An SNN-Based Semantic Role Labeling Model with Its Network Parameters Optimized Using an Improved PSO Algorithm

Semantic role labeling (SRL) is a fundamental task in natural language processing to find a sentence-level semantic representation. The semantic role labeling procedure can be viewed as a process of competition between many order parameters, in which the strongest order parameter will win by competition and the desired pattern will be recognized. To realize the above-mentioned integrative SRL, we use synergetic neural network (SNN). Since the network parameters of SNN directly influence the synergetic recognition performance, it is important to optimize the parameters. In this paper, we propose an improved particle swarm optimization (PSO) algorithm based on log-linear model and use it to effectively determine the network parameters. Our contributions are two-folds: firstly, a log-linear model is introduced to PSO algorithm which can effectively make use of the advantages of a variety of different knowledge sources, and enhance the decision making ability of the model. Secondly, we propose an improved SNN model based on the improved PSO and show its effectiveness in the SRL task. The experimental results show that the proposed model has a higher performance for semantic role labeling with more powerful global exploration ability and faster convergence speed, and indicate that the proposed model has a promising future for other natural language processing tasks.     

A sensor-based feet motion recognition of graphical user interface controls

Motion sensing has now become one of the crucial parts of modern life. The tech products for entertainment create huge competition in the modern market, with the front runners such as game and handset manufacturers. Users can use their own bodies to control the systems without Conversational User Interfaces (CUIs) and Graph User Interfaces (GUIs). With many benefits, the technologies of NUI is getting more and more important, which also are usually accompanied by the sensor developments. Hence, system designers uses the sensors on the embedded platforms used to develop the system devices for different body movement control interfaces, such as using the gravitational sensor to control the systems or using touchscreen detection. However most of the body movement is restricted to the hand portion for the system platform, making it not as dynamic as the traditional monitor consoles. Thus, this restriction decreases the multitude and availability of controlling modes in system devices. In this study, a sensor-based gait recognition was proposed, in order to provide a novel natural user interface for control systems except the operating modes of gesture.     

Emotional speech feature normalization and recognition based on speaker-sensitive feature clustering

In this paper we propose a feature normalization method for speaker-independent speech emotion recognition. The performance of a speech emotion classifier largely depends on the training data, and a large number of unknown speakers may cause a great challenge. To address this problem, first, we extract and analyse 481 basic acoustic features. Second, we use principal component analysis and linear discriminant analysis jointly to construct the speaker-sensitive feature space. Third, we classify the emotional utterances into pseudo-speaker groups in the speaker-sensitive feature space by using fuzzy k-means clustering. Finally, we normalize the original basic acoustic features of each utterance based on its group information. To verify our normalization algorithm, we adopt a Gaussian mixture model based classifier for recognition test. The experimental results show that our normalization algorithm is effective on our locally collected database, as well as on the eNTERFACE’05 Audio-Visual Emotion Database. The emotional features achieved using our method are robust to the speaker change, and an improved recognition rate is observed.     

Complete model-based equivalence class testing

In this article, we present a model-based black-box equivalence partition testing strategy, together with a formal proof of its completeness properties. The results apply to reactive systems with large, possibly infinite input data types and finite internal and output data ranges that may be enumerated with acceptable effort. The investigation is performed on a semantic level and applies to all concrete test models whose behavioural semantics can be encoded as a variant of state transition systems. Test suite construction is performed in relation to a given fault model \(\mathcal{F}\) for which a finite black-box test suite can be constructed which is complete with respect to \(\mathcal{F}\). It is shown how the test suite generation can be effectively implemented by model-based testing tools, using propositional representations of behavioural model semantics and constraint solvers. A SysML model of the ceiling speed monitoring function of the European Train Control System is presented as a case study, to explain theory application to a concrete modelling formalism.     

Predicting temporal centrality in Opportunistic Mobile Social Networks based on social behavior of people

Predicting the centrality of nodes is a significant problem for different applications in Opportunistic Mobile Social Networks (OMSNs). However, when calculating such metrics, current studies focused on analyzing static networks that do not change over time or using aggregated contact information over a period of time. Furthermore, the centrality measured in the past is not verified whether it is useful as a predictor for the future. In this paper, in order to capture the dynamic behavior of people, we focus on predicting nodes’ future centrality (importance) from the temporal perspective using real mobility traces in OMSNs. Three important centrality metrics, namely betweenness, closeness, and degree centrality, are considered. Through real trace-driven simulations, we find that nodes’ future centrality is highly predictable due to natural social behavior of people. Then, based on the observations in the simulation, we design several reasonable prediction methods to predict nodes’ future temporal centrality. Finally, extensive real trace-driven simulations are conducted to evaluate the performance of our proposed methods. The results show that the Recent Weighted Average Method performs best in the MIT Reality trace, and the recent Uniform Average Method performs best in the Infocom 06 trace. Furthermore, we also evaluate the impact of parameters m and w on the performance of the proposed methods and find proper values of different parameters for each proposed method at the same time.     

薄壁不锈钢管波纹卡粘式连接方法及性能研究

介绍了薄壁不锈钢管波纹卡粘式连接的原理、安装方法、注意事项以及工程应用情况,试验研究了波纹卡粘式连接件的连接强度、密封性、使用寿命等。结果表明:经完全固化后,波纹卡粘式连接件的最大破坏扭矩达到41N·m,接近于DN15薄壁不锈钢管产生塑性变形时的扭矩强度;连接密封性能相关的8项指标全部符合国标要求。在550℃的高温炉内,连接件的泄漏量为0.02m3/h,远小于规定的0.17m3/h。根据阿累尼乌斯图推算出连接件的使用寿命可达70a以上。卫生检测显示连接件中的厌氧胶不会影响水质,清洁卫生。     

面向中小企业的电商平台挖掘系统设计

中小企业电商平台在资金和技术不足的情况下,难以快速有效的构建简单而实用的电商数据挖掘系统.文章从分析电商海量数据资源的来源、提取方法及其流程入手,提出了一个基本的数据挖掘系统方案架构,从而实现为中小企业电商平台服务的一整套数据挖掘系统.该系统架构简单实用、易搭建、易整合,可以为中小企业提供更好的服务支持.     

Excel对学生成绩的频度分析的研究

数据频度分析是Excel的重要功能之一,Excel对学生成绩频度分析有四种方法：COUNTIF函数、COUNTIFS函数、FRE-QUENCY函数和数据透视表。四种方法均能方便快捷的完成对成绩数据的频度分析,完美高效地解决成绩统计问题。