Fujitsu Corporation

Fujistu Limited

Mukul Ranjan Prasad

Indradeep Ghosh

Hiroaki Yoshida

Ripon K Saha

Sreeranga P Rajan

Paparao Palacharla

Sonal Mahajan

Jawahar Jain

Xi Wang

Alexander Horn

Oksana Tkachuk

Rajarshi Mukherjee

Praveen Murthy

Subramanian K Iyer

Maarten H Wiggers

Wei Yang

Thomas W Sidle

Shauvik Roy Choudhary

Lei Liu

Naoki Oguchi

Mehdi Bahrami

Fan Yu

Ali Kheradmand

Michael S Hsiao

Frolin S Ocariza, Jr

Xiang Gao

Ali Mesbah

Casper S Jensen

Wenyu Wang

Fangqi Sun

Kelvin K C Ng

Seemanta Saha

Chenguang Zhu

Yinlin Deng

Eric Campbell

Rohan Bavishi

Alex Horn

Benjamin Mehne

Martin J Rogalla

Shin Hwei Tan

Divya Gopinath

Zhiqiang Zang

Qianqian Wang

Ghosh Indradeep

A method may include obtaining a dataset that may include at least a first column and a second column, with a first title and a first value and a second title and a second value, respectively. The operations may further include determining a total similarity value between the first and second columns based on a metadata similarity value, a semantic similarity value, and/or a unit of measurement similarity value; adding the first and second columns to a cluster if the total similarity value is less than a threshold value; generating a new column to add to the cluster using a feature engineering function, where the new column may include a new title and a new value, the new value determined using the feature engineering function by acting on the first value and/or the second value. Further, the operations may include adding the new column to the dataset.

Unit prediction for custom feature engineering

According to one or more embodiments, operations may include normalizing machine learning (ML) pipelines of existing ML projects stored in a corpus of existing ML projects. The operations may also include extracting functional blocks from the normalized ML pipelines. In addition, the operations may include assigning a label to each of the functional blocks in the normalized ML pipelines. Moreover, the operations may include indexing each of the ML pipelines in the corpus based on the labels assigned to the functional blocks. In addition, the operations may include utilizing the labels assigned to the functional blocks in the corpus to generate a new pipeline to perform a new ML task on a new dataset of a new ML project.

Automatically labeling functional blocks in pipelines of existing machine learning projects in a corpus adaptable for use in new machine learning projects

Operations may include obtaining a repair related to correcting an error of source code of a software program, the repair including an edit to make to the source code. The operations may further include determining a change type of the edit. The operations may additionally include identifying an explanation template from a library of explanation templates based on the change type corresponding to the explanation template in the library of explanation templates. In addition, the operations may include generating a change explanation for the edit based on the explanation template. The change explanation may provide a natural language explanation of the changes made by the edit.

Generation of software program repair explanations

Operations include obtaining a first patch that corrects a first error in a first buggy code snippet of first source code based on the first buggy code snippet and a first repaired code snippet. The operations also include generating a second patch based on the first patch and a bug pattern of a bug scenario that corresponds to the first error. In addition, the operations include generating a third patch based on the second patch, the bug pattern, and a second buggy code snippet of second source code, the third patch correcting a second error in the second buggy code snippet. Moreover, the operations include performing one or more repair operations with respect to the second buggy code snippet based on the third patch.

Automated generation of software patches

Operations may include obtaining a buggy code snippet of source code of a software program in which the buggy code snippet includes a particular error. The operations may also include determining a respective first similarity between the buggy code snippet and a plurality of bug patterns of previously identified bug scenarios. In addition, the operations may include selecting a particular bug pattern based on a determined particular first similarity between the particular bug pattern and the buggy code snippet. Moreover, the operations may include determining a respective second similarity between the particular bug pattern and example code snippets obtained from a plurality of posts. The operations may also include selecting a particular post as providing a potential solution to correct the particular error based on a determined particular second similarity between the particular bug pattern and a particular example code snippet of the particular post.

Automated identification of posts related to software patches

According to one or more embodiments, operations may include storing existing machine learning (ML) projects in a corpus. The operations may also include generating a search query for a new ML project based on a new dataset and a new ML task for the new ML project. In addition, the operations may include searching through the existing ML projects stored in the corpus, based on the search query, for a set of existing ML projects. Moreover, the operations may include merging the ML pipelines of the set of existing ML projects to generate a new ML pipeline for the new ML project. In addition, the operations may include adapting functional blocks of the new ML pipeline for the new ML project to enable the new ML pipeline to be executed to perform the new ML task on the new dataset.

Automatically generating a pipeline of a new machine learning project from pipelines of existing machine learning projects stored in a corpus

According to an aspect of an embodiment, operations may include selecting, from a training dataset, a first data point as a seed data point. The operations may further include generating a population of data points by application of a genetic model on the seed data point. The population of data points may include the seed data point and a plurality of transformed data points of the seed data point. The operations may further include determining a best-fit data point in the generated population of data points based on application of a fitness function on the generated population of data points. The operations may further include executing a training operation on the DNN based on the determined best-fit data point. The operations may further include obtaining a trained DNN for the first data point based on the training operation on the DNN based on the determined best-fit data point.

Data augmentation in training deep neural network (DNN) based on genetic model

According to an aspect of an embodiment, operations may include receiving a first data point associated with a real-time application and predicting a first class for the received first data point, by a Deep Neural Network (DNN) pre-trained for a classification task of the real-time application. The operations may further include extracting, from the DNN, a first set of features and a corresponding first set of weights, for the predicted first class. The extracted first set of features may be associated with a convolution layer of the DNN. The operations may further include determining, by a pre-trained classifier associated with the predicted first class, a confidence score for the predicted first class based on the extracted first set of features and the corresponding first set of weights. The operations may further include generating output information to indicate correctness of the predicted first class based on the determined confidence score.

Validation of deep neural network (DNN) prediction based on pre-trained classifier

According to one or more embodiments, operations may include gathering a set of machine learning (ML) projects from one or more repositories of ML projects based on a filtering criteria. The operations may also include ensuring executability of ML pipelines in the set of ML projects. In addition, the operations may include identifying irrelevant portions of the ML pipelines in the set of ML projects. Moreover, the operations may include generating quality features for the set of ML projects. In addition, the operations may include generating diversity features for the set of ML projects. Moreover, the operations may include selecting a subset of ML projects from the set of ML projects based on the quality features and the diversity features. In addition, the operations may include storing the subset of ML projects in a corpus of ML projects that may be adapted for use in new ML projects.

Automatically curating existing machine learning projects into a corpus adaptable for use in new machine learning projects

Operations include obtaining a machine learning (ML) pipeline skeleton that indicates a set of first functional blocks to use to process a new dataset of a new ML project. The operations also include obtaining a relationship mapping that maps dataset features to respective functional blocks, the relationship mapping indicating correspondences between dataset features of existing datasets of existing ML projects and usage of second functional blocks of existing ML pipelines of the existing ML projects. The operations also include mapping the first functional blocks to respective portions of the new dataset based on the relationship mapping. In addition, the operations include instantiating the pipeline skeleton with respective code snippets that each correspond to a respective first functional block of the set of first functional blocks, the respective code snippets each including one or more respective code elements that are based on the mapping of the first functional blocks.

Machine learning pipeline skeleton instantiation

Operations may include obtaining a buggy code snippet from a question included in a post of a discussion forum, the buggy code snippet including an error and a plurality of lines of code. The operations may further include obtaining text from an answer to the question and identifying keywords from the text. In addition, the operations may include marking a particular line from the plurality of lines as being related to the error based on one or more of the keywords corresponding to one or more elements of the particular line. Moreover, the operations may include performing one or more software program repair operations based on the marking of the particular line.

Automated identification of lines of code related to errors field

According to one or more embodiments, operations may include identifying a plurality of source code edits made between a buggy version of first source code of a first software program and a repaired version of the first source code. The operations may also include identifying a plurality of clusters of the source code edits. Each cluster of the plurality of clusters includes one or more source code edits of the plurality of source code edits. The operations may also include identifying a plurality of valid clusters of the plurality of clusters and identifying one or more invalid clusters of the plurality of clusters based on compiling of the first source code. Moreover, the operations may include identifying, as a first repair of the first error, a particular set of one or more valid clusters of the plurality of valid clusters while excluding the one or more invalid clusters from consideration.

Automated software program repair

Operations may include obtaining a code artifact in which the code artifact is a code snippet is of a software language of interest. The operations may also include identifying fault locations of the code snippet that each correspond to a parsing error. In addition, the operations may include implementing repairs for each identified fault location in which, for each fault location, a respective repair is implemented for its corresponding parsing error such that the code snippet is modified into a parsable code snippet.

Parsability of code snippets

Operations may include generating a first graphical representation of a buggy version of source code of a software program and generating a second graphical representation of a repaired version of the source code. The buggy version may include an error and the error has been corrected in the repaired version. The operations may also include identifying, as matching nodes, a first node of the first graphical representation and a second node of the second graphical representation that match each other. The operations may additionally include applying a same reduction to the first node and to the second node concurrently to generate a reduced buggy version of the source code and a reduced repaired version of the source code. Further, the operations may include generating a repair example with respect to correcting the error in which the repair example is based on the reduced buggy version and the reduced repaired version.

Generation of software program repair examples

A method may include obtaining a first violation in a first software program and identifying one or more characteristics of the first violation. The method may include searching at least one website to identify a set of posts. The method may include obtaining, as a first context, a first portion of the first software program that includes the first violation. The method may include obtaining a set of second contexts, each second context corresponding to a different post of the set of posts. The method may also include determining a set of similarity scores by comparing a first abstract tree representation of the first context with each second abstract tree representation corresponding with the second contexts and selecting a particular post based on the similarity score between the first abstract tree representation and the particular second abstract tree representation associated with the particular post.

Automated searching and identification of software patches

Growing community of inventors

San Jose, CA, United States of America

Mukul Ranjan Prasad