This year's application track will consider two different kinds of papers.
The first will be traditional application papers where constraints technology is used to tackle real-life problems and is deployed in practice. Expectations for these papers are: 1. A clear description of the problem and the way how CP technology was used to solve the problem. 2. Quantifiable results like revenue/profit/productivity improvements, or qualitative impact such as improved maintainability or simpler process integration, including an analysis of the contribution that can attributed to constraints technology.
The second category of papers we will consider are papers where (new or existing) constraints technology is integrated in novel ways with other methods, for example CP for data curation, knowledge representation and data mining. Special consideration will be given to CP methods that (a) bridge, or help bridge, predictive and prescriptive analytics; or (b) help assure and explain analytics; or (c) hypothesize rules to turn sample observations into a compact framework where deterministic reasoning techniques can be leveraged; or (d) support human-machine collaborations. For these applications, the novelty of the integrated system as a whole is a key factor, putting special emphasis on a thorough section on prior arts. Also expected are a comprehensive theoretical and/or experimental analysis of the effectiveness and practicability of the proposed usage of CP in the respective application domain.
CP and Data Science
In the last decade, advances in Data Science methods have seen dramatic improvements in effectiveness, leading to widespread attention and successful real-world applications. Despite this, there is a limit to how much can be done using learning methods alone. We argue that finding effective and efficient approaches to synergize learning and reasoning may enable the solution of far more complex and important problems, and potentially improve the way we solve existing problems.
This has been recognized by many. Over the past years, we see two trends in AI and CP conferences: increasingly data mining problems are being solved using constraint solving technology; and even more research is looking into using machine learning techniques to tune algorithms, choose among alternative solution approaches, learn constraints, objectives and search strategies.
This track aims to invite research that sits at the border between constraint solving and data science, with the broadest possible scope.
CP and Operations Research
Constraint programming (CP) has played a pivotal role in the advancement of new modelling and solution techniques in operations research (OR). While global structure is central to inference methods in traditional OR algorithms, the emphasis on individual constraints in CP provides a more flexible modelling language which allows for the exploitation of local substructure that a global view often misses. The combination of these complementary strengths has led to the development of state-of-the-art methodologies in a variety of practical problems, including scheduling, energy and power systems, machine learning, revenue management, vehicle routing, queuing, and global optimization, to name a few.
The purpose of this track is to serve as a venue for research that explores new techniques for the integration or CP and OR, or for novel and interesting applications of CP in OR problems. Examples of possible topics include advances in decomposition methods (such as Benders decomposition and branch-and-price methods), multicriteria optimization, new ways of exploiting OR relaxations (such as linear programming and semidefinite programming) into CP, and new inference methodologies based on dynamic programming or machine learning techniques.
CP and Music
Since its early beginning, Constraint Programming has had a long history of musical applications, for instance on automatic harmonization, rhythm generation tools, musical generation in a given style, constraint languages for music, etc. The music track welcomes articles on any kind of musical application, including (but not restricted to): music or sound generation or processing, music modelling or analysis, generation of a particular musical aspect (chords, notes, rhythms, etc.) of a musical piece, etc. On the CP side, we welcome submissions for any kind of CP techniques, whether they are used in a classical way (solving) or a less classical way (modelling languages, use of CP solving traces, auralization, etc.).
Testing and Verification
The last decade has witnessed a considerable improvement in the efficiency and expressive power of CP solvers, with a consequent impact on (software and hardware) testing and verification application. A deeper integration of solvers and applications is expected with on going research on Constraint Programming (CP) techniques. The Testing and Verification track of CP'2018 will focus on a broad range of topics, without being limited to the ones mentioned below:
- Constraint-based hardware verification
- Constraint-based software testing
- Constraints in formal verification
- Constraints in static and dynamic analysis
- CP solvers for testing applications
- Verification of CP models
- Testing of CP solvers
- Formal verification of CP solvers and optimizers
- Automatic test generation with CP solvers
Multiagent and Parallel CP
The Multiagent and Parallel CP track solicits papers addressing original research on the intersection between constraint solutions and autonomous agents and their interaction, as well as parallel methodologies for constraint-based solutions. Topics of interest include:
- Distributed problem solving
- Teamwork, coalition formation
- Agents’ preferences and preference elicitation
- Single and multi-agent planning and scheduling
- Agent cooperations for practical applications
- Cooperative games
- Non-cooperative games
- Game theory for practical applications
- Parallel search and propagation solutions
- Parallel multi-agent solutions
- Cloud computing for constrained solutions
- GPU approaches to constraint (logic) technologies
- Applications of parallel CP/COP/SAT/ASP
CP, Optimization, and Power System Management
The power system problem has unique characteristics such as fast evolving dynamics, large data sources, and need for real-time operational control. Recently, there is a high demand for efficient algorithms that can address integration of the renewable generation, environmental issues, and development of smart grid technologies. As such information processing techniques are expected to make the real-time operation of power systems more intelligent and reliable. In the last decade, with tremendous advancement in the field of constraint solving, optimization algorithms and the computer hardware technology, existing power system problems can be solved more effectively. This track intends to investigate state-of-the-art work related to all constraint solving and optimization based methodologies that results in efficient operation of the power systems.
Topics of interest:
- Optimization based control
- Carbon emission reduction in energy systems using optimization
- Optimization based approach for integrating renewable energy sources
- Optimal power flow
- Optimal control and scheduling of multi energy systems
- Optimization based approaches for mitigating power quality issues
- Distributed and decentralized optimization