Compression and Context-Aware Benchmark Dataset

The Compression and Context-Aware Benchmark dataset contains 10,000 structured records, which is realistically exhibiting the collection of multi-modal input samples representing diverse data types commonly encountered in cloud-edge systems, intelligent storage, federated environments, and compression pipelines. The dataset is designed to support research in adaptive compression, semantic preprocessing, optimization, and contextual decompression evaluation.It incorporates semantically annotated modalities (text, image, sensor, logs, structured forms), contextual parameters (entropy, fidelity, priority), optimization meta-data (compression strategy), and federated feedback variables (latency, energy, node profile), offering end-to-end simulation capability for modern compression-decompression frameworks.This dataset was created using probabilistic and empirical modeling to reflect realistic patterns observed in actual multi-modal compression and edge-device decompression workloads. Domain-informed entropy scores, strategy selection pathways, and context-sensitive fidelity parameters were modeled using conditional distributions, ensuring the dataset mirrors real-time operational scenarios in smart environments.Parameters such as compression ratios, latency scores, and energy profiles were drawn using bounds informed by benchmarked studies in advanced optimization, edge computing, and federated or split learning deployments.AttributeDescriptionInput_IDUnique identifier for each input sampleModalityType of data: Textual, Image, Sensor, Log, StructuredFormatFile format: CSV, JSON, MP4, XML, JPG, TXTSize_MBRaw size in megabytesEntropy_LevelStatistical entropy class: Low, Medium, HighRetrieval_PriorityPriority for user\/system retrieval: Low, Medium, HighAnnotation_NotesSemantic description of content (e.g., healthcare, telemetry, media)Semantic_Dependency_IndexGraph-based dependency measure (0\u20131)Proximity_Vector_ScoreNormalized semantic proximity metricOriginal_Fidelity_ScorePre-compression semantic fidelity score (0.8\u20131.0)Selected_Compression_StrategyChosen compression approach (e.g., Zstd, HIRAC++, Brotli)Estimated_Compression_RatioPredicted compression ratioActual_Compression_RatioMeasured post-compression ratioFitness_ScoreFitness of strategy using PSO algorithmDecompression_Latency_msLatency measured during decompression (in milliseconds)Energy_Usage_mJEnergy consumed during decompression (in millijoules)Fidelity_DeviationDifference between original and decompressed fidelityNode_TypeExecution node: Edge, Cloud, Mobile