To improve the systematicity and traceability of HSE system evaluation indicators, this study addresses limitations in traditional indicator selection, including strong dependence on expert experience, high subjectivity, overlapping indicator boundaries and unclear source evidence. Based on a first-level indicator framework determined by enterprise experts, HSE standards, system documents and related literature were used as the corpus. A domain-specific entity annotation scheme was designed, covering management measures, responsible actors, evidence information, risk objects, resource support, tools and technologies, and improvement actions. A BIO-annotated dataset was then constructed, and the BERT-BiLSTM-CRF model was employed for domain entity recognition. On this basis, second-level evaluation indicators were generated and screened through knowledge fusion, graph-path retrieval and topic consolidation. The results show that the BERT-BiLSTM-CRF model achieved a precision of 91.2%, a recall of 89.8% and an F1-score of 90.5%. Based on entity recognition and topic consolidation, an HSE system evaluation indicator system comprising 9 first-level indicators and 36 second-level indicators was developed, providing a basis for subsequent indicator weighting and comprehensive evaluation modeling.
Kexin Sun, Hui-Ling Na, Jianwei Zhang et al.· International journal of sof...· 0 citations
Finding people from a natural-language description (``ML engineers transitioning to research roles in biotech'') is increasingly delegated to LLM agents and framed as information retrieval. We argue that it is fundamentally a requirements engineering task: such a request is an under-determined requirement with implicit constraints, many valid answers, and no acceptance criterion, so useful answers require eliciting, validating, and verifying the requirement before search can matter. We present \sys{}, to our knowledge the first interactive, requirements-driven candidate-sourcing agent (it elicits, validates, retrieves, and verifies a vague people-request into a justified slate through bounded elicitation, workflow templates, a two-stage commit protocol, and bidirectional termination guards) and \bench{}, a benchmark that runs the requirements lifecycle (criteria-anchored validation, multi-model evidence-grounded oracle construction, and cost-aware verification). Across $21$ systems and all $691$ requirements, \sys{} dominates breadth ($100%$ coverage at $2.5\times$ the yield) and is \emph{near-orthogonal} to the field, with $90%$ of the people it returns are surfaced by \emph{none} of $20$ strong LLM-plus-web baselines combined. Beyond breadth, an evidence-grounded judging of every system shows \sys{} \emph{recalls} the most relevant real people: $0.241$ of the union pool, $1.9\times$ the next system, with a bootstrap $95%$ interval disjoint from every baseline. \sys{} is thus the strongest \emph{sourcing} engine (the deepest real, reachable candidate pool), while precision-ranking LLMs serve as~complementary verifiers.
Yuanpeng He, Fan Li, Xiangyu Ru et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
