Rumour stance classification, defined as classifying the stance of specific social media posts into one of supporting, denying, querying or commenting on an earlier post, is becoming of increasing interest to researchers. While most previous work has focused on using individual tweets as classifier inputs, here we report on the performance of sequential classifiers that exploit the discourse features inherent in social media interactions or ‘conversational threads’. Testing the effectiveness of four sequential classifiers – Hawkes Processes, Linear-Chain Conditional Random Fields (Linear CRF), Tree-Structured Conditional Random Fields (Tree CRF) and Long Short Term Memory networks (LSTM) – on eight datasets associated with breaking news stories, and looking at different types of local and contextual features, our work sheds new light on the development of accurate stance classifiers. We show that sequential classifiers that exploit the use of discourse properties in social media conversations while using only local features, outperform non-sequential classifiers. Furthermore, we show that LSTM using a reduced set of features can outperform the other sequential classifiers; this performance is consistent across datasets and across types of stances. To conclude, our work also analyses the different features under study, identifying those that best help characterise and distinguish between stances, such as supporting tweets being more likely to be accompanied by evidence than denying tweets. We also set forth a number of directions for future research. 相似文献
Engaging students with multiple solution problems is considered good practice. Solutions to problems consist of the outcomes
of the problem as well as the methods employed to reach these outcomes. In this study we analyze the results obtained from
two groups of kindergarten children who engaged in one task, the Create an Equal Number Task. This task had five possible
outcomes and five different methods which may be employed in reaching these outcomes. Children, whose teachers had attended
the program Starting Right: Mathematics in Kindergartens, found more outcomes and employed more methods than children whose
teachers did not attend this program. Results suggest that the habit of mind of searching for more than one outcome and employing
more than one method may be promoted in kindergarten. 相似文献
Metric space is a universal and versatile model of similarity that can be applied in various areas of non-text information retrieval. However, a general, efficient and scalable solution for metric data management is still a resisting research challenge. In this work, we try to make an important step towards such management system that would be able to scale to data collections of billions of objects. We propose a distributed index structure for similarity data management called the Metric Index (M-Index) which can answer queries in precise and approximate manner. This technique can take advantage of any distributed hash table that supports interval queries and utilize it as an underlying index. We have performed numerous experiments to test various settings of the M-Index structure and we have proved its usability by developing a full-featured publicly-available Web application. 相似文献
This study compares students’ opportunities to engage in transformational (rule-based) algebraic activity between 2 classes taught by the same teacher and across 2 topics in beginning algebra: forming and investigating algebraic expressions and equivalence of algebraic expressions. It comprises 2 case studies; each involves a teacher teaching in two 7th grade classes. All 4 classes used the same textbook. Analysis of classroom videotapes (15–19 lessons in each class) revealed that the opportunities to engage in transformational algebraic activity related to forming and investigating algebraic expressions were similar in each teacher’s 2 classes. By contrast, substantial differences were found between 1 teacher’s classes with regard to the opportunities to engage in transformational algebraic activity related to equivalence of algebraic expressions. The discussion highlights the contribution of the interplay among the mathematical topic, the teacher, and the class to shaping students’ learning opportunities. Specifically, the mathematical topic appeared to play a prominent role in certain situations, with the topic involving deductive reasoning generating high variation in classes of 1 teacher but not in the other’s.
Educational Studies in Mathematics - We used an assessment platform to study the potential of rich student data obtained online to influence classroom instruction and help teachers respond to... 相似文献
Whereas most English language sub-typing schemes for dyslexia (e.g., Castles & Coltheart, 1993) have focused on reading accuracy for words varying in regularity, such an approach may have limited utility for reading
disability sub-typing beyond English in which fluency rather than accuracy is the key discriminator of developmental and individual
differences in reading ability. The present study investigated the viability of an accuracy/fluency-based typology in a regular
orthography, pointed Hebrew. We sought evidence of true or “hard” accuracy/rate subtypes in the strict (double dissociation)
sense of selective impairment on only one dimension in the presence of normal levels of performance on the other dimension. In a nationally representative sample of fourth graders, we were able to identify
a specific accuracy-disabled sub-group as well as an equally specific rate-disabled subgroup. Validating this subdivision,
we show that the nature of reading performance in these subgroups and their converging cognitive/linguistic profiles are unique
and distinctive on variables other than the measures used to define them. While the rate-specific disability appeared to reflect
a general deficit in speed of processing affecting reading rate, and rapid automatized naming of print-related material, the
accuracy-only disability subgroup displayed selective deficits in phonological awareness and morphological knowledge. Biosocial,
demographic, and instructional factors, furthermore, did not explain the sub-group differences. It appears that both these
subtypes are equally prevalent each counting close to 10% of the population. 相似文献
The purpose of this study was to investigate gender-related differences in the relationship between the development of formal reasoning skills and learning interests during the early adolescent stage. For this purpose, 249 students, from seventh to ninth grade, were assessed for their level of mastery of formal reasoning skills by a test based on videotaped simple experiments. Learning interests were assessed by a written response to an open question. Results showed that adolescent boys develop patterns of formal reasoning before their girl classmates. In addition, boys tend to prefer science and technology subjects, while girls tend to prefer language, social studies, and humanities. Analysis of interactions showed that boys' tendency toward science and technology is positively correlated to their age and development of formal reasoning, while girls' tendency to the above subjects is positively related to their development of formal reasoning capacity, but inversely related to their age. Possible explanations to the above-described findings and suggestions for instructional modes that may increase girls' interest in science and technology are discussed. 相似文献