Fuzzy Words Distribution in Management Science Research Articles

By using ZHANG Qiao＇s method for identifying fuzzy words （1998）, the present paper makes a statistic analysis of fuzzy words appearing in about 43,000 words of 6 academic research articles in management science. A distribution curve of fuzzy words in management research articles is drawn in accordance with the structures of the articles, and it is proved to be reasonable for tile discourse function of each section. Furthermore, the fuzzy words are divided into three categories and eleven sub-categories, including Category Ⅰ-fuzzy words, Category Ⅱ-fuzzy words＋non-fuzzy part / non-fuzzy part＋fuzzy words, Category Ⅲ-approximators＋fuzzy words. Distribution of fuzzy words varies in different sections of research articles in management science, because of different communicative purposes or discourse functions. Introduction section and Conclusion section have a higher occurrence rate of fuzzy words than Abstract section and Methodology and Data section.

The Geographical Distribution and Historical Strata of Appellation Words Referring to ＂FATHER＂ in Jin Dialects --Exploring the Causes for the Formation of Historical Strata of Dialect Words

Based on geographical distribution and literature survey, the present study constructs the historical strata of the appellation words referring to ＂FATHER＂ in Jin Dialects. It holds that the patterns of the appellation words are the result of superposition of different historical periods. But their leading words directly inherit the appellation words of standard language of the Song Dynasty and are mainly monosyllabic words. Words used before the Song Dynasty as living fossils of the words only retain in some districts or localities. The different evolution route of＂pa （爸）＂ and ＂tiε （爹）＂ either in Jin dialects or in mandarin is a further proof that Jin dialects and mandarin evolves asynchronously. In addition, it maintains that the formulation of the historical strata is the interaction of the horizontal transmission of languages and systemic changes of local dialects.

Effects of Arbuscular Mycorrhizal Fungi (AMF) on Distribution of Pb Forms in Rhizosphere of Upland Rice

[Objective] The aim was to study the effects of AMF on distribution of Pb in different chemical forms in rhizosphere soil of upland rice. [Method] A pot experiment was conducted to explore effects of AMF inoculation on distribution of Pb in different forms in rhizosphere of rice (Oryzal sativa L.) with Pb in different concentrations (0, 300 and 600 mg/kg). [Result] With inoculation adopted, mycorrzhial colonization rate of upland rice under Pb pdlution root declined substantially with Pb increasing in soils (P<0.05). Compared with non-inoculation, rhizosphere pH significantly enhanced by inoculation; when Pb was at 300 mg/kg, glomalin content in soils improved significantly by inoculation; when Pb was at 600 mg/kg, glomalin content in soils declined substantially (P<0.05). In addition, inoculation significantly improved contents of Pb in exchangeable and organic forms, but lowered Pb in carbonate bound and Fe-Mn oxides bound forms (P<0.05). [Conclusion] The research indicated that AMF inoculation would change distribution of Pb in different forms in rhizosphere soils of upland rice.

Influence of spectral distribution on accommodation-vergence and reading performance

Background:To assess the influence of different spectral energy distribution on accommodation,vergence and reading performance.Methods:A Randomized experimental study was conducted after getting the approval of the Ethical Committee of University of Hyderabad.Forty participants with an age group of 18-21 years was integrated,out of which 50%was male and 50%was female.Subjects with emmetropia and no history of ocular pathology were included in the study.Near point of accommodation(NPA)&near point of convergence(NPC)was measured with the help of royal air force(RAF)ruler followed by near visual task of a readability passage.Results:A statistically significant result was obtained when reading rate,reading speed and NPC was compared among different spectral distribution of light(P<0.001)except NPA(P=0.43).Post hoc analysis showed a significant difference(P<0.001)when tungsten was compared with fluorescent light(FLOU),compact fluorescent light(CFL),and light emitting diode(LED)for reading rate,reading speed and NPC.But there is no noteworthy difference exist when fluorescent was compared with CFL for reading rate(P=0.530)&reading speed(P=0.595).Similarly,LED also showed no considerable difference when compared with CFL(P=0.682)and fluorescent(P=0.490)for NPC.When NPA was assessed within the group LED showed insignificant difference with CFL(P=0.205)and fluorescent(P=0.275)similar like fluorescent and tungsten(P=0.482).Conclusions:This study concluded that reading performance(reading rate and reading speed)and NPC has a significance change if we use inappropriate lighting during visual tasks.It will cause visual fatigue and strain after sustained near work.In addition,tungsten spectral energy influences the convergence which can also show an impact on reading and near visual tasks because of its brightness and miosis.Prolonged reading and working under this lighting can cause convergence disorders and visual fatigue.

Covariance Estimation Using High-Frequency Data： An Analysis of Nord Pool Electricity Forward Data

The modeling of volatility and correlation is important in order to calculate hedge ratios, value at risk estimates, CAPM （Capital Asset Pricing Model betas）, derivate pricing and risk management in general. Recent access to intra-daily high-frequency data for two of the most liquid contracts at the Nord Pool exchange has made it possible to apply new and promising methods for analyzing volatility and correlation. The concepts of realized volatility and realized correlation are applied, and this study statistically describes the distribution （both distributional properties and temporal dependencies） of electricity forward data from 2005 to 2009. The main findings show that the logarithmic realized volatility is approximately normally distributed, while realized correlation seems not to be. Further, realized volatility and realized correlation have a long-memory feature. There also seems to be a high correlation between realized correlation and volatilities and positive relations between trading volume and realized volatility and between trading volume and realized correlation. These results are to a large extent consistent with earlier studies of stylized facts of other financial and commodity markets.

A frequency domain estimation and compensation method for system synchronization parameters of distributed-HFSWR

To analyze the influence of time synchronization error,phase synchronization error,frequency synchronization error,internal delay of the transceiver system,and range error and angle error between the unit radars on the target detection performance,firstly,a spatial detection model of distributed high-frequency surface wave radar(distributed-HFSWR)is established in this paper.In this model,a method for accurate extraction of direct wave spectrum based on curve fitting is proposed to obtain accurate system internal delay and frequency synchronization error under complex electromagnetic environment background and low signal to noise ratio(SNR),and to compensate for the shift of range and Doppler frequency caused by time-frequency synchronization error.The direct wave component is extracted from the spectrum,the range estimation error and Doppler estimation error are reduced by the method of curve fitting,and the fitting accuracy of the parameters is improved.Then,the influence of frequency synchronization error on target range and radial Doppler velocity is quantitatively analyzed.The relationship between frequency synchronization error and radial Doppler velocity shift and range shift is given.Finally,the system synchronization parameters of the trial distributed-HFSWR are obtained by the proposed spectrum extraction method based on curve fitting,the experimental data is compensated to correct the shift of the target,and finally the correct target parameter information is obtained.Simulations and experimental results demonstrate the superiority and correctness of the proposed method,theoretical derivation and detection model proposed in this paper.

基于MapReduce模型的分布式Word文档破解

利用Word加密机制中存在的漏洞,实现了与加密密钥长度无关的常量时间破解,同时提出了基于MapReduce架构实现Word文档破解的方案,简化了分布式程序设计。实验表明了该方案的有效性,分布式计算中节点之间的并行消耗低。

Paragraph Vector Representation Based on Word to Vector and CNN Learning

Document processing in natural language includes retrieval,sentiment analysis,theme extraction,etc.Classical methods for handling these tasks are based on models of probability,semantics and networks for machine learning.The probability model is loss of semantic information in essential,and it influences the processing accuracy.Machine learning approaches include supervised,unsupervised,and semi-supervised approaches,labeled corpora is necessary for semantics model and supervised learning.The method for achieving a reliably labeled corpus is done manually,it is costly and time-consuming because people have to read each document and annotate the label of each document.Recently,the continuous CBOW model is efficient for learning high-quality distributed vector representations,and it can capture a large number of precise syntactic and semantic word relationships,this model can be easily extended to learn paragraph vector,but it is not precise.Towards these problems,this paper is devoted to developing a new model for learning paragraph vector,we combine the CBOW model and CNNs to establish a new deep learning model.Experimental results show that paragraph vector generated by the new model is better than the paragraph vector generated by CBOW model in semantic relativeness and accuracy.

基于Word2vec的语义查询扩展方法

查询扩展是信息检索领域重要研究内容。为了解决信息检索过程中用户提交查询时描述不准确以及查询词不匹配的问题,提出一种基于Word2vec的语义查询扩展方法。使用分布式神经语言概率模型Word2vec训练低维词向量,选取扩展词候选集,利用面向扩展词的查询向量生成方法过滤候选集,使选取的扩展词能更有效地体现整个查询的语义及语法相关性。实验结果表明基于Word2vec的语义查询扩展方法使查全率及查准率均有提高,因此该方法能很好地应用于查询扩展领域。

Application of Word Embedding to Drug Repositioning

As a key technology of rapid and low-cost drug development, drug repositioning is getting popular. In this study, a text mining approach to the discovery of unknown drug-disease relation was tested. Using a word embedding algorithm, senses of over 1.7 million words were well represented in sufficiently short feature vectors. Through various analysis including clustering and classification, feasibility of our approach was tested. Finally, our trained classification model achieved 87.6% accuracy in the prediction of drug-disease relation in cancer treatment and succeeded in discovering novel drug-disease relations that were actually reported in recent studies.

基于词-主题-文本异质网络的短文本分类方法

针对现有分类方法未考虑长距离词的语义相关性和文本间潜在主题共享的问题,提出一种基于词-主题-文本异质网络(WTDHN)的短文本分类方法。通过Word2vec训练词的上下文语义向量;构建词相关性矩阵以充足的词共现信息增强短文本各级别语义学;构建以词、主题和文本为节点的异质网络,并采用图卷积学习节点之间的高阶邻域信息,丰富短文本语义。相较于基准分类模型,该方法在五个公开短文本数据集上的分类准确率平均提高1.56%。

基于语义规则增强的蒙古语情感分布学习

为完善基于先验知识的标记增强方法对于情绪信息的捕捉,提出一种基于语义规则增强的蒙古语情感分布学习方法(semantic rule enhancement based Mongolian emotion distribution learning, SRE-MEDL)。在情感轮和情感词典的基础上,引入程度词典和否定词典,得到各种情感词组合,以此制定相应的语义规则计算情感词权重,将其融入到标记增强中。在情感分布学习中融入从情感分布空间到实例特征空间的反向重构映射来弥补正向映射引起的原始信息丢失问题。对比实验结果显示,在蒙古语和中英文常用数据集上,SRE-MEDL方法在标记增强任务和情感分布学习中的表现均优于现有方法。

Solid particle size characterization by a high-frequency collision response in pneumatic particulate flow

A novel triaxial vibration method is developed for the real-time characterization of the solid particle size distribution(PsD)in pneumatic particulate flow,which is critical for chemical industry.In this work,the particle-wall collision and friction behaviours were analysed by the time-domain statistical and timefrequency joint methods to narrow the high-frequency response range by the initial experiment of free fall for a single particle,interparticle,and multiple particles.Subsequently,verification experiments of PSD characterization in pneumatic flow were performed.First,the quantitative triaxial energy response model that considers the particle size,shape,and mass factors were established.Second,a good agreement of the particle number identification was found between the triaxial vibration energy and mean particle size of 150-550μm.Moreover,the performance with the best accuracy was focused on a range of 42-43 kHz in the x-axis and z-axis and 36.8-38.8 kHz in the y-axis.Finally,the individual particle energy was inversely analysed by the triaxial vibration response within the optimized frequency bands,and the PSD was characterized in real-time by a low error rate,that is,5.2% from the XZ-axis direction of sand(42-43 kHz)and 5.6% from the XYZ-axis of glass(30.9-33.9 kHz,46.2-47.2 kHz,38.3-41.3 kHz for each axis response).Therefore,this research complements the existing approaches for PsD characterization in particulate multiphase flow.

汉语名-量构词词语结构类型及句法生成机理

汉语名-量构词除了表达名语素的概念意义外,还自带类指集合这一特殊语法意义,这与名-量构词特有的组构特征及生成机制密切相关。以往研究多聚焦于名-量构词的结构类型和词汇意义的描写,缺乏对此类构词组构方式和语义变化内在语法理据的探讨。文章基于分布式形态学(DM)理论,从句法的视角探寻名-量构词形成的句法条件,并依托句法结构阐释名-量构词语法意义的来源。文章认为量语素是功能性语素成分,名-量构词是依托句法结构而生成的附加式名词结构形式。名语素与量语素结合,受到功能性量语素的驱动,在词语生成过程中,通过移位获取了名词短语结构中不同层级功能性中心成分的语法特征,进而词汇化为带有语法意义的词汇,其本质是语法词。

汉语常用词演变研究应注意的几个问题

汉语常用词演变研究是汉语词汇史研究的热点之一,研究成果颇多,但也存在一些不足之处有待改进。其中,有五个问题值得格外注意:一是语料的可靠性、有效性和全面性;二是全面收罗概念场成员;三是充分考虑书写形式;四是树立历史发展观;五是在更开阔的视野下考察常用词演变。希望通过这一探讨,能够对今后汉语常用词演变的深入研究有所裨益。

《朱子语类》“冒罩”“眊燥”“莽广”等词源流发展考

《朱子语类》叠置了不同地域的口语语料,其中一些特色方俗词的使用有地域差异。表“粗疏”义的北系方言词“冒罩”源自“毷氉”,“眊燥”是其同义连绵词,衍生有“冒躁”“冒懆”“毛糙”等丰富变体。表“莽撞”义的“莽广”等词与之义近,是南系方言词。

中古“牵拉”概念场词义的演变分析

选取中古时期“牵拉”义词“引”“牵”“曳”“拽”“拖”“挽”六词为研究对象,从共时角度梳理使用情况并得出“牵”“引”为主导词,“挽”“曳”为主要词汇,“拖”“拽”为边缘词汇。进而分析概念要素并辨析词义差别:“引”指不限方向的广义牵拉;“牵”指拉着向前走;“曳”“拽”“拖”指横牵不能或不愿移动之物;“挽”指牵动无法移动的重物。并从认知角度揭示出“引”“牵”“拽”通过隐喻方式衍生出“牵拉”义。

一种新的词语相似度计算方法

提出了一种新的词语相似度计算方法。该方法利用词语关联分布规范化因子,对互信息中目标词和基词的关联度度量方法进行了修正,通过计算目标词和基词的关联度,构造目标词的属性向量,由目标词的属性向量,利用夹角余弦法计算出目标词语相似度。实验结果验证了该方法的有效性。

国内外共词分析研究的文献计量分析

以研究共词分析法的文献为研究对象,以国外的"ISI Web of Science"和国内CNKI(学术期刊数据库)两个较为大型数据库作为数据来源,利用文献计量方法对共词分析相关文献的增长与分布、学科分布、被引分析、作者分布、高频关键词进行分析。以期对共词分析的深入研究与广泛应用提供借鉴与参考。

科技论文关键词特征及其对共词分析的影响

针对传统共词分析中高频词共现矩阵的构建方法提出了一些疑问，包括：抽取高频词作为分析对象的可靠性、高频词矩阵对领域内重要共现关系的保留程度、关键词的语义类型特征和关键词缺失可能带来的影响。通过实证数据揭示了科技论文的关键词词频、共现关系、语义类型的分布特征，并分析了它们对共词分析方法的影响，包括：基于关键词的共词分析只能分析热门知识节点，共词网络实质上是建立在不稳定的单次关联基础之上，而高频词矩阵则会丢失大量重要的共现关系，这些问题是由关键词的语义类型特征决定的，该特征是实现词语间差异化乃至语义化处理的重要切入点。另外，本文在对比关键词增补前后的共词矩阵后发现，增补关键词实质上无法优化高频词矩阵对所分析领域的代表性。在结尾部分，提出了两种可尝试的思路：一是结合关键词频次和共现关系强度抽取分析对象；二是以关键词语义类型为维度构建多维共现矩阵以更好地挖掘多种语义关联。