Customised text entry devices for motor-impaired users

The standard QWERTY keyboard is the principal text entry device for word processing and computer-based communications. For many motor-impaired individuals, and in particular those without intelligible speech, the low text entry rates they can typically achieve is a major problem. For some, the QWERTY design is completely inappropriate. Alternative designs that can appreciably increase these rates would greatly enhance their ability to communicate. This paper considers and compares several approaches to the design of text entry devices for motor-impaired users. A general method for customising (i e, optimising) these designs is employed, and consideration is given to designs requiring significantly fewer input switches than the 26 or more keys required by QWERTY. Use is made of language statistics in the design procedure, and the increased availability of inexpensive, powerful computers is directly exploited.     

Text/shape classifier for mobile applications with handwriting input

The paper provides a practical solution to a real-time text/shape differentiation problem for online handwriting input. The proposed structure of the classification system comprises stroke grouping and stroke classification blocks. A new set of features is derived that has low computational complexity. The method achieves 98.5 % text/shape classification accuracy on a benchmark dataset. The proposed stroke grouping machine learning approach improves classification robustness in relation to different input styles. In contrast to the threshold-based techniques, this grouping adaptation enhances the overall discriminating accuracy of the text/shape recognition system by 11.3 %. The solution improves system’s response on a touch-screen device.     

基于手势识别的文字输入系统

基于视觉的手势识别是实现新一代人机交互的关键技术。通过手势识别向屏幕输入文字以供搜索查找的系统基本没有,在现有的手势识别基础上,利用汉语字母和数字对应的手语作为输入手势,采用微软的kinect获取深度图像,对其进行手势分割。通过Canny算法提取手势的边缘,利用小波矩提取特征,得到手势字母,实现了具有手势识别以及基于文字输入功能的系统。实验表明该系统能够准确有效地实现汉字的输入。     

文本录入测试的VB编程

通过编程实现对文字录入的比较测试,统计速度及得分。     

文本录入测试的VB编程

通过编程实现对文字录入的比较测试，统计速度及得分。     

Graphical input: Can computers understand people?

Computer graphics provides a primary means of interaction between architects and CAD systems. It is argued that existing graphical input procedures to CAD systems are universally bad. Conventional drawing operations are described in order to identify required properties of good input systems. Inadequacies of current systems are traced to choice of computing technology and the study discusses opportunities offered by new technology, as in the example of Prolog.     

面向情感聚类的文本相似度计算方法研究

在文本情感分析时,使用无监督的聚类方法,可以有效节省人力和数据资源,但同时也面临聚类精度不高的问题。相似性是文本聚类的主要依据,该文从文本相似度计算的角度,针对情感聚类中文本—特征向量的高维和稀疏问题,以及对评论文本潜在情感因素的表示问题,提出一种基于子空间的文本语义相似度计算方法(RESS)。实验结果表明,基于RESS的文本相似度计算方法,有效解决了文本向量的高维问题,更好地表达了文本间情感相似性,并获得较好的聚类结果。     

Mobile touchscreen user interfaces: bridging the gap between motor-impaired and able-bodied users

Touchscreen mobile devices are highly customizable, allowing designers to create inclusive user interfaces that are accessible to a broader audience. However, the knowledge to provide this new generation of user interfaces is yet to be uncovered. The goal was to thoroughly study mobile touchscreen interfaces and provide guidelines for informed design. The paper presents an evaluation performed with 15 tetraplegic and 18 able-bodied users that allowed to identify their main similarities and differences within a set of interaction techniques (Tapping, Crossing, and Directional Gesturing) and parameterizations. Results show that Tapping and Crossing are the most similar and easy to use techniques for both motor-impaired and able-bodied users. Regarding Tapping, error rates start to converge at 12 mm, showing to be a good compromise for target size. As for Crossing, it offered a similar level of accuracy; however, larger targets (17 mm) are significantly easier to cross for motor-impaired users. Directional Gesturing was the least inclusive technique. Regarding position, edges showed to be troublesome. For instance, they have shown to increase Tapping precision for disabled users, while decreasing able-bodied users’ accuracy when targets are too small (7 mm). It is argued that despite the expected error rate disparity, there are clear resemblances between user groups, thus enabling the development of inclusive touch interfaces. Tapping, a traditional interaction technique, was among the most effective for both target populations, along with Crossing. The main difference concerns Directional Gesturing that in spite of its unconstrained nature shows to be inaccurate for motor-impaired users.     

面向短文本的特征选择及文本表示

短文本由于其稀疏性、实时性、非标准性等特点,在文本特征选择和文本表示方面存在较多问题,从而影响文本分类精度。针对文本特征选择方面存在较高的特征维数灾难的问题,提出一种二阶段的文本特征选择算法。首先在互信息算法的基础上,引入平衡因子、频度、集中度、词性及词在文本中的位置等5个指标对互信息值进行计算,然后将排序结果靠前的特征集初始化进行遗传算法的训练从而得到最优特征集合。因为TFIDF在计算时针对的是整篇语料而没有考虑类间分布不均的情况,在计算IDF公式时引入方差,并将改进后的TFIDF公式对Word2Vec词向量进行加权表示文本。将改进算法应用在人工构建的百科用途短文本语料集中进行实验,实验结果表明改进的文本特征选择算法和文本表示算法对分类效果有2%~5%的提升。     

一种有效的脆弱型文本数字水印技术

给出一种有效的用于文本认证的脆弱型数字水印方案。通过提取文本的分块特征,结合密钥技术,该方案能够有效地检测文本是否被窜改,并能定位窜改发生的位置,从而可达到检验文本完整性的目的。而在作完整性检验时,该方案并不需要其他的辅助性认证数据。实验表明,该方案是一种比较有效的文本认证方案。     

含未知输入的广义系统的状态与输入估计

研究了含未知输入的广义系统的输入解耦观测器设计问题. 在系统脉冲能控的条件下通过系统输入-状态对的非奇异变换, 把此问题等价地转化为正常状态空间系统的相应问题. 用大家熟知的方法设计正常状态空间系统的观测器, 从而得到广义系统的输入解耦观测器. 然后用广义系统的观测器状态和系统输出的线性组合渐近估计系统的状态与未知输入.     

改进的OPTICS算法及其在文本聚类中的应用

基于密度的OPTICS聚类算法以可视化的结果输出方式直观呈现语料结构,但由于其结果组织策略在处理稀疏点时的局限性,算法实际性能未能得到充分发挥。本文针对此缺陷提出一种有效的结果重组织策略以辅助稀疏点的重新定位,并针对文本领域的特点改变距离度量方法,形成了OPTICS-Plus文本聚类算法。在真实文本分类语料上的实验表明,我们的结果重组织策略能够辅助算法产生更为清晰反映语料结构的可达图,与K-means算法的比较则证实了OPTICS-Plus具有较为良好的聚类性能。     

文本挖掘的时态文本关联规则算法研究

由于数据库的频繁更新,时态数据库隐藏了大量的未知信息,因此针对实时更新的数据库应产生相应的时态关联规则.虽然关联规则算法已经被深入广泛地研究,但在文本数据中时态关联规则算法的研究还不多见.在深入了解时态关联规则算法及其在文本数据中的研究价值后,以时态文本为对象进行了时态关联规则算法的研究,建立了时态文本数据的时间表示模型,提出了文本时态关联规则算法SPFM,最后通过实验对算法进行了有效性验证,结果表明该算法是正确可行的.     

文本特征区域与文本过滤的匹配机制

为了根据用户的信息需求,在因特网上搜索相关文本,该文提出了一种文本过滤的匹配机制,其基本思想是：利用基于词典的概念扩张方法,改进用户模板。计算扩张的用户模板与文本的全局相似度,获取初步的过滤结果;在文本特征区域,进行标题、摘要段、首段和尾段等片断的局部相似度计算,以综合评价文本与用户模板的匹配情况。该方法可操作性强,效果明显。     

HowNet在文本挖掘中的应用

该文介绍HowNet在文本挖掘中的应用,利用HowNet从中文语义的角度计算中文词语相似度,计算词语之间的相关性,为实现更深层次的信息处理做准备。     

Visual input for pen-based computers

The design and implementation of a camera-based, human-computer interface for acquisition of handwriting is presented. The camera focuses on a standard sheet of paper and images a common pen; the trajectory of the tip of the pen is tracked and the contact with the paper is detected. The recovered trajectory is shown to have sufficient spatio-temporal resolution and accuracy to enable handwritten character recognition. More than 100 subjects have used the system and have provided a large and heterogeneous set of examples showing that the system is both convenient and accurate     

Recoding-derived bounds for input encoding

This paper uses theory and experiment to help explain why state assignment algorithms which use two-level-based cost measures often give good multi-level logic implementations. First, we develop theorems that give conditions under which an input encoding that results in multi-cube functions in the minimized Boolean network can be re-encoded to change the multi-cube functions into smaller functions to produce a smaller network. Second, we measure the properties of some typical finite-state machines to determine how well they fit the requirements of the theorems. The good fit between the requirements of the theorems and the properties of typical state machines helps explain why state assignment algorithms designed for two-level-logic implementations are relatively successful in designing state assignments for multi-level logic.     

Text simplification resources for Spanish

In this paper we present the development of a text simplification system for Spanish. Text simplification is the adaptation of a text for the special needs of certain groups of readers, such as language learners, people with cognitive difficulties, and elderly people, among others. There is a clear need for simplified texts, but manual production and adaptation of existing text is labour-intensive and costly. Automatic simplification is a field which attracts growing attention in Natural Language Processing, but, to the best of our knowledge, there are no existing simplification tools for Spanish. We present a corpus study which aims to identify the operations a text simplification system needs to carry out in order to produce an output similar to what human editors produce when they simplify news texts. We also present a first prototype for automatic simplification, which shows that the most important simplification operations can be successfully treated.     

中文文本分类器的设计

文本分类是指在给定分类体系下,根据文本的内容自动确定文本类型的过程。文章应用球形的k-均值算法确定每个文本的类标签,并通过Boosting算法构建分类器。构建的分类器具有以下特点:分类器的设计针对未知类标签的语料库,实用性好;分类器能随着语料库中文本的变化而增加新的类,具有很好的可扩展性;分类器基于Boosting算法,具有很好的分类精度。