Search Results (4)

Deep learning has enabled significant progress in source code generation, aiming to reduce the manual, error-prone, and time-consuming aspects of software development. While many existing models rely on recurrent neural networks (RNNs) with sequence-to-sequence architectures, these approaches struggle with the long and complex token sequences typical in source code. To address this, we propose a grammar-based convolutional neural network (CNN) combined with a tree-based representation to enhance accuracy and efficiency. Our model achieves state-of-the-art results on the benchmark HEARTHSTONE dataset, with a BLEU score of 81.4 and an Acc+ of 62.1%. We further evaluate the model on our proposed dataset, AST2CVCode, designed for computer vision applications, achieving 86.2 BLEU and 51.9% EM. Additionally, we introduce BLEU+, an enhanced evaluation metric tailored for functional correctness in code generation, which achieves a BLEU+ score of 92.0% on the AST2CVCode dataset. These results demonstrate the effectiveness of our approach in both model architecture and evaluation methodology. Full article

Fires are an essential aspect of the grassland ecosystem across the Great Plains of North America. Wildfires can also transform surrounding rocks to appear like hearths or hearthstones used by prehistoric people. A grassland fire that swept through part of a historic ranch located along the eastern escarpment of the Southern High Plains of Texas has created surface features that mimicked the appearance of hearths. Fourteen wildfire features resembling hearths have been documented, and thermally modified rocks from the surface of three of these features were analyzed to investigate the impact of natural fires on the landscape. The results demonstrate that wildfires can create features resembling hearths when an adjacent shrub is burned. An excavation and detailed analysis, however, suggest that (1) the tops of thermally modified rocks from a wildfire will often have a relatively darker Munsell color value in comparison to their bottom halves, and (2) wildfire features will likely have a thinner cross-section of ash and larger pieces of charcoal produced from the incomplete combustion of the nearby shrub and deadfall. The broader implications are useful for understanding site formation processes within temperate grassland settings in other places. Full article

This study aims to examine firstly the motivations of esports spectators driving them to engage in consumption behaviour, and secondly, two spectator groups (League of Legends, LOL; Hearthstone) to compare the findings of the independence model and the competing model. In recent decades, the concept of esports has emerged as a major component of the sports industry and, therefore, of the global economy. However, the basic functioning of this new sector is relatively poorly understood. This study considers consumer motivations as they relate to esports and aims to assess how selected motivations interact. The motivations chosen in five categories were adopted from the Uses and Gratifications Theory. The independence model (based on Uses and Gratifications Theory (UGT)) and competing model (based on multiple theoretical perspectives) were applied to the LOL and Hearthstone spectator groups. Data (n = 574) were collected via online surveys with cross-validation measured and established between the two groups. The findings showed that social integrative motivations positively impacted consumption behaviour across game genres. Affective motivation partially mediated the relationship between social integrative motivation and consumption behaviour in LOL, and cognitive and personal integrative motivations positively influenced consumption behaviour in Hearthstone. The tension-release motivation had no significant association with consumption behaviour for spectators of either game. The findings can help the commercial interests of different esports game genres to predict why people consume particular esports and thus aid effective marketing strategies. Full article

Monte Carlo Tree Search has proved to be very efficient in the broad domain of Game AI, though it suffers from high dimensionality in cases of large branching factors. Several pruning techniques have been proposed to tackle this problem, most of which require explicit domain knowledge. In this study, an approach using neural networks to determine the number of actions to be pruned, depending on the iterations run and the total number of possible actions, is proposed. Multi-armed bandit simulations with the UCB1 formula are employed to generate suitable datasets for the networks’ training and a specifically designed process is followed to select the best combination of the number of iterations and actions for pruning. Two pruning Monte Carlo Tree Search variants are investigated, based on different actions’ expected rewards’ distributions, and they are evaluated in the collectible card game Hearthstone. The proposed technique improves the performance of the Monte Carlo Tree Search algorithm in different setups of computational limitations regarding the available number of tree search iterations and is significantly boosted when combined with supervised learning trained-state value predicting models. Full article