Fine-SE: Integrating Semantic Features and Expert Features for Software Effort Estimation
Reliable effort estimation is of paramount importance to software planning and management, especially in industry that requires effective and on-time delivery. \textcolor{blue}{Various estimation approaches have been proposed (\eg planning poker and analogy), but they may be subjective and non-automatic, which are not applicable to other projects. \textcolor{violet}{(C${\textbf{A.19}}$, C${\textbf{B.5}}$)}} In recent years, deep learning approaches for effort estimation that rely on learning expert features or semantic features respectively have been extensively studied and have been found to be promising. Semantic features and expert features describe software tasks from different perspectives, however, in the literature, the best combination of these two features has not been explored to enhance effort estimation. Additionally, there are a few studies that discussed which expert features are useful for estimating effort in the industry. To this end, we investigate the potential 13 expert features that can be used to estimate effort by interviewing 26 enterprise employees. After that, we propose a novel model, called Fine-SE, that leverages semantic features and expert features for effort estimation. To validate our model, a series of evaluations are conducted on more than 30,000 software tasks from 17 industrial projects of a global ICT enterprise and four open-source software (OSS) projects. The evaluation results indicate that Fine-SE provides higher performance than the baselines on evaluation measures (\ie mean absolute error, mean magnitude of relative error, and performance indicator), particularly in industrial projects with large amounts of software tasks, which is a significant improvement in effort estimation. In comparison with expert estimation, Fine-SE improves the performance of evaluation measures by 32.0%-45.2% in within-project estimation. In comparison with the state-of-the-art models, Deep-SE and GPT2SP, it achieved an improvement of 8.9%-91.4% in industrial projects. The empirical results demonstrate the value of integrating expert features with semantic features in effort estimation.
Fri 19 AprDisplayed time zone: Lisbon change
14:00 - 15:30 | Analytics 5Research Track / Journal-first Papers at Amália Rodrigues Chair(s): Sridhar Chimalakonda Indian Institute of Technology, Tirupati | ||
14:00 15mTalk | An Exploratory Investigation of Log Anomalies in Unmanned Aerial Vehicles Research Track Dinghua Wang , Shuqing Li The Chinese University of Hong Kong, Guanping Xiao Nanjing University of Aeronautics and Astronautics, Yepang Liu Southern University of Science and Technology, Yulei Sui UNSW, Pinjia He Chinese University of Hong Kong, Shenzhen, Michael Lyu The Chinese University of Hong Kong | ||
14:15 15mTalk | ModuleGuard: Understanding and Detecting Module Conflicts in Python Ecosystem Research Track Ruofan Zhu Zhejiang University, Xingyu Wang Zhejiang University, Chengwei Liu Nanyang Technological University, Zhengzi Xu Nanyang Technological University, Wenbo Shen Zhejiang University, China, Rui Chang Zhejiang University, Yang Liu Nanyang Technological University | ||
14:30 15mTalk | Empirical Analysis of Vulnerabilities Life Cycle in Golang Ecosystem Research Track Jinchang Hu , Lyuye Zhang Nanyang Technological University, Chengwei Liu Nanyang Technological University, Sen Yang Academy of Military Science, Song Huang Army Engineering University of PLA, Yang Liu Nanyang Technological University | ||
14:45 15mTalk | Fine-SE: Integrating Semantic Features and Expert Features for Software Effort Estimation Research Track Yue Li Nanjing University, Zhong Ren State Key Laboratory of Novel Software Technology, Software Institute, Nanjing University Nanjing, Jiangsu, China, Zhiqi Wang State Key Laboratory of Novel Software Technology, Software Institute, Nanjing University Nanjing, Jiangsu, China, Lanxin Yang Nanjing University, Liming Dong Nanjing University, He Zhang Nanjing University | ||
15:00 7mTalk | Concretization of Abstract Traffic Scene Specifications Using Metaheuristic Search Journal-first Papers Aren Babikian McGill University, Oszkár Semeráth Budapest University of Technology and Economics, Daniel Varro Linköping University / McGill University | ||
15:07 7mTalk | Technical leverage analysis in the Python ecosystem Journal-first Papers Ranindya Paramitha University of Trento, Fabio Massacci University of Trento; Vrije Universiteit Amsterdam | ||
15:14 7mTalk | Automated Mapping of Adaptive App GUIs from Phones to TVs Journal-first Papers Han Hu Faculty of Information Technology, Monash University, ruiqi dong Swinburne University of Technology, John Grundy Monash University, Thai Minh Nguyen Monash University, huaxiao liu Jilin University, Chunyang Chen Technical University of Munich (TUM) Link to publication DOI Pre-print | ||
15:21 7mTalk | Assessing the Early Bird Heuristic (for Predicting Project Quality) Journal-first Papers Link to publication DOI Pre-print |