{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,9]],"date-time":"2026-07-09T08:13:26Z","timestamp":1783584806074,"version":"3.55.0"},"publisher-location":"Cham","reference-count":17,"publisher":"Springer Nature Switzerland","isbn-type":[{"value":"9783032042873","type":"print"},{"value":"9783032042880","type":"electronic"}],"license":[{"start":{"date-parts":[[2025,8,22]],"date-time":"2025-08-22T00:00:00Z","timestamp":1755820800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"},{"start":{"date-parts":[[2025,8,22]],"date-time":"2025-08-22T00:00:00Z","timestamp":1755820800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026]]},"DOI":"10.1007\/978-3-032-04288-0_9","type":"book-chapter","created":{"date-parts":[[2025,8,21]],"date-time":"2025-08-21T08:23:19Z","timestamp":1755764599000},"page":"143-157","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking AI-Facilitated UI Test-Script Generation: A Reproducible Evaluation Framework"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-2868-5633","authenticated-orcid":false,"given":"Alexander","family":"Poth","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Olsi","family":"Rrjolli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Huiyu","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,8,22]]},"reference":[{"key":"9_CR1","doi-asserted-by":"crossref","unstructured":"Leotta, M., Yousaf, H. Z., Ricca, F., Garcia, B.: AI-generated test scripts for web e2e testing with ChatGPT and copilot: a preliminary study. In: Proceedings of the 28th International Conference on Evaluation and Assessment in Software Engineering, pp. 339\u2013344 (2024)","DOI":"10.1145\/3661167.3661192"},{"key":"9_CR2","doi-asserted-by":"crossref","unstructured":"McIntosh, T.R., Susnjak, T., Arachchilage, N., Liu, T., Xu, D., Halgamuge, M.N.: Inadequacies of large language model benchmarks in the era of generative artificial intelligence. IEEE Trans. Artif. Intell. (2025)","DOI":"10.1109\/TAI.2025.3569516"},{"key":"9_CR3","unstructured":"Chen, M., et al.: Evaluating large language models trained on code (2021). arXiv preprint arXiv:2107.03374 (2021)"},{"key":"9_CR4","unstructured":"Hendrycks, D., et al.: Measuring coding challenge competence with apps. arXiv preprint arXiv:2105.09938 (2021)"},{"issue":"6624","key":"9_CR5","doi-asserted-by":"publisher","first-page":"1092","DOI":"10.1126\/science.abq1158","volume":"378","author":"Y Li","year":"2022","unstructured":"Li, Y., et al.: Competition-level code generation with alphacode. Science 378(6624), 1092\u20131097 (2022)","journal-title":"Science"},{"key":"9_CR6","doi-asserted-by":"crossref","unstructured":"Poth, A., Rrjolli, O., Arcuri Technology adoption performance evaluation applied to testing industrial REST APIs. Autom. Softw. Eng. 32(1), 5","DOI":"10.1007\/s10515-024-00477-2"},{"key":"9_CR7","unstructured":"Deng, C., Zhao, Y., Cohan, A.: Investigating data contamination in modern benchmarks for large language models (2023). arXiv preprint arXiv:2311.09783"},{"key":"9_CR8","unstructured":"Sun, Y., Wang, H., Li, D., Wang, G., Zhang, H.: The Emperor\u2019s New Clothes in Benchmarking? A Rigorous Examination of Mitigation Strategies for LLM Benchmark Data Contamination (2025). arXiv preprint arXiv:2503.16402"},{"key":"9_CR9","unstructured":"Xu, R., Wang, Z., Fan, R.Z., Liu, P.: Benchmarking benchmark leakage in large language models (2024). arXiv preprint arXiv:2404.18824"},{"key":"9_CR10","unstructured":"Yang, S., Chiang, W.L., Zheng, L., Gonzalez, J.E., Stoica, I.: Rethinking benchmark and contamination for language models with rephrased samples (2023). arXiv preprint arXiv:2311.04850"},{"key":"9_CR11","doi-asserted-by":"crossref","unstructured":"Romano, A., Song, Z., Grandhi, S., Yang, W., Wang, W.: An empirical analysis of UI-based flaky tests. In: 2021 IEEE\/ACM 43rd International Conference on Software Engineering (ICSE), pp. 1585\u20131597. IEEE (2021)","DOI":"10.1109\/ICSE43902.2021.00141"},{"key":"9_CR12","unstructured":"BrowserStack, 15 Challenges in Automation Testing and Their Solutions. https:\/\/www.browserstack.com\/guide\/challenges-in-automated-testing"},{"key":"9_CR13","doi-asserted-by":"publisher","unstructured":"Poth, A., Prang, T., Rrjolli, O.: Accessibility testing and sustainability\u2014considerations, concepts, and lasting implementations. In: Yilmaz, M., Clarke, P., Riel, A., Messnarz, R., Greiner, C., Peisl, T. (eds.) Systems, Software and Services Process Improvement. EuroSPI 2024. Communications in Computer and Information Science, vol. 2180, pp. 76\u221290. Springer, Cham (2024). https:\/\/doi.org\/10.1007\/978-3-031-71142-8_6","DOI":"10.1007\/978-3-031-71142-8_6"},{"key":"9_CR14","doi-asserted-by":"crossref","unstructured":"Poth, A., Rrjolli, O., Riel, A.: Integration-and system-testing aligned with cloud-native approaches for DevOps. In: 2022 IEEE 22nd International Conference on Software Quality, Reliability, and Security Companion (QRS-C), pp. 201\u2013208. IEEE (2022)","DOI":"10.1109\/QRS-C57518.2022.00038"},{"key":"9_CR15","unstructured":"Levenshtein, V.I.: Binary codes capable of correcting deletions, insertions, and reversals. Soviet Phys. doklady 10(8), 707\u2013710 (1966)"},{"key":"9_CR16","doi-asserted-by":"publisher","unstructured":"Poth, A., Heimann, C.: How to innovate software quality assurance and testing in large enterprises?. In: European Conference on Software Process Improvement, pp. 437\u2013442. Cham: Springer International Publishing (2018). https:\/\/doi.org\/10.1007\/978-3-319-97925-0_37","DOI":"10.1007\/978-3-319-97925-0_37"},{"key":"9_CR17","doi-asserted-by":"publisher","unstructured":"Poth, A., Heimann, C.: Frugal innovation approaches combined with an agile organization to establish an innovation value stream. In: Yilmaz, M., Clarke, P., Riel, A., Messnarz, R. (eds.) Systems, Software and Services Process Improvement. EuroSPI 2023. Communications in Computer and Information Science, vol. 1891, pp. 260-274. Springer, Cham (2023). https:\/\/doi.org\/10.1007\/978-3-031-42310-9_19","DOI":"10.1007\/978-3-031-42310-9_19"}],"container-title":["Communications in Computer and Information Science","Systems, Software and Services Process Improvement"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/978-3-032-04288-0_9","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,9]],"date-time":"2026-07-09T07:17:28Z","timestamp":1783581448000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/978-3-032-04288-0_9"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,22]]},"ISBN":["9783032042873","9783032042880"],"references-count":17,"URL":"https:\/\/doi.org\/10.1007\/978-3-032-04288-0_9","relation":{},"ISSN":["1865-0929","1865-0937"],"issn-type":[{"value":"1865-0929","type":"print"},{"value":"1865-0937","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,22]]},"assertion":[{"value":"22 August 2025","order":1,"name":"first_online","label":"First Online","group":{"name":"ChapterHistory","label":"Chapter History"}},{"value":"EuroSPI","order":1,"name":"conference_acronym","label":"Conference Acronym","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"European Conference on Software Process Improvement","order":2,"name":"conference_name","label":"Conference Name","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"Riga","order":3,"name":"conference_city","label":"Conference City","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"Latvia","order":4,"name":"conference_country","label":"Conference Country","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"2025","order":5,"name":"conference_year","label":"Conference Year","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"17 September 2025","order":7,"name":"conference_start_date","label":"Conference Start Date","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"19 September 2025","order":8,"name":"conference_end_date","label":"Conference End Date","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"32","order":9,"name":"conference_number","label":"Conference Number","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"eurospi2025","order":10,"name":"conference_id","label":"Conference ID","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"https:\/\/conference.eurospi.net","order":11,"name":"conference_url","label":"Conference URL","group":{"name":"ConferenceInfo","label":"Conference Information"}}]}}