Scope & Topics
The workshop will cover techniques, systems, and applications centered on making foundation models scalable, efficient, and adaptable for
large-scale data mining.
Efficient Training & Adaptation
- Low-rank approximation, pruning, distillation, quantization
- Adaptive and dynamic architectural reconfiguration
- Lightweight fine-tuning (LoRA, prefix tuning, adapters)
- Continual, incremental, and federated adaptation strategies
- Transfer learning under limited data or compute
Scalable Foundation Model Design
- Efficient LLMs, vision transformers, multimodal models
- Memory-efficient representation learning
- Sparse, modular, or expert-based architectures
- Foundation models for edge or distributed settings
Foundation Model–Enabled Data Mining
- Integration of FM for classical data mining tasks
- Representation learning for structured & multimodal data
- Large-scale pattern discovery using FM
- Foundation models for graph/time-series mining
- Cross-domain knowledge transfer
Systems & Deployment
- Distributed training & inference optimizations
- Energy-efficient and cost-aware execution
- Hardware–software co-design for scalable FM
- Resource-aware serving, caching, compression systems
Applications
- Healthcare analytics, fraud detection, finance, recommendation
- Manufacturing, IoT and cyber-physical systems
- Smart cities, transportation, environmental monitoring
- Social media mining, language analytics, multimodal KDD applications
Paper Submission
-
Paper Submission Portal (CMT):
https://cmt3.research.microsoft.com/EFMxDM2026
-
To create a CMT account, please visit:
https://cmt3.research.microsoft.com/docs/help/general/account-creation.html
-
For a step-by-step guide on how to submit a paper as an author, see:
https://cmt3.research.microsoft.com/docs/help/author/author-submission-form.html
- Submitted papers must be in English.
- Maximum length: 12 pages (including references and appendices).
- Abstract limited to 200 words.
-
Formatting must follow Springer LNCS/LNAI guidelines (10pt font).
Springer LNCS Formatting Instructions
- Submissions will undergo a double-blind review process.
-
The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.
- All papers must be submitted electronically via CMT in PDF format.
- Supplementary materials (PDF) may be included, but reviewers are not obliged to consider them.
- Submissions violating the policy will be rejected without review.
- Submissions must be original and not under review elsewhere.
- Blind review requirement: remove author names/affiliations from the manuscript and supplementary files.
- Self-citations must be written in the third person.
-
Accepted papers will be made available on the PAKDD 2026 website.
- At least one author of each accepted paper must register and present in person at PAKDD 2026.
Formatting Guidelines
