Stream Processing (SP) engines play a crucial role in realtime analysis within the Big Data landscape, handling infinite data streams to analyze massive, noisy, and heterogeneous information flows.While initially inheriting programming interfaces from Hadoop MapReduce, a recent trend involves adopting declarative languages for expressing analyses.The Event Processing Language (EPL) and its implementation Esper, a mature query language in streaming and event processing, have gained prominence. EPL, with SQL-like syntax, uniquely combines Complex Event Processing (CEP) and streaming analytics.However, it lacks formal semantics.This work addresses this gap by formalizing a core fragment of EPL, focusing on the aspects of Data Definition Language (DDL) and Data Manipulation Language (DML).The formalization resolves semantic ambiguities, identifies potentially harmful constructs, and specifies EPL's data and processing model.This effort addresses a major gap in the formalization of stream processing languages, aligning with recent initiatives from similar domains like graph query languages.

Datalog has grown in popularity with its integration into various commercial and open-source systems. One significant application of Datalog is in stream reasoning scenarios, particularly in the context of Internet of Things and Edge Computing, where it supports the derivation of implicit consequences from incomplete streaming data. The dynamic nature of edge infrastructure, with constantly changing data and deduction rules, poses additional challenges for incremental reasoning. This paper introduces Zodiac, a method for reasoning under rule amendments, and ZodiacEdge, a system implementing this method. Zodiac is based on Datalog stratification and semi-naive evaluation but includes a novel data structure that supports incremental maintenance of deductions for fast rule amendments, even in the presence of negation and aggregation. ZodiacEdge is designed to work with RDF data, accepting Datalog programs with predicates of a maximum arity of two.

Continuous Queries (CQs) are designed to operate over infinite data streams; the paradigm gained prominence with the rise of Stream Processing (SPs). Central to CQs are window operators as they enforce bounded computation by partitioning streams into finite subsets. Although several window operators exist —e.g., slide-by-tuple, session-window, and frames—commercial systems largely adopt a few due to implementation complexity, theoretical opacity, and input-dependent non-determinism. This demonstration shows Play2Win, an interactive playground that empowers users to explore and compare various windowing strategies under a unified system semantics. Our platform offers three key contributions: (I) a real-time environment for experimenting with different window operators; (II) a graph-based representation of the window state that eases direct comparison; and (III) a compositional framework for rapid prototyping of novel windowing mechanisms. The demonstration explore multiple datasets across different scenarios, fostering a deeper understanding of window operators for querying streams.

Property Graphs are a versatile and expressive data model that has gained widespread adoption due to their flexibility in supporting labeled and attributed nodes and edges. They are well-established in research communities and are becoming widespread in companies and organizations across various sectors. They have been boosted by a fervent ISO/IEC standardization activity, leading to dedicated query and schema languages. While the current standards are still evolving, opportunities remain to enrich them with features such as composability. The plethora of existing query languages reflects a rich and diverse ecosystem, which ongoing unification efforts aim to align. This tutorial aims to deepen the understanding of Property Graph standards by showcasing their strengths, highlighting recent unification efforts, clarifying the central role of schema constraints, and exploring the rich landscape of research and industrial opportunities shaping the future of graph data management.

Continuous Queries (CQs) run indefinitely, processing infinite data streams and producing continuous outputs. They commonly use window functions to segment streams intofi nite chunks for computation. Ensuring data integrity in CQs is challenging, involving, for example, streaming joins for binary constraints. Current methods, like dropping or repairing inconsistent data, can harm throughput and increase latency. This paper proposes a novel approach using provenance-based techniques to map violations in input streams to CQ results with minimal overhead. This ensures continuous dataflow and maintains the analytical integrity of CQs. Our study explores the feasibility and efficiency of this method, addressing a significant gap in applying provenance techniques to streaming data. While provenance-based techniques have proven effective for static data, their application in streaming contexts remains unexplored. Our solution addresses this gap, achieving a stable throughput across increasingly demanding memory loads wrt to the baselines, spacing between a 10% increase for medium-sized buffers (i.e., the windows), up to 80% for heavier loads. Moreover, results show the minimal impact of annotation (up to 25%) in the total execution runtime, demonstrating the effectiveness of our graph-based approach.

Abstract Data streams which are now massively and constantly arriving from Internet of Things devices, sensors and social media, require efficient processing, querying and reasoning within a given timeframe. With this in mind, the RDF data model, the cornerstone of the Web of Data, supports a feature-rich stream processing ecosystem that takes into account the temporal dimension associated with events. These timestamped streams support advanced temporal analysis ranging from time-based queries, temporal anomaly detection to temporal reasoning. This survey is the first to provide a comprehensive overview of the field of RDF stream processing, focusing on (query) languages, systems, and benchmarks. For each of these areas, we present salient dimensions, propose a taxonomy of existing work, detail the concepts at the core of each approach and describe their main technical aspects and implementation. We hope that the survey will help readers understand this scientifically rich field and identify the most relevant method for various usage scenarios.

International audience

Stream processing is designed to query unbounded and timely-ordered data flows in real-time while guaranteeing low latency and high throughput.

Data streams are becoming omnipresent on the Web. The Stream Reasoning (SR) paradigm, which combines Stream Processing with Semantic Web techniques, has been successful in processing these data streams. The progress in SR research has led to several applications in domains such as the Internet of Things , social media analysis, Smart Cities, and many others. Each of these applications produces and consumes data streams, however, there are no fixed guidelines on how to manage data streams on the Web, as there are for their static counterparts. More specifically, there is no fixed life cycle for Streaming Linked Data (SLD) yet. Tommasini et al. (2020) introduced an initial proposal for a SLD life cycle , however, it has not been verified if the proposed life cycle captures existing applications and no guidelines were given for each step. In this paper, we survey existing SR applications and identify if the life cycle proposed by Tommasini et al. fully captures the surveyed applications. Based on our analysis, we found that some of the steps needed reordering or being split up. This paper proposes an update of the life cycle and surveys the existing literature for each life cycle step while proposing a number of guidelines and best practices. Compared to the initial proposal by Tommasini et al., we drill down into the details of the processing step which was previously neglected. The updated life cycle and guidelines serves as a blueprint for future SR applications. A life cycle for SLD that allows to efficiently manage data streams on the web, brings us a step closer to the realization of the SR vision.

Leveraging Big Data (BD) processing frameworks to process large-scale Resource Description Framework (RDF) datasets holds a great interest in optimizing query performance. Modern BD services are complicated data systems, where tuning the configurations notably affects the performance. Benchmarking different frameworks and configurations provides the community with best practices towards selecting the most suitable configurations. However, most of these benchmarking efforts are classified as descriptive or diagnostic analytics. Moreover, there is no standardization for comparing and contrasting these benchmarks based on quantitative ranking techniques. This paper aims to fill this timely research gap by proposing ranking criteria (called Bench-ranking) that provide prescriptive analytics via ranking functions. In particular, Bench-ranking starts by describing the current state-of-the-art single-dimensional ranking limitations. Next, we discuss the recent benchmarking requirements for sophisticated approaches over multi-dimensional ranking. Finally, we discuss the ranking criteria goodness by reviewing its conformance and coherence metrics. We validate Bench-ranking by conducting an empirical study using large RDF datasets under a relational BD engine, i.e., Apache Spark-SQL. The proposed ranking techniques provide the practitioners with clear insights to make an informed decision, especially with experimental trade-offs for such complex solution space.

Streaming Analytics (SA) and Complex Event Recognition (CER) are of paramount importance in searching for an ultimate Big Data solution that can simultaneously address Data Velocity, Variety, and Volume. Indeed, the growing popularity of streaming data has pushed the boundaries of existing data systems, fostering the rise of Stream Processing Engines (SPE). However, Data Velocity never appears isolated. Streams are huge, heterogeneous, and noisy as they come from multiple sources. Horizontally-scalable SPEs like Flink and KSQL-DB allow continuous stream analytics using SQL-like languages. On the other hand, CER engines like OracleCEP and DroolFusion use regular languages for (parallel) pattern detection over heterogeneous streams. This paper takes a first step towards a unifying solution. To this extent, we present KELPr, an in-memory distributed CER engine designed extending the Dual Streaming Model and implemented on top of Kafka Streams.