On evaluating recommender systems for API usages

Bruch, Marcel; Schäfer, Thorsten; Mezini, Mira

doi:10.1145/1454247.1454254

Cited by 13 publications

(13 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar limitations apply to qualitative approaches that manually evaluate the result of applying the tool to some selected tasks, see Bruch et al [1] for a discussion.…”

Section: Discussionmentioning

confidence: 96%

See 1 more Smart Citation

On recommending meaningful names in source and UML

Kuhn

2010

Proceedings of the 2nd International Workshop on Recommendation Systems for Software Engineering

View full text Add to dashboard Cite

Meaningful method names are crucial for the readability and maintainability of software. Existing naming conventions focus on syntactic details, leaving programmers with little or no support in choosing meaningful (domain) names. In this paper we propose to build a recommendation system that supports software developers and software architects when naming identifiers in source code as well as when naming elements in UML diagrams. We discuss related work, outline the design of such a recommendation system and discuss possible evaluation strategies.

show abstract

“…Similar limitations apply to qualitative approaches that manually evaluate the result of applying the tool to some selected tasks, see Bruch et al [1] for a discussion.…”

Section: Discussionmentioning

confidence: 96%

“…It has been suggested to evaluate recommendations systems by replaying recorded IDE sessions [11], or by querying the code base for pairs of questions and answers [1]. For example, for each name in the source code we could extract a query with its type and context (but not its name) and then expect the name as correct answer.…”

Section: Discussionmentioning

confidence: 99%

On recommending meaningful names in source and UML

Kuhn

2010

Proceedings of the 2nd International Workshop on Recommendation Systems for Software Engineering

View full text Add to dashboard Cite

show abstract

“…Standard evaluation techniques that refer directly to the performance metrics [22] related to time, storage requirements and computation complexity is used to evaluate the performance of mROSE.…”

Section: Mrose Evaluationmentioning

confidence: 99%

“…Typically most filtering algorithms can be divided into two separate steps [22]: Fist, there is a model building step, usually executed off-line, followed by a second execution step, which is always executed on-line. Preprocessing, representation, calculation and recommendation, prediction generation, which appear in most discussed filtering algorithms, can be as part of off-line step.…”

Section: Computation Complexitymentioning

confidence: 99%

mROSE: To Determine Tool Selection and to Understand Model-Driven Software Evolution

Madhavi¹,

Akepogu²

2012

IJCA

View full text Add to dashboard Cite

Growing interest in the model driven approaches has largely increased the number of tools into the model driven development environment. Previous research has shown that the stakeholders often do not use or know all of the tools available in the model evolution environment that they regularly use. The common solution to this problem is to provide a means to search through passive help documents. However, this approach requires a stakeholder to be able to express their desires in a form understood by search engine. So, choosing the right tool for MoDSE tasks has become difficult because of the diverse nature of numerous tools available. To overcome this limitation, this paper aims to present a prototypical recommendation system, named mROSE, to provide timely and useful recommendations to stakeholders. Two empirical studies were conducted to investigate if mROSE helps or hinders the stakeholders in MoDSE, if so under what conditions. First one was longitudinal user study and the second one was a laboratory user study. Performance of mROSE was also evaluated by using some of the existing metrics. These studies confirmed that mROSE can help stakeholders to choose right tools more efficiently and users liked the idea of having a recommendation system for MoDSE environment, like mROSE. These studies also revealed future directions that would improve the functionality of mROSE.

show abstract

“…Several factors such as the current location in code, e.g., whether the developer is currently working in the control-flow of a framework method [6], or the availability of certain other variables in the current scope affect the relevance of a method.…”

Section: Introductionmentioning

confidence: 99%

Learning from examples to improve code completion systems

Bruch

Monperrus

Mezini

2009

Proceedings of the 7th Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Found

Self Cite

305

222

View full text Add to dashboard Cite

The suggestions made by current IDE's code completion features are based exclusively on static type system of the programming language. As a result, often proposals are made which are irrelevant for a particular working context. Also, these suggestions are ordered alphabetically rather than by their relevance in a particular context. In this paper, we present intelligent code completion systems that learn from existing code repositories. We have implemented three such systems, each using the information contained in repositories in a different way. We perform a large-scale quantitative evaluation of these systems, integrate the best performing one into Eclipse, and evaluate the latter also by a user study. Our experiments give evidence that intelligent code completion systems which learn from examples significantly outperform mainstream code completion systems in terms of the relevance of their suggestions and thus have the potential to enhance developers' productivity.

show abstract

On evaluating recommender systems for API usages

Cited by 13 publications

References 14 publications

On recommending meaningful names in source and UML

On recommending meaningful names in source and UML

mROSE: To Determine Tool Selection and to Understand Model-Driven Software Evolution

Learning from examples to improve code completion systems

Contact Info

Product

Resources

About