Streams of data are now massively, constantly arrive from Internet of Things Devices (IoT), sensors and social media, and require rapid processing, querying and integration with background knowledge in order to support further data analysis.
So far, the current state-of-the-art in RDF Stream processing has provided either centralized engines that cannot deal with massive RDF data streams or distributed engines that offer limited reasoning capabilities.
Summarization techniques on the other hand have already proved their value for indexing data, query answering, reasoning, source selection, graph visualization, and schema discovery. However, to the best of our knowledge they have not yet been exploited for stream data, which remains a completely unexplored area.
High-quality Summaries
Efficient Exact QA
Approximate QA
Incremental Algorithms
Stream Processing
Summarization techniques have already proved their value for indexing data, query answering, estimating the size of query results, source selection, graph visualization, and schema discovery . However, to the best of our knowledge they have not yet been exploited for stream data, which remains a completely unexplored area.
Having established effective and efficient algorithms for summarizing stream data for background knowledge and for the streaming window dataset the natural next step is to devise algorithms for answering queries that require the combination of both dynamic stream data with historic static or staged data. This will result in a unified platform that will enable reasoning on stream data, enabling further analysis tasks
Due to the rapidly changing information and query needs, it might be the case that the summary itself might not be adequate to return exact answers. In this case, we intend to offer both approximate answers that will be able to be delivered instantly over the summaries and then to investigate exploration operations that will allow us to eventually expand current summaries for retrieving exact answers
Due to unpredictable input data streams, we expect that summaries should be dynamically updated. We foresee that, for the stream data in the current sliding window, each time a new summary will be generated from scratch with minimal footprint and computational impact. However, as the historic stream data are staged and combined with the background knowledge the summaries for the historic data should be incrementally updated rather that recomputed from scratch. As such, the third objective of this proposal is to devise incremental updating algorithms for summaries over the stream data
Develop summarization techniques for stream data capitalizing the world-leading expertise of the PI and his team on building semantic summaries, investigating approaches of both supernode creation and node selection.
Georgia Troullinou, Kostas Stefanidis, Dimitris Plexousakis and Haridimos Kondylakis, DIAERESIS: Knowledge Graph Partitioning for Efficient Query Answering, TheWebConf 2024.
Georgia Troullinou, Giannis Agathangelos, Haridimos Kondylakis, Kostas Stefanidis, Dimitris Plexousakis, DIAERESIS: RDF Data Partitioning and Query Processing on SPARK, Semantic Web Journal, 2024.
Giannis Vassiliou, Nikos Papadakis and Haridimos Kondylakis, SummaryGPT: Leveraging ChatGPT for Summarizing Knowledge Graphs, ESWC 2023.
Angela Bonifati, Stefania Dumbrava, Haridimos Kondylakis, Georgia Troullinou, Giannis Vassiliou, PING: Progressive Querying on RDF Graphs, ISWC 2025.
Giannis Vassiliou, Nikolaos Papadakis and Haridimos Kondylakis, iSummary: Demonstrating Workload-based, Personalized Summaries for Knowledge Graphs, ISWC 2023.
Giannis Vassiliou, Fanouris Alevizakis, Nikolaos Papadakis and Haridimos Kondylakis, iSummary: Workload-based, Personalized summaries for Knowledge Graphs, ESWC 2023
Kenza Kellou-Menouer, Nikolaos Kardoulakis, Georgia Troullinou, Zoubida Kedad, Dimitris Plexousakis, and Haridimos Kondylakis, Tutorial on Semantic Schema Discovery: principles, methods and future research directions, ISWC 2022
Alexander Miraka, Nikolaos Kardoulakis, Kenza Kellou-Menouer, Georgia Troullinou, Zoubida Kedad, Dimitris Plexousakis, and Haridimos Kondylakis, HInT: Hybrid and Incremental Type Discovery for Semantic Graphs, ISWC 2022
Kenza Kellou-Menouer, Nikolaos Kardoulakis, Georgia Troullinou, Zoubida Kedad, Dimitris Plexousakis, Haridimos Kondylakis, A survey on Semantic Schema Discovery, PVLDB Posters, 2022
Georgia Eirini Trouli, Alexandros Pappas, Georgia Troullinou, Lefteris Koumakis, Nikos Papadakis, Haridimos Kondylakis, SumMER: Structural Summarization for RDF/S KGs, Algorithms, 16(1), 18, 2022
Georgia Troullinou, Haridimos Kondylakis, Matteo Lissandrini, Davide Mottin: SOFOS: Demonstrating the Challenges of Materialized View Selection on Knowledge Graphs. ACM SIGMOD (2021)
Georgia Eirini Trouli, Georgia Troullinou, Lefteris Koumakis, Nikolaos Papadakis and Haridimos Kondylakis, SumMER: Summarizing RDF/S KBs using Machine LEaRning, ISWC 2021
Giannis Vassiliou, Georgia Troullinou, Nikos Papadakis, Kostas Stefanidis, Evangelia Pitoura, Haridimos Kondylakis, Coverage-Based Summaries for RDF KBs, ESWC 2021
Nikolaos Kardoulakis, Kenza Kellou-Menouer, Georgia Troullinou, Zoubida Kedad, Dimitris Plexousakis, Haridimos Kondylakis: HInT: Hybrid and Incremental Type Discovery for Large RDF Data Sources. SSDBM 2021: 97-108
Giannis Vassiliou, Georgia Troullinou, Nikolaos Papadakis, Haridimos Kondylakis, WBSum: Workload-based Summaries for RDF/S KBs, SSDBM 2021
Kenza Kellou-Menouer, Nikolaos Kardoulakis, Georgia Troullinou, Zoubida Kedad, Dimitris Plexousakis, Haridimos Kondylakis, A survey on Semantic Schema Discovery, VLDB Journal (2021)
| Deliverable Number | Deliverable Name |
|---|---|
| D4.1 | System Design and Architecture |
| D1.1 | First periodic management report |
| D3.1 | Transparent summaries for historic stream data |
| D3.2 | Incremental update of summaries |
| D1.2 | Second periodic management report |
| D2.1 | Processing recent window stream data |
| D3.3 | Query answering using summaries |
| D5.1 | Query answering using summaries |
| D6.1 | Report on dissemination and exploitation activities of the project |
| D4.2 | Combining historic with dynamic stream data |
| D1.3 | Final report |
| D5.2 | Final Evaluation Report |
| D6.2 | Final report on dissemination and exploitation activities of the project |