Adaptive Dialogue Strategy Selection through Imprecise Probabilistic Query Answering

In a human-computer dialogue system, the dialogue strategy can range from very restrictive to highly flexible. Each specific dialogue style has its pros and cons and a dialogue system needs to select the most appropriate style for a given user. During the course of interaction, the dialogue style can change based on a user’s response and the system observation of the user. This allows a dialogue system to understand a user better and provide a more suitable way of communication. Since measures of the quality of the user’s interaction with the system can be incomplete and uncertain, frameworks for reasoning with uncertain and incomplete information can help the system make better decisions when it chooses a dialogue strategy. In this paper, we investigate how to select a dialogue strategy based on aggregating the factors detected during the interaction with the user. For this purpose, we use probabilistic logic programming (PLP) to model probabilistic knowledge about how these factors will affect the degree of freedom of a dialogue. When a dialogue system needs to know which strategy is more suitable, an appropriate query can be executed against the PLP and a probabilistic solution with a degree of satisfaction is returned. The degree of satisfaction reveals how much the system can trust the probability attached to the solution.

Adaptive uncertain information fusion to enhance plan selection in BDI agent systems

Correctly modelling and reasoning with uncertain information from heterogeneous sources in large-scale systems is critical when the reliability is unknown and we still want to derive adequate conclusions. To this end, context-dependent merging strategies have been proposed in the literature. In this paper we investigate how one such context-dependent merging strategy (originally defined for possibility theory), called largely partially maximal consistent subsets (LPMCS), can be adapted to Dempster-Shafer (DS) theory. We identify those measures for the degree of uncertainty and internal conflict that are available in DS theory and show how they can be used for guiding LPMCS merging. A simplified real-world power distribution scenario illustrates our framework. We also briefly discuss how our approach can be incorporated into a multi-agent programming language, thus leading to better plan selection and decision making.

Threshold-based substream selection for closed-loop spatial multiplexing

We propose a low-complexity closed-loop spatial multiplexing method with limited feedback over multi-input-multi-output (MIMO) fading channels. The transmit adaptation is simply performed by selecting transmit antennas (or substreams) by comparing their signal-to-noise ratios to a given threshold with a fixed nonadaptive constellation and fixed transmit power per substream. We analyze the performance of the proposed system by deriving closed-form expressions for spectral efficiency, average transmit power, and bit error rate (BER). Depending on practical system design constraints, the threshold is chosen to maximize the spectral efficiency (or minimize the average BER) subject to average transmit power and average BER (or spectral efficiency) constraints, respectively. We present numerical and Monte Carlo simulation results that validate our analysis. Compared to open-loop spatial multiplexing and other approaches that select the best antenna subset in spatial multiplexing, the numerical results illustrate that the proposed technique obtains significant power gains for the same BER and spectral efficiency. We also provide numerical results that show improvement over rate-adaptive orthogonal space-time block coding, which requires highly complex constellation adaptation. We analyze the impact of feedback delay using analytical and Monte Carlo approaches. The proposed approach is arguably the simplest possible adaptive spatial multiplexing system from an implementation point of view. However, our approach and analysis can be extended to other systems using multiple constellations and power levels.

Where the Lake Meets the Sea: Strong Reproductive Isolation Is Associated with Adaptive Divergence between Lake Resident and Anadromous Three-Spined Sticklebacks

Contact zones between divergent forms of the same species are often characterised by high levels of phenotypic diversity over small geographic distances. What processes are involved in generating such high phenotypic diversity? One possibility is that introgression and recombination between divergent forms in contact zones results in greater phenotypic and genetic polymorphism. Alternatively, strong reproductive isolation between forms may maintain distinct phenotypes, preventing homogenisation by gene flow. Contact zones between divergent freshwater-resident and anadromous stickleback (Gasterosteus aculeatus L.) forms are numerous and common throughout the species distribution, offering an opportunity to examine these contrasting hypotheses in greater detail. This study reports on an interesting new contact zone located in a tidally influenced lake catchment in western Ireland, characterised by high polymorphism for lateral plate phenotypes. Using neutral and QTL-linked microsatellite markers, we tested whether the high diversity observed in this contact zone arose as a result of introgression or reproductive isolation between divergent forms: we found strong support for the latter hypothesis. Three phenotypic and genetic clusters were identified, consistent with two divergent resident forms and a distinct anadromous completely plated population that migrates in and out of the system. Given the strong neutral differentiation detected between all three morphotypes (mean FST = 0.12), we hypothesised that divergent selection between forms maintains reproductive isolation. We found a correlation between neutral genetic and adaptive genetic differentiation that support this. While strong associations between QTL linked markers and phenotypes were also observed in this wild population, our results support the suggestion that such associations may be more complex in some Atlantic populations compared to those in the Pacific. These findings provide an important foundation for future work investigating the dynamics of gene flow and adaptive divergence in this newly discovered stickleback contact zone.

Directed Selection using Reinforcement Learning for the Examination Timetabling Problem

Traditional heuristic approaches to the Examination Timetabling Problem normally utilize a stochastic method during Optimization for the selection of the next examination to be considered for timetabling within the neighbourhood search process. This paper presents a technique whereby the stochastic method has been augmented with information from a weighted list gathered during the initial adaptive construction phase, with the purpose of intelligently directing examination selection. In addition, a Reinforcement Learning technique has been adapted to identify the most effective portions of the weighted list in terms of facilitating the greatest potential for overall solution improvement. The technique is tested against the 2007 International Timetabling Competition datasets with solutions generated within a time frame specified by the competition organizers. The results generated are better than those of the competition winner in seven of the twelve examinations, while being competitive for the remaining five examinations. This paper also shows experimentally how using reinforcement learning has improved upon our previous technique.