The Late Permian Emeishan Large Igneous Province (ELIP) is commonly regarded as being located in the western part of the Yangtze craton, SW China, with an asymmetrical shape and a small area. This area, however, is ju...The Late Permian Emeishan Large Igneous Province (ELIP) is commonly regarded as being located in the western part of the Yangtze craton, SW China, with an asymmetrical shape and a small area. This area, however, is just a maximum estimation because some parts of the ELIP were not recognized or dismembered and destroyed during the Triassic to Cenozoic tectonism. In this paper, the chemostratigraphical data of the Zongza block, the Garze-Litang belt and the Songpan-Garze block suggest that the Late Permian basalts in these areas have remarkable similarities to the ELIP basalts in petrography and geochemistry. Flood basalts in the Sanjiangkou area are composed of the lower part of the low-Ti (LT) tholeiite and the upper part of the high-Ti (HT) tholeiite, which is the same as the flood basalts on the western margin of the Yangtze craton. Flood basalts in the Zongza and Songpan-Garze areas, which are far from the Yangtze craton, consist of HT tholeiite only. This is the same as the flood basalts within the Yangtze craton. Therefore we argue that these contemporary basalts all originated from the Emeishan mantle plume, and the ELIP could have a significant westward extension with an outcropped area of over 500,000 km2. This new scenario shows that the LT tholeiite occurs on the western margin of the Yangtze craton, while the HT tholeiite overlying the LT basalts occupies the whole area of the ELIP.展开更多
An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods ...An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods such as MI-based rule evaluating, weighted rule quantification and element-based n-gram probability approximation are presented. Dynamic Viterbi algorithm is adopted to search the best path in lattice. To strengthen the model, transformation-based error-driven rules learning is adopted. Applying proposed model to Chinese Pinyin-to-character conversion, high performance has been achieved in accuracy, flexibility and robustness simultaneously. Tests show correct rate achieves 94.81% instead of 90.53% using bi-gram Markov model alone. Many long-distance dependency and recursion in language can be processed effectively.展开更多
Successful prediction of protein domain boundaries provides valuable information not only for the computational structure prediction of muhi-domain proteins but also for the experimental structure determination. A nov...Successful prediction of protein domain boundaries provides valuable information not only for the computational structure prediction of muhi-domain proteins but also for the experimental structure determination. A novel method for domain boundary prediction has been presented, which combines the support vector machine with domain guess by size algorithm. Since the evolutional information of multiple domains can be detected by position specific score matrix, the support vector machine method is trained and tested using the values of position specific score matrix generated by PSI-BLAST. The candidate domain boundaries are selected from the output of support vector machine, and are then inputted to domain guess by size algorithm to give the final results of domain boundary, prediction. The experimental results show that the combined method outperforms the individual method of both support vector machine and domain guess by size.展开更多
文摘The Late Permian Emeishan Large Igneous Province (ELIP) is commonly regarded as being located in the western part of the Yangtze craton, SW China, with an asymmetrical shape and a small area. This area, however, is just a maximum estimation because some parts of the ELIP were not recognized or dismembered and destroyed during the Triassic to Cenozoic tectonism. In this paper, the chemostratigraphical data of the Zongza block, the Garze-Litang belt and the Songpan-Garze block suggest that the Late Permian basalts in these areas have remarkable similarities to the ELIP basalts in petrography and geochemistry. Flood basalts in the Sanjiangkou area are composed of the lower part of the low-Ti (LT) tholeiite and the upper part of the high-Ti (HT) tholeiite, which is the same as the flood basalts on the western margin of the Yangtze craton. Flood basalts in the Zongza and Songpan-Garze areas, which are far from the Yangtze craton, consist of HT tholeiite only. This is the same as the flood basalts within the Yangtze craton. Therefore we argue that these contemporary basalts all originated from the Emeishan mantle plume, and the ELIP could have a significant westward extension with an outcropped area of over 500,000 km2. This new scenario shows that the LT tholeiite occurs on the western margin of the Yangtze craton, while the HT tholeiite overlying the LT basalts occupies the whole area of the ELIP.
文摘An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods such as MI-based rule evaluating, weighted rule quantification and element-based n-gram probability approximation are presented. Dynamic Viterbi algorithm is adopted to search the best path in lattice. To strengthen the model, transformation-based error-driven rules learning is adopted. Applying proposed model to Chinese Pinyin-to-character conversion, high performance has been achieved in accuracy, flexibility and robustness simultaneously. Tests show correct rate achieves 94.81% instead of 90.53% using bi-gram Markov model alone. Many long-distance dependency and recursion in language can be processed effectively.
基金Supported by the National Natural Science Foundation of China （No. 60435020）
文摘Successful prediction of protein domain boundaries provides valuable information not only for the computational structure prediction of muhi-domain proteins but also for the experimental structure determination. A novel method for domain boundary prediction has been presented, which combines the support vector machine with domain guess by size algorithm. Since the evolutional information of multiple domains can be detected by position specific score matrix, the support vector machine method is trained and tested using the values of position specific score matrix generated by PSI-BLAST. The candidate domain boundaries are selected from the output of support vector machine, and are then inputted to domain guess by size algorithm to give the final results of domain boundary, prediction. The experimental results show that the combined method outperforms the individual method of both support vector machine and domain guess by size.
