International Business Machines Corporation

Yan Luo

Yan Xu

Liujia Shao

Sibin Fan

Gary Mazo

Peter Haumer

Monvorath Phongpaibul

Tala El Hallak

Stanislav Georgiev

Seana Hogan

John Carl DelMonaco

William John Scheid, Iii

Hongxia Li

Rahamim Katan

An approach to code refactor renaming may be provided. Source code with a naming convention for functions and classes can be presented to a machine learning model. The model may identify the names for functions and classes. The identified names may be tokenized. Docstrings associated with functions and classes may be identified. Code for the identified functions and classes and associated may be input into a feature vector generation mechanism. A model may be trained mapping the generated feature vectors to tokenized identified names, via regression. The model can be utilized to analyze input code with the same naming convention to predict names for functions and classes, allowing for the recommendation of function and class names in accordance with the programming code naming convention.

Code refactor renaming recommender

Techniques for context-based word embedding for programming artifacts are described herein. An aspect includes determining a plurality of keywords based on a corpus of programming artifacts, the corpus of programming artifacts including source code corresponding to a software project. Another aspect includes determining a plurality of context/keyword pair sets based on the plurality of keywords and the corpus of programming artifacts, wherein each context/keyword pair set of the plurality of context/keyword pair sets includes a first keyword, a second keyword, and a context type corresponding to a co-occurrence of the first keyword and the second keyword in the corpus of programming artifacts. Another aspect includes constructing a word embedding matrix based on the plurality of context/keyword pair sets.

Context-based word embedding for programming artifacts

Techniques for context-based keyword grouping for business rule mining are described herein. An aspect includes determining, based on a first corpus, a first list of keywords. Another aspect includes constructing a co-occurrence matrix based on the first list of keywords. Another aspect includes applying a clustering algorithm to the co-occurrence matrix to determine a first plurality of keyword groups. Another aspect includes presenting the first plurality of keyword groups to a user via a user interface.

Context-based keyword grouping

In an approach to creating code snippet auto-commenting models utilizing a pre-training model leveraging dependency data, one or more computer processors create a generalized pre-training model trained with one or more dependencies and one or more associated dependency embeddings, wherein dependencies include frameworks, imported libraries, header files, and application programming interfaces associated with a software project. The one or more computer processors create a subsequent model with a model architecture identical to the created pre-training model. The one or more computer processors computationally reduce a training of the created subsequent model utilizing one or more trained parameters, activations, memory cells, and context vectors contained in the created pre-training model. The one or more computer processors create deploy the subsequent model to one to more production environments.

Pretraining utilizing software dependencies

Methods, systems, and computer program products for automated breakpoint creation using machine learning are provided. Aspects include obtaining a bug report for a software and source code for the software and analyzing the bug report to determine a bug type for the bug report, where analyzing the bug report includes using a bug type labeling model. Aspects also include analyzing the source code to identify a code snippet in the source code based on the bug type, where analyzing the source code includes using a source code detection model. Aspects further include inserting a breakpoint in the source code at the code snippet.

Automated breakpoint creation

Aspects include a method for generating an initial ranking of a plurality of discovered keywords, in which the initial ranking is based on confidence levels. The method includes presenting the initial ranking to a user as a list that includes the plurality of discovered keywords. The method includes, in response to receiving a text string as a first input from the user, filtering the list of discovered keywords to generate a refined list of discovered keywords. The filtering includes searching the list of discovered keywords to rank keywords that match the input string using a fuzzy search. The method includes mapping a discovered keyword of the refined list to a business term using a fuzzy search. The method includes updating the initial ranking based on the mapping of the discovered keyword.

Discovering and displaying business artifact and term relationships

In an approach to generating natural language rules based on detected code snippets, one or more computer processors detect a code snippet. The one or more computer processors extract code information from the detected code snippet. The one or more computer processors feed the extracted code information into a cognitive model, wherein the cognitive model utilizes one or more historical code snippets based on the extracted code information and one or more natural language rules based on the extracted code information. The one or more computer processors generate, based on one or more calculations by the cognitive model, a natural language rule for the code snippet.

Predictive natural language rule generation

Techniques for automatic code coverage file recommendation are described herein. An aspect includes receiving historical code coverage data. Another aspect includes clustering the historical code coverage data. Another aspect includes performing content filtering based on the clustered historical code coverage data to determine a content filtering preferred file list. Another aspect includes performing collaborative filtering based on the clustered historical code coverage data to determine a collaborative filtering preferred file list. Another aspect includes combining the content filtering preferred file list and the collaborative filtering preferred file list to determine a code coverage file recommendation list. Another aspect includes providing the code coverage file recommendation list to a user.

Automatic code coverage file recommendation

A computer-implemented method includes training a classification model to produce a mapping from a set of feature vectors to a set of confidence levels. Each feature vector describes a respective term, and each confidence level indicates a likelihood that the respective term is a keyword. A plurality of program artifacts are tokenized into a plurality of terms. For each term in the plurality of terms, a respective feature vector is determined for describing the term, based on the plurality of program artifacts. For each term in the plurality of terms, based on the respective feature vector, using the mapping, a respective confidence level is determined for indicating a likelihood that the term is a keyword of the plurality of program artifacts.

Predicting keywords in an application

Embodiments of the invention include methods, systems, and computer program products for root cause analysis. Aspects of the invention include receiving, by a processor, operations data associated with a plurality of applications. A trend analysis is performed on the operations data to determine an operations issue associated with at least one of the plurality of applications. And a root-cause analysis is performed on the operations issue to identify a set of candidate applications from the plurality of applications that may be a cause of the operations issue.

Growing community of inventors

Concord, CA, United States of America

Yan Luo