DOM tree an additional request produced in response to performance of simulated human interaction with the neighboring node is identified, if the additional request matches the baseline request, the label is assigned to the neighboring node, and the neighboring node is selected to be the annotated node.

Abstract: Integration code usable to cause a computing device to determine which category from a plurality of categories corresponds to an interface of an interface provider is generated based at least in part on output from a machine learning algorithm trained to categorize interfaces. The computing device is caused, by providing the integration code to the computing device, to execute the integration code to cause the computing device to evaluate characteristics of an interface of an interface provider, determine a category of an interface of the interface provider, and interact with the interface in a manner that accords with the category.

Abstract: A request is received from a computing device for substitute data, with access to the substitute data being contingent upon successful multi-factor authentication of the first service. Signature data based on the request is generated using a first key of public-private key pair. Credential proof and the signature is provided to a second service, which verifies the credential proof as a first factor of the multi-factor authentication and verifies, using a second key, the signature as a second factor of the multi-factor authentication. The substitute data is obtained as a result of authentication by the second service. The computing device is caused, by providing the substitute data to the computing device, to input the substitute data into the interface in place of data associated with the first entity.

Abstract: A first set of records and a second set of records different from the first set of records are obtained. A score that indicates indicating confidence that the first set of records and the second set of records correspond to a same individual is computed. As a result of the score reaching a value relative to a threshold, a device of an individual associated with the second set of records is caused to prompt the individual to confirm control of a first asset identified by first data within the first set of records. A message to a destination associated with the first asset is provided. As a result of a response to the message indicating that the individual is the same individual, an association between the second set of records and the first asset is stored and the device is caused to display details about the first and second sets of records.

Abstract: A first interface having a first control object and a second control object is executed, where the first control object is associated with a value, a label, and a digital image. The first control object in engaged. The second control object is engaged to execute a second interface. That the second interface includes an occurrence of the label is identified. Locations of the label and the digital image in the first interface and the occurrence of the value in the second interface are determined. Integration code is generated based on the locations. The integration code is provided to the device to cause, as a result of execution of the integration code by a device, the device to display data obtained based on the locations.

Abstract: A first hierarchy path associated with a first interface of a first provider is determined, with the first interface being of a first type. A second hierarchy path associated with a second interface of the first provider is determined, with the second interface being of a second type different from the first type. The first hierarchy path is determined to be a mismatch from the second hierarchy path. A third hierarchy path corresponding to another interface of the first type is obtained from a second provider. A dictionary is generated based on a determination that the third hierarchy path matches the first hierarchy path. A vector that corresponds to a third interface is generated based on the dictionary. As a result of the third interface being determined to be of the first type based on the vector, a device is caused to perform an operation against the third interface.

Abstract: A first interface is accessed. Initial data of a second interface different from the first interface is obtained. A first object in the first interface that is associated with a first value is engaged. Refreshed data of the second interface is obtained. A correlation between the first value and a second object in the refreshed data of the second interface is determined at least in part as a result of a difference between the initial data and the refreshed data corresponding to the first value. That the second object is a final object and the second interface is a final interface is determined based on the correlation. Integration code is generated based on a location of the final object in the second interface, where the integration code causes the device to engage the first object and engage a second object in the first interface to execute the final interface.

Abstract: Software program code of an interface is obtained. A particular object in the program code is identified by at least causing the computer system to identify a set of candidate objects in the software program code and, for each candidate object of the set of candidate objects, produce a modified set of candidate objects. The system is caused to at least simulate human interaction with the candidate object, detect a result of simulating the human interaction, and modify a ranking of the candidate object based on the result. The particular object is determined from the modified set of candidate objects based on the ranking. Integration code based on the particular object. Integration code is generated based on the particular object. The integration code is provided to a device to cause, as a result of execution by the device, the device to engage the particular object in the interface.

Abstract: A first interface having a control object that is associated with a first option and a second option is launched. The control object is engaged at a first time to select the first option. A first set of differences between a first current state of the interface and a preconfigured state of the first interface is determined. The control object is engaged at a second time to select the second option. A second set of differences between a second current state of the interface and the preconfigured state of the first interface is determined. Integration code that, upon execution by a client device, causes the client device to modify a second user interface is generated based on the first set of differences or the second set of differences. The integration code is provided to the client device to cause the client device to modify the second user interface.

Abstract: A first characteristic associated with a first interface of an interface type and a second characteristic associated with a second interface of an additional interface type are determined. A machine learning algorithm is trained using the first characteristic, the second characteristic, and the interface type as a ground truth for the first characteristic. A template script usable to identify whether a given interface of the same interface provider is of the interface type or the additional interface type is generated. The parameter for the template script is determined based the machine learning algorithm. A device is caused to, as a result of the device executing the template script with the parameter to identify that the given interface is of the interface type, perform an operation specific to the interface type.

Abstract: An input value is derived at least in part from information associated with present data associated with a user and details of historical data associated with the user, with the historical data including a value indicating an option selected by the user and the option being a member of a plurality of options available to associate with the present data. A plurality of scores is obtained at least in part by passing, for each of the plurality of options, the input value as input to a plurality of decision rules. A preferred option is determined based at least in part on the plurality of scores. An interface is caused to display a plurality of selectable elements corresponding to the plurality of options that include the preferred option. A response that indicates selection of the preferred option is received via the interface. The present data is associated with the preferred option.

Abstract: Embodiments described herein simplify ordering of products from online shops. This may involve an agent server receiving captured contents of a product webpage from a client device and autonomously deriving product information therefrom. Alternatively, the agent service can receive a web address corresponding to a product webpage, and the agent server can visit the product webpage and capture its contents. The agent server then uses the derived product information to autonomously control ordering of the product for the user from an online shop, without requiring that the user have an account with the online shop, without requiring that the user enter personal and payment information in a check-out webpage of the online shop, without requiring that the user select a buy or shopping cart button associated with the online shop, and without requiring the online shop utilize an API to enable the agent server to interact with the online shop.