Abstract: An example generator tool generates an example illustrating correct usage of a command of a command line interface. A command may include a command name, zero or more subcommands, and one or more parameters with a corresponding parameter value. A template containing the correct syntax of the command is obtained from a template database. Parameter values for the template are generated from a neural transformer with attention given the command template.

Abstract: A source code generation system uses a neural transformer model with attention to predict candidate method bodies given a method docstring, method signature, and one or more method templates. The method templates are derived from intent-snippet pairs from StackOverflow question/answer pairs or template methods from GitHub. Joint embeddings are generated for the method bodies of the method templates and associated method docstrings for quick retrieval. A code completion system uses the source code generation system to generate candidate method bodies to complete a method signature and/or method docstring using the method templates.

Abstract: An automated program repair system uses a neural transformer model with attention to predict a bug-free version of a method having a source code bug identified in an associated stack trace. The neural transformer model is pre-trained with English language text and the source code of a target programming language. The pre-trained neural transformer model is trained to create synthetic bugs in bug-free methods. The bug-free methods with the synthetic bugs are executed with a test case to obtain a stack trace of the source code bug. The method with the synthetic bug, without the bug, and its stack trace are used to train the neural transformer model to predict repairs for buggy methods.

Abstract: Runtime errors in a source code program are detected in advance of execution by machine learning models. Features representing a context of a runtime error are extracted from source code programs to train a machine learning model, such as a random forest classifier, to predict the likelihood that a code snippet has a particular type of runtime error. The features are extracted from a syntax-type tree representation of each method in a program. A model is generated for distinct runtime errors, such as arithmetic overflow, and conditionally uninitialized variables.

Abstract: A data mining technique is used to find large frequently-occurring source code patterns from methods/APIs that can be used in code development. Simplified trees that represent the syntactic structure and type and method usage of a source code fragment, such as a method, are mined to find closed and maximal frequent subtrees which represent the largest frequently-occurring source code patterns or idioms associated with a particular type and method usage. These idioms are then used in an idiom web service and/or a code completion system to assist users in the development of source code programs.

Abstract: A neural transformer model with attention is trained to predict candidates to complete a line of source code with a zero-inference capability. The model is trained on an unsupervised training dataset that includes features from source code written in multiple programming languages. The features include a file-level context and a local context, where the file-level context includes a global context, a class context, a function context, and/or a method context for each class, function and/or method of the source code programs used in the training dataset. The local context includes method bodies, function bodies, and/or stand-alone code of main method routines. From these features, the model is able to learn to predict an ordered sequence of code elements that complete a line of source code in a programming language seen and not seen during training.

Abstract: A code completion tool uses a deep learning model to predict the likelihood of a method completing a method invocation. In one aspect, the deep learning model is a LSTM trained on features that represent the syntactic context of a method invocation derived from an abstract tree representation of the code fragment.

Abstract: An automated system for resolving program merges uses neural transformers with attention. In one aspect, a neural encoder transformer model is trained from developer-resolved merge conflicts to learn to predict a resolution strategy that aids a developer in constructing a merged program. In a second aspect, a neural decoder transformer model is trained on the syntax and semantics of different source code programming languages to predict a merge resolution consisting of interleaved lines of source code from programs A, B, or O, where programs A and B contain changes to code base O.

Abstract: An automated system for resolving program merges uses a sequence-to-sequence supervised machine learning model trained from developer-resolved merge conflicts to learn to predict a merge resolution to resolve a three-way program merge. The model utilizes an embedding of the merge tuple (A, B, O) which represents the program syntax, program semantics and the intent of the program inputs. The model uses a pointer mechanism to construct the resolved program in terms of the lines of source code found in the input programs.

Abstract: A code completion tool uses a neural transformer model to generate candidate sequences to complete a line of source code. The neural transformer model is trained using a conditional language modeling objective on a large unsupervised dataset that includes source code programs written in several different programming languages. The neural transformer model is used within a beam search that predicts the most likely candidate sequences for a code snippet under development.

Abstract: A code completion system uses neural components to rank the unordered list of code completion candidates generated from an existing static analyzer. The candidates represent the next sequence of tokens likely to complete a partially-formed program element as a developer is typing in a software development tool. A re-ranking component generates a ranked order of the candidates based on a context embedding of the code context and candidate embeddings of the candidates, where both embeddings are based a common token encoding.

Abstract: The present disclosure is directed to providing supplemental content to one or more client devices requesting multimedia content. The supplemental content may be determined or selected based on the availability of one or more device channels, such as a display, speakers, or other component of the one or more client devices capable of providing an output. The supplemental content may also be selected based on one or more characteristics of the requested multimedia content, such as a genre, subject matter, or duration. Furthermore, the supplemental content may be determined or selected based on the portions of the requested multimedia content that are the most prominent or significant, such as any audio content, any video content, and/or any textual content. The supplemental content may be provided to the one or more client devices such that it is displayed before, during, or after the display of the requested multimedia content.

Abstract: A unit test generation system employs a neural transformer model with attention to generate candidate unit test sequences given a focal method of a programming language. The neural transformer model is pre-trained with source code programs and natural language text and fine-tuned with mapped test case pairs. A mapped test case pair includes a focal method and a unit test case for the focal method. In this manner, the neural transformer model is trained to learn the semantics and statistical properties of a natural language, the syntax of a programming language and the relationships between the code elements of the programming language and the syntax of a unit test case.

Abstract: An assert statement generator employs a neural transformer model with attention to generate candidate assert statements for a unit test method that tests a focal method. The neural transformer model is pre-trained with source code programs and natural language text and fine-tuned with test-assert triplets. A test-assert triplet includes a source code snippet that includes: (1) a unit test method with an assert placeholder; (2) the focal method; and (3) a corresponding assert statement. In this manner, the neural transformer model is trained to learn the semantics and statistical properties of a natural language, the syntax of a programming language, and the relationships between the code elements of the programming language and the syntax of an assert statement.

Abstract: A code completion tool uses a neural transformer model to generate candidate sequences to complete a line of source code. The neural transformer model is trained using a conditional language modeling objective on a large unsupervised dataset that includes source code programs written in several different programming languages. The neural transformer model is used within a beam search that predicts the most likely candidate sequences for a code snippet under development.

Abstract: Language interoperability between source code programs not compatible with an interprocedural static code analyzer is achieved through language-independent representations of the programs. The source code programs are transformed into respective intermediate language instructions from which a language-independent control flow graph and a language-independent type environment is created. A program compatible with the interprocedural static code analyzer is generated from the language-independent control flow graph and the language-independent type environment in order to utilize the interprocedural static code analyzer to detect memory safety faults.

Abstract: An interactive question and answer (Q&A) service provides pairs of questions and corresponding answers related to the content of a web page. The service includes pre-configured Q&A pairs derived from a deep learning framework that includes a series of neural networks trained through joint and transfer learning to generate questions for a given text passage. In addition, pre-configured Q&A pairs are generated from historical web access patterns and sources related to the content of the web page.

Abstract: A code completion tool uses machine learning models to more precisely predict the likelihood of a method invocation completing a code fragment that follows one or more method invocations of different classes in a same document during program development. In one aspect, the machine learning model is a n-order Markov chain model that is trained on features that represent characteristics of the context of method invocations found in commonly-used programs from a sampled population. The machine learning model is implemented as a hash table contained a ranked order of hash values in descending order of probability of completing a partially-formed method invocation.

Abstract: The present disclosure is directed to providing supplemental content to one or more client devices requesting multimedia content. The supplemental content may be determined or selected based on the availability of one or more device channels, such as a display, speakers, or other component of the one or more client devices capable of providing an output. The supplemental content may also be selected based on one or more characteristics of the requested multimedia content, such as a genre, subject matter, or duration. Furthermore, the supplemental content may be determined or selected based on the portions of the requested multimedia content that are the most prominent or significant, such as any audio content, any video content, and/or any textual content. The supplemental content may be provided to the one or more client devices such that it is displayed before, during, or after the display of the requested multimedia content.

Abstract: Systems and methods of visual mining of user behavior patterns are disclosed. A plurality of clickstreams may be received. Each clickstream may represent a corresponding sequence of user actions. A visual representation of each clickstream may be caused to be displayed on a device. Each visual representation may comprise a distinct graphical element for each user action of the corresponding clickstream, and each visual representation may be configured to indicate a frequency level of the corresponding clickstream. Each distinct graphical element may comprise a geometric shape and a corresponding color that distinctly represents the corresponding user action. Graphical elements of each visual representation may comprise a size that is proportional to the frequency level of the corresponding clickstream. An indication of a selection of one of the visual representations may be received, and additional information about the corresponding clickstream may be caused to be displayed on the device.