Search

Serious concerns about the way research is organized collectively are increasingly being raised. They include the escalating costs of research and lower research productivity, low public trust in researchers to report the truth, lack of diversity, poor community engagement, ethical concerns over research practices, and irreproducibility. Open science (OS) collaborations comprise of a subset of open practices including open access publication, open data sharing and the absence of restrictive intellectual property rights with which institutions, firms, governments and communities are experimenting in order to overcome these concerns. We gathered two groups of international representatives from a large variety of stakeholders to construct a toolkit to guide and facilitate data collection about OS and non-OS collaborations. Ultimately, the toolkit will be used to assess and study the impact of OS collaborations on research and innovation. The toolkit contains the following four elements: 1) an annual report form of quantitative data to be completed by OS partnership administrators; 2) a series of semi-structured interview guides of stakeholders; 3) a survey form of participants in OS collaborations; and 4) a set of other quantitative measures best collected by other organizations, such as research foundations and governmental or intergovernmental agencies. We opened our toolkit to community comment and input. We present the resulting toolkit for use by government and philanthropic grantors, institutions, researchers and community organizations with the aim of measuring the implementation and impact of OS partnership across these organizations. We invite these and other stakeholders to not only measure, but to share the resulting data so that social scientists and policy makers can analyse the data across projects.

Coordinating agents in a complex environment is a hard problem, but it can become even harder when certain characteristics of the tasks, like the required number of agents, are unknown. In these settings, agents not only have to coordinate themselves on the different tasks, but they also have to learn how many agents are required for each task. To contribute to this problem, we present in this paper a selective perception reinforcement learning algorithm which enables agents to learn the required number of agents that should coordinate their efforts on a given task. Even though there are continuous variables in the task description, agents in our algorithm are able to learn their expected reward according to the task description and the number of agents. The results, obtained in the RoboCupRescue simulation environment, show an improvement in the agents overall performance.

This paper presents a POMDP approximation method, called RTBSS, which is based on a look-ahead search in order to plan in a real-time dynamic environment. The basis of our approach is to avoid computing full policies in POMDP problems. Our approach is especially motivated by real-time environments where the state space is too large to consider traditional offline algorithms. We then proceed with an online approach to find at each step, the action that maximizes the agent expected utility. To this end, we present the formalism behind our approach. Then, we present how the approach was applied on three different environments. Let us mention finally that this approach was successfully implemen ted for the RoboCupRescue 2004 international competition where we finished second.

We address this article in relation to blogging and personal knowledge publishing (PKP). Derived from blogging, personal knowledge publishing is a form of Web-based communication that lowers social and linguistic barriers, facilitating knowledge migration across disciplinary boundaries. The aggregate output of this practice could provide a promising basis for intelligent systems development. These individual-centered, community-discussion-support systems have emerged as an interesting complement to journals, mailing lists, and other media for interdisciplinary-minded individuals. PKP offers individual researchers a tool for building strong interdisciplinary research networks. These networks enable individuals to find relevant literature and experts outside their core research community.

The Internet has by now become a truly global and multilingual community: non-English speakers presently account for two-thirds of Internet users. Yet, while dozens of languages are being used simultaneously on the same network, the language barriers make it difficult for connection and collaboration to happen across languages.

Partially Observable Markov Decision Processes (POMDPs) provide a rich framework for sequential decision-making under uncertainty in stochastic domains. However, solving a POMDP is often intractable except for small problems due to their complexity. Here, we focus on online approaches that alleviate the computational complexity by computing good local policies at each decision step during the execution. Online algorithms generally consist of a lookahead search to find the best action to execute at each time step in an environment. Our objectives here are to survey the various existing online POMDP methods, analyze their properties and discuss their advantages and disadvantages; and to thoroughly evaluate these online approaches in different environments under various metrics (return, error bound reduction, lower bound improvement). Our experimental results indicate that state-of-the-art online heuristic search methods can handle large POMDP domains efficiently.

This paper discusses the design and implementation of processes and tools to support the collaborative creation and maintenance of multilingual wiki content. A wiki is a website where a large number of participants are allowed to create and modify content using their Web browser. This simple concept has revolutionized collaborative authoring on the web, enabling among others, the creation of Wikipedia, the world's largest online encyclopedia. On many of the largest and highest profile wiki sites, content needs to be provided in more than one language. Yet, current wiki engines do not support the efficient creation and maintenance of such content. Consequently, most wiki sites deal with the issue of multilingualism by spawning a separate and independent site for each language. This approach leads to much wasted effort since the same content must be researched, tracked and written from scratch for every language. In this paper, we investigate what features could be implemented in wiki engines in order to deal more effectively with multilingual content. We look at how multilingual content is currently managed in more traditional industrial contexts, and show how this approach is not appropriate in a wiki world. We then describe the results of a User-Centered Design exercise performed to explore what a multilingual wiki engine should look like from the point of view of its various end users. We describe a partial implementation of those requirements in our own wiki engine (LizzyWiki), to deal with the special case of bilingual sites. We also discuss how this simple implementation could be extended to provide even more sophisticated features, and in particular, to support the general case of a site with more than two languages. Finally, even though the paper focuses primarily on multilingual content in a wiki context, we argue that translating in this “Wiki Way”, may also be useful in some traditional industrial settings, as a way of dealing better with the fast and ever-changing nature of our modern internet world., WIKISym 2006: the Conference Wiki of the 2006 International Symposium, August 21-23, 2006, Odense, Denmark

Coordinating agents in a complex environment is a hard problem, but it can become even harder when certain characteristics of the tasks, like the required number of agents, are unknown. In those settings, agents not only have to coordinate themselves on the different tasks, but they also have to learn how many agents are required for each task. To achieve that, we have elaborated a selective perception reinforcement learning algorithm to enable agents to learn the required number of agents. Even though there were continuous variables in the task description, the agents were able to learn their expected reward according to the task description and the number of agents. The results, obtained in the RoboCupRescue, show an improvement in the agents overall performance.

In the RoboCupRescue simulation, the PoliceForce agents have to decide which roads to clear to help other agents to navigate in the city. In this article, we present how we have modelled their environment as a POMDP and more importantly we present our new online POMDP algorithm enabling them to make good decisions in real-time during the simulation. Our algorithm is based on a look-ahead search to find the best action to execute at each cycle. We thus avoid the overwhelming complexity of computing a policy for each possible situation. To show the efficiency of our algorithm, we present some results on standard POMDPs and in the RoboCupRescue simulation environment.

In this paper, we present an online method for POMDPs, called RTBSS (Real-Time Belief Space Search), which is based on a look-ahead search to find the best action to execute at each cycle in an environment. We thus avoid the overwhelming complexity of computing a policy for each possible situation. By doing so, we show that this method is particularly efficient for large real-time environments where offline approaches are not applicable because of their complexity. We first describe the formalism of our online method, followed by some results on standard POMDPs. Then, we present an adaptation of our method for a complex multiagent environment and results showing its efficiency in such environments.

In this paper, we introduce an approach called RTBSS (Real-Time Belief Space Search) for real-time decision making in large POMDPs The approach is based on a look-ahead search that is applied online each time the agent has to make a decision RTBSS is particularly interesting for large real-time environments where offline solutions are not applicable because of their complexity.

The current and charge signals on individual pixels of CdTe γ-ray cameras are calculated. The time and position dependent free carrier densities following the absorption of a γ-ray in the detector are first calculated by solving the continuity equations. These density distributions are then combined with the physical electric field and the pixel weighting field to provide the current induced through an individual pixel from which the charge signal is also obtained. Simulations are made for various pixel sizes. These simulations confirm that the energy resolution of electron collecting pixel detectors must improve when smaller pixels are used since small pixels are less sensitive to holes.

Choosing between multiple alternative tasks is a hard problem for agents evolving in an uncertain real-time multiagent environment. An example of such environment is the RoboCupRescue simulation, where at each step an agent has to choose between a number of tasks. To do that, we have used a reinforcement learning technique where an agent learns the expected reward it should obtain if it chooses a particular task. Since all possible tasks can be described by a lot of attributes, we have used a selective perception technique to enable agents to narrow down the description of each task.

Weblogs have recently emerged as a popular means of sharing information on the Web. While they effectively foster the networking of participants on a one-to-one basis, so far they have lacked the capacity of allowing the establishment of many-to-many communication relationships. The paper describes recent work on facilitating group forming processes and the sharing of content among weblog authors with shared interests. We have designed, implemented and tested the Internet Topic Exchange, a system that enables weblog posts to be shared among open groups in a form that we call topic channels. After nearly a year of operation, more than 200 topic channels have been created; several of them have been very active and have brought together many participants. This suggests that our approach to enabling weblog authors and topical content to cluster while retaining the advantages of personal publishing is a viable one.

In this abstract, we present a complex multiagent environment, the RoboCupRescue simulation, and show some of the learning opportunities for the coordination of agents in this environment.