the energy consumption of electrical household appliances is strongly influenced by the behavior of the people who live there. the ignorance of using the electrical appliances led to an increase in electricity bills e...
详细信息
Transformers have gained popularity in the softwareengineering (SE) literature. these deep learning models are usually pre-trained through a self-supervised objective, meant to provide the model with basic knowledge ...
详细信息
ISBN:
(纸本)9781665457019
Transformers have gained popularity in the softwareengineering (SE) literature. these deep learning models are usually pre-trained through a self-supervised objective, meant to provide the model with basic knowledge about a language of interest (e.g., Java). A classic pre-training objective is the masked language model (MLM), in which a percentage of tokens from the input (e.g., a Java method) is masked, withthe model in charge of predicting them. Once pre-trained, the model is then fine-tuned to support the specific downstream task of interest (e.g., code summarization). While there is evidence suggesting the boost in performance provided by pre-training, little is known about the impact of the specific pre-training objective(s) used. Indeed, MLM is just one of the possible pre-training objectives and recent work from the natural language processing field suggest that pre-training objectives tailored for the specific downstream task of interest may substantially boost the model's performance. For example, in the case of code summarization, a tailored pre-training objective could be the identification of an appropriate name for a given method, considering the method name to generate as an extreme summary. In this study, we focus on the impact of pre-training objectives on the performance of transformers when automating code-related tasks. We start with a systematic literature review aimed at identifying the pre-training objectives used in SE. then, we pre-train 32 transformers using both (i) generic pre-training objectives usually adopted in SE;and (ii) pre-training objectives tailored to specific code-related tasks subject of our experimentation, namely bug-fixing, code summarization, and code completion. We also compare the pre-trained models with non pre-trained ones and show the advantage brought by pre-training in different scenarios, in which more or less fine-tuning data are available. Our results show that: (i) pre-training helps in boosting performance on
Collaboration platforms, such as Github and Slack, are a vital instrument in the day-to-day routine of softwareengineering teams. the data stored in these platforms has a significant value for data-driven methods tha...
详细信息
ISBN:
(纸本)9781450393034
Collaboration platforms, such as Github and Slack, are a vital instrument in the day-to-day routine of softwareengineering teams. the data stored in these platforms has a significant value for data-driven methods that assist with decision-making and help improve software quality. However, the distribution of this data across different platforms leads to the fact that combining it is a very time-consuming process. Most existing algorithms for socio-technical assistance, such as recommendation systems, are based only on data directly related to the purpose of the algorithms, often originating from a single system. In this work, we explore the capabilities of a multimodal recommendation system in the context of softwareengineering. Using records of interaction between employees in a software company in messenger channels and repositories, as well as the organizational structure, we build several channel recommendation models for a softwareengineering collaboration platform, and compare them on historical data. In addition, we implement a channel recommendation bot and assess the quality of recommendations from the best models with a user study. We find that the multimodal recommender yields better recommendations than unimodal baselines, allows to mitigate the overfitting problem, and helps to deal with cold start. Our findings suggest that the multimodal approach is promising for other recommendation problems in softwareengineering.
Human-Robot Collaboration (HRC) is showing the potential of widespread application in today's human-centric smart manufacturing, as prescribed by Industry 5.0. To enable safe and efficient collaboration, numerous ...
详细信息
Many enterprises today grapple with supply disruptions in a tumultuous post-COVID-19 environment caused by black swan events, resulting in lost sales and hindering growth. Supply networks, previously optimized for cos...
详细信息
the aviation industry is one of the most dynamic sectors, with safety and quality of service dependent on skilled aircraft maintenance personnel. European legislation, in particular Regulation (EC) No 2042/2003 and su...
详细信息
Digital image forgery (DIF) detection has gained importance in recent years. Because of the availability of lowcost, high-resolution digital cameras, issues related to confirming the authenticity of digital images hav...
详细信息
the proceedings contain 13 papers. the special focus in this conference is on Logic and engineering of Natural Language Semantics. the topics include: Logic Operators and Quantifiers in Type-theory of A...
ISBN:
(纸本)9783031439766
the proceedings contain 13 papers. the special focus in this conference is on Logic and engineering of Natural Language Semantics. the topics include: Logic Operators and Quantifiers in Type-theory of Algorithms;slurs’ Variability, Emotional Dimensions, and Game-theoretic Pragmatics;measurement theory Meets Mereology in Multidimensionality in Resemblance Nominalism;events and Relative Clauses;the Semantic Markedness of the Japanese Negative Preterite: Non-existence of (Positive) Eventualities vs. Existence of Negative Eventualities;granularity in Number and Polarity Effects;contrafactives and Learnability: An Experiment with Propositional Constants;formalizing Argument Structures with Combinatory Categorial Grammar;a Proof-theoretic Analysis of the Meaning of a Formula in a Combination of Intuitionistic and Classical Propositional Logic;constraining Parse Ambiguity with Grammatical Codes;detecting Modality and Evidentiality: Against Purely Temporal-Aspectual Analyses of the German Semi-Modal Drohen.
softwareengineering is a human-centric activity influenced by human aspects such as age, gender, culture, language, attitudes, behaviors, and skills. While many of the studies in the softwareengineering area focused...
详细信息
Pressure for higher productivity and faster delivery is increasingly pervading software organizations. this can lead software engineers to act like chess players playing a gambit-making sacrifices of their technically...
详细信息
ISBN:
(纸本)9781665457019
Pressure for higher productivity and faster delivery is increasingly pervading software organizations. this can lead software engineers to act like chess players playing a gambit-making sacrifices of their technically sound estimates, thus submitting their teams to time pressure. In turn, time pressure can have varied detrimental effects, such as poor product quality and emotional distress, decreasing productivity, which leads to more time pressure and delays: a hard-to-stop vicious cycle. this reveals a need for moving on from the more passive strategy of yielding to pressure to a more active one of defending software estimates. therefore, we propose an approach to support software estimators in acquiring knowledge on how to carry out such defense, by introducing negotiation principles encapsulated in a set of defense lenses, presented through a digital simulation. We evaluated the proposed approach through a controlled experiment withsoftware practitioners from different companies. We collected data on participants' attitudes, subjective norms, perceived behavioral control, and intentions to perform the defense of their estimates in light of the theory of Planned Behavior. We employed a frequentist and a bayesian approach to data analysis. Results show improved scores among experimental group participants after engaging withthe digital simulation and learning about the lenses. they were also more inclined to choose a defense action when facing pressure scenarios than a control group exposed to questions to reflect on the reasons and outcomes of pressure over estimates. Qualitative evidence reveals that practitioners perceived the set of lenses as useful in their current work environments. Collectively, these results show the effectiveness of the proposed approach and its perceived relevance for the industry, despite the low amount of time required to engage with it.
暂无评论