Abstract. In the community of sentiment analysis, supervised learning techniques have been shown to perform very well. When transferred to another domain, however, a supervised sentiment classifier often performs extremely bad. This is so-called domain-transfer problem. In this work, we attempt to attack this problem by making the maximum use of both the old-domain data and the unlabeled new-domain data. To leverage knowledge from the old-domain data, we proposed an effective measure, i.e., Frequently Co-occurring Entropy (FCE), to pick out generalizable features that occur frequently in both domains and have similar occurring probability. To gain knowledge from the newdomain data, we proposed Adapted Naïve Bayes (ANB), a weighted transfer version of Naive Bayes Classifier. The experimental results indicate that proposed approach could improve the performance of base classifier dramatically, and even provide much better performance than the transfer-learning baseline, i.e. the Naïve Bayes Transfer Classifier (NTBC).
the larger subunit of RCA (45-46 kDa), which is more thermostable and increased in response to moderate heat stress, and the smaller isoform (38-39 kDa) of RCA may play important roles in maintaining the photosynthetic capability by EBR under stress conditions.
In this paper we present a novel strategy, DragPushing, for improving the performance of text classifiers. The strategy is generic and takes advantage of training errors to successively refine the classification model of a base classifier. We describe how it is applied to generate two new classification algorithms; a Refined Centroid Classifier and a Refined Naïve Bayes Classifier. We present an extensive experimental evaluation of both algorithms on three English collections and one Chinese corpus.The results indicate that in each case, the refined classifiers achieve significant performance improvement over the base classifiers used. Furthermore, the performance of the Refined Centroid Classifier implemented is comparable, if not better, to that of state-of-the-art support vector machine (SVM)-based classifier, but offers a much lower computational cost.
Mowing is an important land management practice for natural semi-arid regions. A growing body of empirical evidence shows that different mowing regimes affect the functioning of grassland ecosystems. However, the responses of plant functional traits to long-term mowing and their allometric scaling under long-term mowing are poorly understood. For a better understanding of the effects of mowing on grassland ecosystems, we analyzed the allometric traits of leaves and stems of Leymus chinensis (Trin.) Tzvel., a dominant grass species in eastern Eurasian temperate grassland, at different mowing intensities (no clipping, clipping once every two years, once a year and twice a year). Experiments were conducted on plots established over a decade ago in a typical steppe of Xilinhot, Inner Mongolia, China. Results showed that most of the functional traits of L. chinensis decreased with the increased mowing intensity. The responses of leaves and stems to long-term mowing were asymmetric, in which leaf traits were more stable than stem traits. Also significant allometric relationships were found among most of the plant functional traits under the four mowing treatments. Sensitive traits of L. chinensis (e.g. leaf length and stem length) were primary indicators associated with aboveground biomass decline under high mowing intensity. In conclusion, the allometric growth of different functional traits of L. chinensis varies with different long-term mowing practices, which is likely to be a strategy used by the plant to adapt to the mowing disturbances.
This paper takes the ecological water conveyance project (EWCP) that transfers water from the Bosten Lake, to Daxihaizi Reservoir, and finally to the Taitema Lake as a case study to analyze the dynamic change of the groundwater depth, the vegetation responses to the elevation of the groundwater depth as well as the relationship between the groundwater depth and the natural vegetation. The results from many years' monitoring in field indicate: (1) the groundwater depth has been elevating gradually with the increase in the times of watering and the elevation range has been expanding continuously in the lower reaches of Tarim River. Correspondingly, the natural vegetation has a favorable response to the elevation of the groundwater depth. The change of the natural vegetation has accordance with that of the groundwater depth. Such facts not only show that groundwater is a key factor to the growth of the native vegetation but also prove it is feasible that the degraded ecosystem can be restored and protected by the EWCP; (2) the results of analysis of the spatial-temporal response of the natural vegetation to watering reveals that the beneficial influence of the EWCP on the ecosystem in the lower Tarim River is a long-term process; (3) in terms of the function and structure of ecosystem after watering in the lower reaches of Tarim River, the EWCP does not still reach the goal of ecological restoration at a large spatial scale at present. Based on such monitoring results, some countermeasures and suggestions for the future restoration strategy are proposed so as to provide a theoretical basis for restoring and protecting the ecosystem in Tarim River, and meanwhile it can also provide some scientific references for implementing the similar ecological projects in other areas.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.