面向时间序列的微博话题演化模型研究

话题演化研究有助于追踪用户的喜好和话题的发展趋势,对于舆情预警具有重要意义。目前,话题演化方法注重运用话题生成模型实现话题演化分析,忽略了话题中时间因素和背景词的存在。以传统话题生成模型LDA为基础,将其扩展为微博话题生成模型MTLDA。MTLDA模型增加了对背景词的考虑,提高了话题生成的效率,同时对微博话题集进行时间片划分,利用KL距离计算相邻时间片话题距离,分析话题演化情况。以新浪微博数据为例进行实验,结果表明,MTLDA模型通过时间片划分完成了微博话题的生成,话题演化结果与实际情况吻合。     

Intuitionistic uncertain linguistic partitioned Bonferroni means and their application to multiple attribute decision-making

The Bonferroni mean (BM) was originally introduced by Bonferroni and generalised by many other researchers due to its capacity to capture the interrelationship between input arguments. Nevertheless, in many situations, interrelationships do not always exist between all of the attributes. Attributes can be partitioned into several different categories and members of intra-partition are interrelated while no interrelationship exists between attributes of different partitions. In this paper, as complements to the existing generalisations of BM, we investigate the partitioned Bonferroni mean (PBM) under intuitionistic uncertain linguistic environments and develop two linguistic aggregation operators: intuitionistic uncertain linguistic partitioned Bonferroni mean (IULPBM) and its weighted form (WIULPBM). Then, motivated by the ideal of geometric mean and PBM, we further present the partitioned geometric Bonferroni mean (PGBM) and develop two linguistic geometric aggregation operators: intuitionistic uncertain linguistic partitioned geometric Bonferroni mean (IULPGBM) and its weighted form (WIULPGBM). Some properties and special cases of these proposed operators are also investigated and discussed in detail. Based on these operators, an approach for multiple attribute decision-making problems with intuitionistic uncertain linguistic information is developed. Finally, a practical example is presented to illustrate the developed approach and comparison analyses are conducted with other representative methods to verify the effectiveness and feasibility of the developed approach.     

Cloud decision model for selecting sustainable energy crop based on linguistic intuitionistic information

In recent years, sustainable energy crop has become an important energy development strategy topic in many countries. Selecting the most sustainable energy crop is a significant problem that must be addressed during any biofuel production process. The focus of this study is the development of an innovative multi-criteria decision-making (MCDM) method to handle sustainable energy crop selection problems. Given that various uncertain data are encountered in the evaluation of sustainable energy crops, linguistic intuitionistic fuzzy numbers (LIFNs) are introduced to present the information necessary to the evaluation process. Processing qualitative concepts requires the effective support of reliable tools; then, a cloud model can be used to deal with linguistic intuitionistic information. First, LIFNs are converted and a novel concept of linguistic intuitionistic cloud (LIC) is proposed. The operations, score function and similarity measurement of the LICs are defined. Subsequently, the linguistic intuitionistic cloud density-prioritised weighted Heronian mean operator is developed, which served as the basis for the construction of an applicable MCDM model for sustainable energy crop selection. Finally, an illustrative example is provided to demonstrate the proposed method, and its feasibility and validity are further verified by comparing it with other existing methods.     

区间二元语义Bonferroni集成算子及其决策应用

区间二元语义值是一种常用的不确定环境下决策信息表达形式。考虑到决策信息的交叉影响作用,定义了区间二元语义值的Bonferroni平均算子以及相应的加权形式,在此基础上,给出了组合形式的区间二元语义值的加权Bonferroni平均算子的概念,并研究了算子的幂等性、单调性等数学性质,给出了基于C-I2TLWBA算子区间二元语义值的集成模型和决策应用。实例表明了该模型具有较好的有效性。     

基于动态P-模糊语言算子的VIKOR决策方法

针对直觉模糊信息解决动态多属性决策问题时存在的不足,将Pythagorean模糊语言信息引入到动态多属性决策问题,提出一种基于Pythagorean模糊语言信息集成算子的多准则妥协排序（VIKOR）决策方法。引入Pythagorean模糊语言得分函数、精确函数、距离计算公式等概念,提出动态 Pythagorean模糊语言加权平均（DPFLWA）算子,并研究DPFLWA算子的基本性质。最后,基于DPFLWA算子和VIKOR方法,构建一种动态 Pythagorean模糊语言多属性决策方法。通过第三方逆向物流服务商的选择实例,表明该方法的可行性和有效性。     

时变论域下的语言动力学轨迹

为了描述在不同时刻内涵发生变化的模糊集合, 本文通过综合模糊集合、语言动力系统(Linguistic dynamic systems,LDS) 及动态规划, 提出了时变论域与动态模糊规则理论, 将时变论域分为离散型与连续型两类, 每一类分为递增型、递减型及波动型三种, 并讨论如何在时变论域上建立动态模糊规则以及进行词计算, 最后给出时变论域下的语言动力学轨迹.     

Automatic acquisition of semantic collocation from corpora

The real difficulty in developing practical NLP systems is due to the fact that we do not know in advance what actualinstances of knowledge should be used in the application system, even though we know in advance whattypes of knowledge are required. An effective method for extracting linquistic knowledge from corpora is needed. We propose automatic linguistic knowledge acquisition from sublanguage corpora. The system combines existing linquistic knowledge and human intervention with corpus based techniques. The algorithm involves a Gradual Approximation, which works to converge linguistic knowledge gradually towards desirable results. We conducted three experiments. The first experiment revealed the characteristic of this algorithm and the others proved the effectiveness of this algorithm for a real corpus. The results show the algorithm is promising, though there are some problems; the practical problem of the parameters, the formalism problems to include more linguistic features and the combination with other linguistic clues for more development. We would like to continue the research to perform further experiments and to improve the algorithm.     

基于语言学知识的查询个性化潜力预测

大多数关于个性化信息检索的研究都是针对所有查询的,很少有研究试图回答哪些查询将受益于个性化信息检索。从大规模知识库中挖掘大量的语言学知识,用于预测查询的个性化潜力,这些知识包括概念词、歧义词、同义词等。使用语言学知识作为特征,预测查询的个性化潜力,可以减少查询日志的数据稀疏问题的影响。实验结果表明该方法的有效性和可行性。     

Beyond argumentation and comparison/contrast: Extending the Socrates CAI design principles to classroom teaching and the interpretation and production of other forms of discourse

This study compares pencil-and-paper and computer-assisted versions of a college process/model program in critical thinking and academic writing to a traditional composition program. Students in the experimental sections used more linguistic markers of argument and comparison/contrast, attempted more arguments and made stronger arguments. CAI students also did better than students in the pencil-and-paper sections on some measures. Applications of computer assisted techniques in studying meaning in poetry, teaching technical writing and in teaching practice, suggest new approaches to collaborative thinking and writing. Thomas Bacig is a Professor of Humanities at the University of Minnesota, Duluth with research i interests in: computer assisted instruction in writing, reading and thinking; science fiction; and forest history. His most recent publication is: How Computer-Assisted Instruction Informs the Design of Conventional Classroom Activities in English Composition (with D. Larmouth), Proceedings of the 19th Annual Small College Computing Symposium, 1987. Robert Evans is Associate Professor of Philosophy at the University of Minnesota, Duluth with research interests in logic, American philosophy and philosophy of law. His most recent publication is a review of Volume 11, John Dewey, The Later Works containing Dewey's major political work, Liberalism and Social Action, in Transactions of the Charles S. Peirce Society, 15, winter 1989.Donald Larmouth is Professor of Linguistics and Dean of Arts, Sciences and Graduate programs at the University of Wisconsin, Green Bay. He has research interests in dialect geography, computers in composition and language policy. His most recent publication is Does Linguistic Heterogenity Erode National Unity? in Thomas Tonneson's Ethnicity and Language, Institute on Race and Ethnicity of the University of Wisconsin System, 1988.Kenneth Risdon is Assistant Professor of Composition at the University of Minnesota, Duluth with research interests in writing with computers, use of computer networks, and computer analysis of text.     

过程控制中的一种数据挖掘算法

结合粗糙集理论,提出了一种适合工业控制领域使用的数据挖掘算法.工业控制的采样数据以决策表的形式表示,用模糊c-方法对采样信息表进行模糊语言化,依据适合实际应用的语言值关联规则支持度和信任度,通过属性约简,剔除冗余规则,挖掘出隐含的关联规则.以水泥生产线中分解炉控制为例,采用现场实时采集的数据集,挖掘出了相关控制对象之间的关联关系.关联规则的挖掘,有助于在先验经验不足的条件下获取模糊规则,通过动态的模糊模型的建立,实现一定程度的智能控制.