RESimEnv

Dataset Description This dataset is a synthetic requirements engineering dataset designed to support research in software requirements analysis, requirements discovery, conflict detection, and stakeholder-aware system design. Each record represents a software system scenario from a specific application domain and contains structured information commonly produced during requirements elicitation and analysis. The dataset covers a diverse range of domains including e-commerce marketplaces, healthcare systems, educational platforms, financial applications, IoT systems, government service portals, logistics platforms, and collaboration tools. For each scenario, the dataset provides a high-level system description, a complete set of ground-truth functional requirements, a subset of visible requirements available to the analyst, and hidden requirements that represent omitted or undiscovered system needs. In addition to functional requirements, the dataset includes non-functional requirements related to performance, security, privacy, compliance, reliability, scalability, and usability. Stakeholder information is provided for each scenario, including stakeholder roles, interests, priorities, and potential conflicts. Domain entities representing key business concepts are also included to support modeling and entity extraction tasks. A distinctive feature of the dataset is the inclusion of requirement conflicts. Each scenario contains at least one explicitly defined conflict between requirements, along with a conflict type such as temporal, privacy, compliance, business rule, access control, or stakeholder conflict. These annotations enable research on automated conflict detection, conflict classification, and requirements negotiation. The dataset further includes a difficulty label (easy, medium, or hard) that reflects the complexity of the scenario based on factors such as domain criticality, stakeholder interactions, regulatory constraints, and requirement dependencies. Overall, the dataset is intended to support a wide range of Requirements Engineering and Artificial Intelligence research tasks, including hidden requirement prediction, requirement recommendation, requirements completion, conflict detection, stakeholder analysis, domain classification, requirement quality assessment, use case generation, user story generation, and multi-task learning for automated requirements engineering. Key Components Component Description Scenario ID Unique identifier for each scenario Domain Application domain of the software system Rough Idea High-level description of the system Ground Truth Requirements Complete set of functional requirements Visible Requirements Requirements initially available to analysts Hidden Requirements Missing requirements to be inferred or discovered Non-Functional Requirements Performance, security, compliance, and quality requirements Stakeholders Actors, roles, interests, and conflicts Domain Entities Core business entities within the system Requirement Conflicts Explicitly annotated conflicting requirements Difficulty Level Complexity classification of the scenario Potential Research Applications Hidden Requirement Discovery Requirement Recommendation Systems Requirement Conflict Detection and Resolution Functional and Non-Functional Requirement Classification Stakeholder-Aware Requirements Analysis Domain Entity Extraction Automatic Use Case Generation User Story Generation Software Project Complexity Prediction Multi-Task Learning for Requirements Engineering This dataset provides a realistic simulation of early-stage requirements engineering artifacts and can serve as a benchmark for evaluating intelligent systems that assist analysts in eliciting, validating, and managing software requirements.