Special Sessions

ISMIS is an established and prestigious conference for exchanging the latest research results in building intelligent systems. Held twice every three years, the conference provides a medium for exchanging scientific research and technological achievements accomplished by the international community.

The scope of ISMIS is intended to represent a wide range of topics on applying Artificial Intelligence techniques to areas as diverse as decision support, automated deduction, reasoning, knowledge based systems, machine learning, computer vision, robotics, planning, databases, information retrieval, etc. The focus is on research in intelligent systems. The conference addresses issues involving solutions to problems that are complex to be solved through conventional approaches and that require the simulation of intelligent thought processes, heuristics and applications of knowledge. The integration of these multiple approaches in solving complex problems is of particular importance. ISMIS provides a forum and a means for exchanging information for those interested purely in theory, those interested primarily in implementation, and those interested in specific research and industrial applications.

Data is considered as the new oil for virtually any company around the word. Hence, data science has established itself as a multi-disciplinary field that holistically addresses challenges in managing data. It subsumes, e.g., database sys­tems as well as data analytics. Within data analytics, unsupervised approaches, in particular, clustering, play a crucial role to discover new insights in the data and turn them into information. Most real-life applications are characterized by overlapping data structures. These can be addressed by soft computing approaches such as fuzzy, rough or granular clustering. The objective of the session is to study the state of the art of soft clustering in data science.

Topics: The special session invites researchers and scientists to submit their high-quality, original works in granular and soft clustering for data science. The topics include, but not limited to:

• Fuzzy clustering, rough clustering, granular clustering
• Three-way decision clustering
• Hybrid clustering and meta clustering
• Dynamic clustering
• Data stream clustering
• Applications of clustering

Special Session Chairs:

Traditionally, Formal Methods have been used as rigorous means to prove correctness and safety of software and hardware systems. They are rooted in logic and reasoning, and aim to provide guarantees that the system is behaving correctly, which is necessary in safety-critical contexts. Such guarantees can be provided automatically for conventional software/hardware systems using verification technologies such as model checking or theorem proving. However, in several (critical) application domains (e.g. planning and scheduling, machine learning, autonomous controllers synthesis, business processes) the underpinning reasoning techniques do not offer the needed guarantees, and reasoning capabilities necessary to justify safety of the application. The scope of this special session is intended to represent research results and discussions about the application and/or combination of Formal Methods (model checking, theorem proving, mathematical reasoning …) to solve problems in different areas such as Planning and Scheduling, Machine Learning, Decision Support Systems, Robotics, Autonomy, Business Processes. The session addresses issues involving solutions to problems that will benefit from the adoption of Formal Methods. Examples are for instance guarantee robustness of a plan, or robustness of a (deep) neural, robustness/safety of a decision support system or of a robotic controller or of a business process.</p>

Topics:  This special session aims at bringing together academic and industrial leaders who will present and discuss the results that combine Formal Methods to solve problems in different areas related to Intelligent Systems. The topics include, but not limited to:

• Intelligent Information Systems
• Autonomic and Evolutionary Computation
• Logic for Artificial Intelligence
• Knowledge Integration and Aggregation
• Intelligent Agent Technology
• Intelligent Data Processing and Analytics

Special Session Chairs:

Roads are becoming more and more technological. An ever-increasing amount of traffic information is collected day by day by different entities. This makes it possible and interesting to extract relevant traffic patterns, and possibly associate some of them to different kinds of anomalous behavior, such as accidents, break of driving regulations, road crimes, etc. The focus of this special session is on methodologies and applications of automatic traffic analysis systems that can detect, track, understand and suitably propose to interested stakeholders the behavior of road users, both in general and from the specific perspective of Police activities and investigations.

Topics of interest include (but are not limited) to:

• detection and tracking of road users (vehicles, bikes, trucks, etc.);
• behavior understanding of road users;
• automatic understanding of the environment in traffic scenarios;
• applications related to traffic surveillance;

• vehicle accident analysis.

Among the stakeholders of this task, both as data provider and as final users, is the Italian National Police, and specifically its Traffic Police branch, whose peculiar responsibilities include watching over the Italian road network, ensuring road patrolling and investigating on typical crimes made on the road (e.g. car theft). The Italian Traffic Police has accumulated massive historical datasets about traffic on Italian highways, and is willing to share publicly part of these data for research purposes. Researchers interested in obtaining the Traffic Police (TRAP) dataset may contact the organizing committee.

Organizing committee:

Recent advances in storage, hardware, information technology, communication, and networking have resulted in a large amount of digital data. Unlike the structured small-scale data previously utilized in traditional machine learning tasks, many of the data available nowadays are large-scale complex data, with features like heterogeneity, multiple labels, multiple modalities, and incompleteness, etc. The increasing complexity of these digital data give rise to new challenges in machine learning techniques, calling for novel and innovative methods and implementations to process these complex data in an efficient and effective manner.

The topics of the special session include, but are not limited to:

• Supervised/unsupervised/semi-supervised learning models for complex data
• Dimensionality reduction and feature extraction for high-dimensional data
• Regularization and generalization in machine learning
• Optimization and numerical methods in machine learning for large-scale data
• Approximation in machine learning
• Distributed/parallel optimization algorithms in machine learning
• Deep learning
• Multi-modal/view/instance/task/label/scale learning
• Multi-graph learning
• Learning with heterogeneous data
• Learning with tensorial data
• Learning with imbalanced data
• Learning with incomplete data
• Machine learning in healthcare
• Machine learning in medicine
• Machine learning in social networks
• Cross-media learning
• Web/text/image mining
• Learning for personalization, advertising, and recommendation

Special Session Chairs:

Paper Submission: Authors are invited to submit their manuscripts (maximum 10 pages) electronically in Springer’s LNCS/LNAI style. Any necessary information concerning typesetting can be obtained directly from Springer’s webpage. All submissions will be subject to review by the ISMIS 2018 program committee in consultation with the special session organizers.

Publication: The accepted papers will be published in ISMIS 2018 proceedings in Springer’s LNAI series.

Important Dates: