News
Book Chapter at ACM 2026
Data Series Similarity Search Methods. In Data Series Management and Analytics: From Time Series to High-dimensional Vectors, ACM, 2026
Paper Presented at SIGMOD/PODS 2026
DARTH: Declarative Recall Through Early Termination for Approximate Nearest Neighbor Search
Paper Presented at CVPR 2026
The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment
Paper Accepted at ICML 2026
ParisKV: Fast and Drift-Robust KV-Cache Retrieval for Long- Context LLMs
Journal Accepted at PACMMOD 2026
TaCo: Data-adaptive and Query-aware Subspace Collision for High-dimensional Approximate Nearest Neighbor Search
Journal Accepted at IEEE TKDE 2026
PDET-LSH: Scalable In-Memory Indexing for High-Dimensional Approximate Nearest Neighbor Search with Quality Guarantees
School on the Practice and Theory of Distributed Computing 2026
SPTDC 2026 has officially concluded after four days of deep exploration into the theory and practice of modern distributed computing.
Poster Presented at SPTDC 2026
ATROPOS: Benchmarking Concurrent Persistent Algorithms under Failures
Poster Presented at SPTDC 2026
PPMR: Persistent Polymorphic Memory Resources
Poster Presented at SPTDC 2026
PFRESH: A persistent Lock Free Data Series Index
Poster Presented at SPTDC 2026
Concurrent Double-Ended Priority Queues
Poster Presented at SPTDC 2026
The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment
Presentation at Data Science and Applications Course, UoC
Presentation of HAR.S.H. during the Data Science and Applications Course at the Department of Computer Science at the University of Crete.
Poster Accepted at CVPR 2026
The More, the Merrier: Contrastive Fusion for Higher-Order Multimodal Alignment
Invitation Talk at SPTDC 2026
Vector Similarity Search: Past, Present, and Future
Invited Talk at Nanyang Technological University
Vector Similarity Search: Past, Present, and Future
School on the Practice and Theory of Distributed Computing 2026
Excited to announce SPTDC 2026! Organized by HAR.S.H., this year’s school brings together experts from academia and industry to explore the future of Distributed Computing.
Paper Presented at ICLR 2026
CAuKer: Classification Time Series Foundation Models Can Be Pretrained on Synthetic Data
Workshop at TSALM 2026
Investigating Simple Target-Covariate Relationships for Chronos-2 and TabPFN-TS
Workshop at TSALM 2026
Time Series Foundation models Improve LLM Decisions: A Case Study in Stock Trading
Tutorial at EDBT 2026
The Quest for Faster ANN Vector Search
Invited talk at MDD Summer School 2026
High-Dimensional Vector Indexing and Similarity Search
Creation of Exact Similarity Search Library
Introduction of DaiSy: The Library for Fast and Exact, Data Series and Vector Similarity Search, designed in python
Invited talk at National Metsovio Polytechnic
Vector Similarity Search: Past, Present, and Future
Presentation of HAR.S.H to High Schools Visits
Presenting our research work to high school students during their visits to the Department of Computer Science, at the Uniersity of Crete.
Paper Presented at IEEE 2025
Interpretable Multivariate Anomaly Detector Selection for Automatic Marine Data Quality Control
Paper Presented at PPoPP 2026
Concurrent Balanced Augmented Trees
Paper Presented at PPoPP 2026
Sharded Elimination and Combining for Highly-Efficient Concurrent Stacks
Paper Accepted at PPoPP 2026
Concurrent Balanced Augmented Trees
Paper Accepted at PPoPP 2026
Sharded Elimination and Combining for Highly-Efficient Concurrent Stacks
Best System Paper Award at OPODIS 2025
Recoverable Lock-Free Locks
Paper Presented at OPODIS 2025
Recoverable Lock-Free Locks
Journal at IEEE Transactions on Knowledge and Data Engineering (TKDE), Volume 38, Issue 6
MS-Index: Fast Top-k Subsequence Search for Multivariate Time Series under Euclidean Distance
Invited Presentation at MIT
Concurrent Trees Supporting Complex Queries
Innodays 2025 Exhibition
Presenting HAR.S.H. during Innovation Days at companies and participants
Invited presentation at MIT
Vector Similarity Search: Past, Present, and Future
Invited presentation at Khoury College of Computer Science
Concurrent Trees Supporting Complex Queries
Invited presentation at Managing Specialized and Heterogeneous Architectures Workshop
Parallel Data Series Indexing and Similarity Search on Modern Hardware
Invited presentation at Harvard
Vector Similarity Search: Past, Present, and Future
Invited presentation at Amazon
Vector Similarity Search: Past, Present, and Future
Invited presentation at Google
Vector Similarity Search: Past, Present, and Future
Invited presentation at Columbia University
Vector Similarity Search: Past, Present, and Future
Invited presentation at AT&T Labs
Vector Similarity Search: Past, Present, and Future
Invited Talk at Simons Institute for the Theory of Computing
Combining and Aggregating for Fast Synchronization
Invited presentation at the Foundation Of Research and Technology - Heraklion
Vector Similarity Search: Past, Present, and Future
Participation at Researcher's Night
The HAR.S.H. team participates in Researcher's Night and presents on similarity search and retrieval, as well as zero-shot classification.
Participation at Researcher's Night Chatlab Section
HAR.S.H. team participated in the Researcher's Night Chatlab and talked about similarity search and retrieval, and zero-shot classification to schools and students
Poster Presented at CSD Reunion
Computer Science Department - University of Crete, Classes of '85/'95/'05 reunion.
Plenary Keynote at Guide-AI Workshop, in conjunction with the International Conference on Very Large Data Bases (VLDB)
Invited presentation: Scalable Vector Analytics: A Story of Twists and Turns
Poster Presented at CSD Reunion
Computer Science Department - University of Crete, Classes of '85/'95/'05 reunion.
Poster Presented at VLDB 2025
The LAW theorem: Local Reads and Linearizable Asynchronous Replication
Paper Presented at VLDB 2025
The LAW theorem: Local Reads and Linearizable Asynchronous Replication
Tutorial at ACM Conference on Knowledge Discovery & Data Mining 2025
Advances in Time-Series Anomaly Detection: Algorithms, Benchmarks, and Evaluation Measures
Tutorial at International Joint Conference on Artificial Intelligence 2025
"Advances in Time-Series Anomaly Detection: Algorithms, Benchmarks, and Evaluation Measures"
Paper Accepted at 39th International Symposium on Distributed Computing
Brief Announcement: Concurrent Double-Ended Priority Queues
Paper Accepted at 39th International Symposium on Distributed Computing
PIPQ: A Strict Insert-Optimized Concurrent Priority Queue
Paper Presented at ACM 2025
NILMFormer: Non-Intrusive Load Monitoring that Accounts for Non-Stationarity
Paper Presented at ACM 2025
Evaluating and Generating Query Workloads for High-Dimensional Vector Similarity Search
Workshop at Highlights of Parallel Computing 2025
Lock-Free Augmented Trees (Abstract)
Invited presentation at Technical University of Berlin
Invited presentation at Snowflake
Poster Presented at ACM-W Celebration of Women in Computing: 7th Summit on Gender Equality in Computing
Concurrent Double-Ended Priority Queues
Poster Presented at ACM-W Celebration of Women in Computing: 7th Summit on Gender Equality in Computing
Concurrent Binary Search Trees Supporting Split and Join
Journal Accepted at Distributed Computing Volume 38
When is Recoverable Consensus harder than Consensus?
University of Crete Press Release at the Beginning of HARSH
HARSH Poster – February 2025
HARSH FactSheet
Similarity Search in Large-Scale Time Series Collections with Hardware Awareness
HARSH: Hardware-Aware extReme-scale Similarity search
- Project Code: ΥΠ3ΤΑ-0560901
- Project Start Date: April 15, 2025
- Project End Date: June 30, 2026
- Funding Body: Ministry of Education, Religious Affairs and Sports, Greece 2.0, National Recovery and Resilience Plan
- Host Institution: University of Crete (UoC), Department of Computer Science
