Extending WHIRL with background knowledge for improved text classification

Intelligent use of the many diverse forms of data available on the Internet requires new tools for managing and manipulating heterogeneous forms of information. This paper uses WHIRL, an extension of relational databases that can manipulate textual data using statistical similarity measures developed by the information retrieval community. We show that although WHIRL is designed for more general similarity-based reasoning tasks, it is competitive with mature systems designed explicitly for inductive classification. In particular, WHIRL is well suited for combining different sources of knowledge in the classification process. We show on a diverse set of tasks that the use of appropriate sets of unlabeled background knowledge often decreases error rates, particularly if the number of examples or the size of the strings in the training set is small. This is especially useful when labeling text is a labor-intensive job and when there is a large amount of information available about a particular problem on the World Wide Web.

The Role of Anatomy Computer-Assisted Learning on Spatial Abilities of Medical Students

Currently, medical education context poses different challenges to anatomy, contributing to the introduction of new pedagogical approaches, such as computer-assisted learning (CAL). This approach provides insight into students' learning profiles and skills that enhance anatomy knowledge acquisition. To understand the influence of anatomy CAL on spatial abilities, a study was conducted. A total of 671 medical students attending Musculoskeletal (MA) and Cardiovascular Anatomy (CA) courses, were allocated to one of three groups (MA Group, CA Group, MA + CA Group). Students' pre-training and post-training spatial abilities were assessed through Mental Rotations Test (MRT), with scores ranging between 0-24. After CAL training sessions, students' spatial abilities performance improved (9.72 ± 4.79 vs. 17.05 ± 4.57, P < 0.001). Although male students in both MA Group and CA Group show better baseline spatial abilities, no sex differences were found after CAL training. The improvement in spatial abilities score between sessions (Delta MRT) was correlated with Musculoskeletal Anatomy training sessions in MA Group (r = 0.333, P < 0.001) and MA + CA Group (r = 0.342, P < 0.001), and with Cardiovascular Anatomy training sessions in CA Group (r = 0.461, P = 0.001) and MA + CA Group (r = 0.324, P = 0.001). Multiple linear regression models were used, considering the Delta MRT as dependent variable. An association of Delta MRT to the amount of CAL training and the baseline spatial abilities was observed. The results suggest that CAL training in anatomy has positive dose-dependent effect on spatial abilities.     

马克思主义哲学视角下的语篇语言学

语篇语言学又称话语语言学,兴起于20世纪70年代,是现代语言学的重要分支之一,它的诞生与发展,使语言学的研究脱离了形式一逻辑的固有模式,从语言的本质出发去研究语言的基本规律。语言学的这个进步,反映了语言学家们从哲学层面对语言的重新审视,不再把语言看作是一成不变的形式化的内容,而是从语言作为人类的交际手段这一基本属性的角度来研究语言本身。从马克思主义哲学的视角来审视语篇语言学的诞生、研究对象和研究方法,无疑能够更好地探究语篇语言学的哲学基础。     

Social media opinion summarization using emotion cognition and convolutional neural networks

Quickly and accurately summarizing representative opinions is a key step for assessing microblog sentiments. The Ortony-Clore-Collins (OCC) model of emotion can offer a rule-based emotion export mechanism. In this paper, we propose an OCC model and a Convolutional Neural Network (CNN) based opinion summarization method for Chinese microblogging systems. We test the proposed method using real world microblog data. We then compare the accuracy of manual sentiment annotation to the accuracy using our OCC-based sentiment classification rule library. Experimental results from analyzing three real-world microblog datasets demonstrate the efficacy of our proposed method. Our study highlights the potential of combining emotion cognition with deep learning in sentiment analysis of social media data.     

Cloud-based big data analytics for customer insight-driven design innovation in SMEs

Fast development of IT and ICT facilitate customers to post a large volume of their concerns and expectation online, which are widely accepted to be a valuable resource for product designers. However, it is found that only a small number of small and medium-sized enterprises (SMEs) have capabilities to leverage customer online insights for design innovation, which often demonstrate a significant share in national economies growth. To discover the beneath reasons regarding the barrier that prevent them to make effective utilization, in this study, as a concrete example, manufacturing SMEs in the South Wales and Greater Manchester industrial areas of the UK are focused and their potential motivations for using and knowledge of big data-based customer analytics are investigated. An exploratory survey was conducted in terms of the type of customer data they have, the storage approaches, the volume of customer data, etc. Next, a carefully devised exploratory study was undertaken to understand how SMEs perceive the relations between customer data and product design, how about their expectations from big customer data analytics and what really challenges SMEs to exploit the value of big customer data. Besides, a demonstration platform is developed to present SMEs an automatic process of analysing customer online reviews and the capacity on customer insights acquisition and strategic decision making. Finally, findings from two focus groups indicate the different managerial and technical considerations required for SMEs considering implementing big data and customer analytics. This study encourages SMEs to welcome big customer data and suggests that a cloud-based approach may be the most appropriate way of giving access to big data analytics techniques.     

一种基于特征编码的Word文档数字水印算法

本文基于文本数字水印的特征编码思想,通过设置Word文档中字符的下划线格式以嵌入和检测水印。实验表明,该算法有较强的不可见性,能较好的嵌入和检测水印,从而达到数字水印的一般要求。     

基于粗集的支持向量机文本分类方法研究

本文提出了一种基于粗糙集的支持向量机文本分类方法。该方法利用粗糙集的约简理论降低了支持向量的维数,同时保证了分类性能。实验表明该方法能获得较好的分类效果。     

院校研究与学习分析

院校研究与学习分析在学习研究方面同异并存,皆利用大数据探寻学生学习的优劣.院校研究主要依赖大学的管理和调查数据,学习分析主要使用学习管理系统收集的数据.院校研究聚焦于帮助大学“知己知彼”,学习分析旨在助力“知学生”“知学习”“知教育”.院校研究为学校层面决策提供实证研究证据,学习分析识别学生个体学习行为从而支持个性化学...     

Authorship attribution based on a probabilistic topic model

This paper describes, evaluates and compares the use of Latent Dirichlet allocation (LDA) as an approach to authorship attribution. Based on this generative probabilistic topic model, we can model each document as a mixture of topic distributions with each topic specifying a distribution over words. Based on author profiles (aggregation of all texts written by the same writer) we suggest computing the distance with a disputed text to determine its possible writer. This distance is based on the difference between the two topic distributions. To evaluate different attribution schemes, we carried out an experiment based on 5408 newspaper articles (Glasgow Herald) written by 20 distinct authors. To complement this experiment, we used 4326 articles extracted from the Italian newspaper La Stampa and written by 20 journalists. This research demonstrates that the LDA-based classification scheme tends to outperform the Delta rule, and the χ² distance, two classical approaches in authorship attribution based on a restricted number of terms. Compared to the Kullback–Leibler divergence, the LDA-based scheme can provide better effectiveness when considering a larger number of terms.     

The use of player physical and technical skill match activity profiles to predict position in the Australian Football League draft

This study investigated the extent to which position in the Australian Football League (AFL) national draft is associated with individual game performance metrics. Physical/technical skill performance metrics were collated from all participants in the 2014 national under 18 (U18) championships (18 games) drafted into the AFL (n = 65; 17.8 ± 0.5 y); 232 observations. Players were subdivided into draft position (ranked 1–65) and then draft round (1–4). Here, earlier draft selection (i.e., closer to 1) reflects a more desirable player. Microtechnology and a commercial provider facilitated the quantification of individual game performance metrics (n = 16). Linear mixed models were fitted to data, modelling the extent to which draft position was associated with these metrics. Draft position in the first/second round was negatively associated with “contested possessions” and “contested marks”, respectively. Physical performance metrics were positively associated with draft position in these rounds. Correlations weakened for the third/fourth rounds. Contested possessions/marks were associated with an earlier draft selection. Physical performance metrics were associated with a later draft selection. Recruiters change the type of U18 player they draft as the selection pool reduces. juniors with contested skill appear prioritised.