Knowledge-directed Adaptation in Multi-level Agents

Most work on adaptive agents have a simple, single layerarchitecture. However, most agent architectures support three levels ofknowledge and control: a reflex level for reactive responses, a deliberatelevel for goal-driven behavior, and a reflective layer for deliberateplanning and problem decomposition. In this paper we explore agentsimplemented in Soar that behave and learn at the deliberate and reflectivelevels. These levels enhance not only behavior, but also adaptation. Theagents use a combination of analytic and empirical learning, drawing from avariety of sources of knowledge to adapt to their environment. We hypothesize that complete, adaptive agents must be able to learn across all three levels.     

Interleaving Planning and Robot Execution for Asynchronous User Requests

ROGUE is an architecture built on a real robot which provides algorithms for the integration of high-level planning, low-level robotic execution, and learning. ROGUE addresses successfully several of the challenges of a dynamic office gopher environment. This article presents the techniques for the integration of planning and execution.ROGUE uses and extends a classical planning algorithm to create plans for multiple interacting goals introduced by asynchronous user requests. ROGUE translates the planner';s actions to robot execution actions and monitors real world execution. ROGUE is currently implemented using the PRODIGY4.0 planner and the Xavier robot. This article describes how plans are created for multiple asynchronous goals, and how task priority and compatibility information are used to achieve appropriate efficient execution. We describe how ROGUE communicates with the planner and the robot to interleave planning with execution so that the planner can replan for failed actions, identify the actual outcome of an action with multiple possible outcomes, and take opportunities from changes in the environment.ROGUE represents a successful integration of a classical artificial intelligence planner with a real mobile robot.     

Adaptive personal agents

Personal agents are an important advance in the management of electronic information. Accurate user profiling is critical to the personalisation of agent-based services. In this paper we describe a framework for personal agents, which provides a range of services using a common user profile. The user profile is automatically learned from observation of the electronic documents a user reads, and tracks user interests over time. We report on a large trial of this framework, which has given us important insights into the effectiveness of personal agent applications. In particular we discuss the benefits obtained from the use of a common profile and the interoperation of several personal agent applications.     

基于移动代理的信息收集方法研究

传统的用软盘收集信息的方法存在一些不足之处，本文介绍了移动代理技术，给出了用移动代理技术来实现信息收集的方法，对该方法进行了安全性分析，并说明了该方法的优点。     

The Value of Temptation

There is an implicit assumption in electronic commerce that induces the buyers to believe that their deals will be handled appropriately. However, after a seller has already committed to a buyer, he may be tempted by several requests though he will not be able to supply them all. We analyze markets in which a finite set of automated buyers interacts repeatedly with a finite set of automated sellers. These sellers can satisfy one buyer at a time, and they can be tempted to break a commitment they already have. We have found the perfect equilibria that exist in markets with a finite horizon, and with an unrestricted horizon. A significant result stemming from our study reveals that sellers are almost always tempted to breach their commitments. However, we also show that if markets' designers implement an external mechanism that restricts the automated buyers actions, then sellers will keep their commitments.     

Influence of Sodium Dodecyl Sulfate and Tween 20 on Fungal Growth and Toluene Degradation in a Vapor-Phase Bioreactor

Biofiltration, a process in which contaminated air is passed through a biologically active bed, can be used to remove volatile organic pollutants from air streams. While most biofilters rely on bacteria to degrade organic pollutants, biofilters utilizing a fungal biofilm have the potential to be an efficient and robust treatment alternative. In this study, two surfactants were evaluated for their ability to activate the spores of the fungus Exophiala lecanii-corni and reduce the start-up period typically observed in fungal vapor-phase bioreactors. Sodium dodecyl sulfate, an anionic surfactant, was found to inhibit growth of E. lecanii-corni. Polyoxyethylene sorbitan monolaurate (Tween 20), a nonionic surfactant, stimulated bud formation and enhanced toluene degradation in E. lecanii-corni cultures. Tween 20 was also found to enhance inoculum development by shortening the lag period during culture growth. However, when bioreactors were presoaked in medium containing Tween 20, washout of the fungal cells occurred during inoculation, inhibiting the initial fungal biofilm development. After a seven-day start-up period, no detrimental effects on reactor performance due to residual Tween 20 were observed, and toluene elimination capacities of up to 150?g/m3?h were achieved.     

Industrial application integration using the unification approach to agent-enabled semantic SOA

Within the enterprise, information infrastructure as the mean to bring together different software applications is the key component to enable cooperation and information and knowledge exchange in an open-distributed environment. The paper addresses the issues of industrial application integration in business processes using agent-enabled service-oriented architectures (SOA). In this paper, we show that agent-enabled SOA can play an important role for service integration. Our framework combines Web services (WS) and intelligent agent technologies orchestrated by a business process management system. This framework looks for unification of agent and WS service models, is grounded in a semantic SOA of an agent platform and is supported by the Component Agent Platform Over .NET Framework agent platform tools. We describe the architecture and illustrate the approach by an industrial application scenario from petroleum wells’ drilling.     

Assessing the Efficiency of Triangular Gold Nanoparticles as NIR Photothermal Agents In Vitro and Melanoma Tumor Model

Photothermal therapy (PTT) is gaining a lot of interest as a cancer treatment option with minimal side effects due to the efficient photothermal agents employed. They are based on nanomaterials that, upon laser irradiation, absorb photon energy and convert it into heat to induce hyperthermia, which destroys the cancer cells. Here, the unique light-to-heat conversion features of three different gold nanotriangular nanoparticles (AuNTs) are evaluated with respect to their absorption properties to select the most efficient nanoheater with the highest potential to operate as an efficient photothermal agent. AuNTs with LSPR response in- and out- of resonance with the 785 nm near-infrared (NIR) excitation wavelength are investigated. Upon 15 min laser exposure, the AuNTs that exhibit a plasmonic response in resonance with the 785 nm laser line show the highest photothermal conversion efficacy of 80%, which correlates with a temperature increase of 22 °C. These photothermal properties are well-preserved in agarose-based skin biological phantoms that mimic the melanoma tumoral tissue and surrounding healthy tissue. Finally, in vitro studies on B16.F10 melanoma cells prove by fluorescence staining and MTT assay that the highest phototoxic effect after NIR laser exposure is induced by AuNTs with LSPR response in resonance with the employed laser line, thus demonstrating their potential implementation as efficient photothermal agents in PTT.     

Development of Anti-LRRC15 Small Fragments for Imaging Purposes Using a Phage-Display ScFv Approach

The human leucine-rich repeat-containing protein 15 (LRRC15) is a membrane protein identified as a marker of CAF (cancer-associated fibroblast) cells whose overexpression is positively correlated with cancer grade and outcome. Nuclear molecular imaging (i.e., SPECT and PET) to track LRRC15 expression could be very useful in guiding further therapeutic strategies. In this study, we developed an ScFv mouse phage-display library to obtain small fragment antibodies against human LRRC15 for molecular imaging purposes. Mice were immunized with recombinant human LRRC15 (hLRRC15), and lymph node cells were harvested for ScFv (single-chain variable fragment) phage-display analysis. The built library was used for panning on cell lines with constitutive or induced expression after transfection. The choice of best candidates was performed by screening various other cell lines, using flow cytometry. The selected candidates were reformatted into Cys-ScFv or Cys-diabody by addition of cysteine, and cloned in mammalian expression vectors to obtain batches of small fragments that were further used in site-specific radiolabeling tests. The obtained library was 1.2 × 10⁷ cfu/µg with an insertion rate >95%. The two panning rounds performed on cells permittedenrichment of 2 × 10⁻³. Screening with flow cytometry allowed us to identify 28 specific hLRRC15 candidates. Among these, two also recognized murine LRCC15 and were reformatted into Cys-ScFv and Cys-diabody. They were expressed transiently in a mammalian system to obtain 1.0 to 4.5 mg of Cys fragments ready for bioconjugation and radiolabeling. Thus, in this paper, we demonstrate the relevance of the phage-display ScFv library approach for the fast-track development of small antibodies for imaging and/or immunotherapy purposes.     

Conjugates of Methylene Blue with Cycloalkaneindoles as New Multifunctional Agents for Potential Treatment of Neurodegenerative Disease

The development of multi-target-directed ligands (MTDLs) would provide effective therapy of neurodegenerative diseases (ND) with complex and nonclear pathogenesis. A promising method to create such potential drugs is combining neuroactive pharmacophoric groups acting on different biotargets involved in the pathogenesis of ND. We developed a synthetic algorithm for the conjugation of indole derivatives and methylene blue (MB), which are pharmacophoric ligands that act on the key stages of pathogenesis. We synthesized hybrid structures and performed a comprehensive screening for a specific set of biotargets participating in the pathogenesis of ND (i.e., cholinesterases, NMDA receptor, mitochondria, and microtubules assembly). The results of the screening study enabled us to find two lead compounds (4h and 4i) which effectively inhibited cholinesterases and bound to the AChE PAS, possessed antioxidant activity, and stimulated the assembly of microtubules. One of them (4i) exhibited activity as a ligand for the ifenprodil-specific site of the NMDA receptor. In addition, this lead compound was able to bypass the inhibition of complex I and prevent calcium-induced mitochondrial depolarization, suggesting a neuroprotective property that was confirmed using a cellular calcium overload model of neurodegeneration. Thus, these new MB-cycloalkaneindole conjugates constitute a promising class of compounds for the development of multitarget neuroprotective drugs which simultaneously act on several targets, thereby providing cognitive stimulating, neuroprotective, and disease-modifying effects.