Semantic language and multi-language MT approach based on SL

Any natural language is regarded as a representation of semantic language.The translation between two languages (I,J) is regarded as a transformation between two representations.A natural language-Ⅰ is translated into another natural language-J only by two steps.One is semantic-analysis of language-Ⅰ based on “semantic-element-representation-base of language-I“,the other is deploying into the representation of language-J based on “semantic-element-representation-base of language-J‘.For translating in N natural languages,it is needed to develop N translation systems only,rather than N（N-1）/2,or 2N systems.     

论智能的级——向二级智能进军!

本文在人工"智能"的含义下,提出智能算法及其级的新概念,指出二级智能的算法的直接运行结果是人们事先不掌握的,甚至有些是人们无法掌握的算法.再运行这些算法(一般地说,其运行过程是人们所不掌握的)才能得到人们所要的结果,最后给出二级智能的一些例子.     

Sloping-and-shaking——Multiway merging and sorting

Most traditional merging and merging-based sorting algorithms are based on 2 sorters or 2 comparators A new merging technique is developed, namely sloping-and-shaking multiway merging, and a corresponding mul-tiway sorting method based only on k-sorters is proposed The sloping-and-shaking merging algorithm merges k sorted lists into one, where k can be any prime number The merging process is not a series of recursive applications of 2-way morging It sorts the keys on the m × k plane in vertical and horizontal directions, then along sloping lines with various slope rates step by step Only k-sorters are needed in the merging or sorting process. The time needed to merge ksorted lists, with m of each, is ( k + [log2( m / k) ]) tk, and the time for sorting N keys is (1 + (p - 1) k + 1/2( p -1) (p - 2)[ log2k])tk, where p - logkN, and tk is the time to sort k keys. The proposed algorithms can be implemented either by hardwared sorting networks, or on general purpose parallel and vector machines The tradition     

基于句对比较的自动获取翻译模板方法改进

翻译模板不仅是机器翻译的重要资源,而且是有用的语言学知识。使用词对齐结果改进了从实例中类比学习翻译模板的方法,并提出了带有函数关系的模板（TTFR）以获取语言之间的依赖关系,如“he…his”对。带函数关系的模板是一个更一般化的模板并能直接用于双向的机器翻译。实验结果表明改进的方法有效地克服类比方法的问题并且提高了自动获取模板的准确率。     

基于语义分析提高脱机手写体识别率的方法

论文讨论了一种通过自然语言语义层次的理解来协助提高脱机手写体识别率的方法。从识别的过程来看,这种方法是一种基于上下文关系的识别“后处理”方法。文中介绍了语义单元、语义单元表示、语义单元表示树、语义单元表示集、语义单元表示树集,并且具体讨论了通过语义单元表示树集的剪枝来提高识别正确率的方法。     

k-Bitonic sort

A k-bitonic sort which generalizes the bitonic sort is proposed. The theorem of the bitonic sort, which merges two monotonic sequences into one order sequence, is extended into the theorem of k-bitonic sort. The k-bitonic sort merges K (=2k or 2k-1) monotonic sequences into one order sequence in steps, where k=[K/2] is an integer and k≥1. The k-bitonic sort is the Batcher's bitonic sort when k=1.     

An efficient multiway merging algorithm

Ｐａｒａｌｌｅｌｓｏｒｔｉｎｇｈａｓｂｅｅｎｓｔｕｄｉｅｄｆｏｒｄｅｃａｄｅｓ[1—10].Ｍｅｒｇｅ＿ｓｏｒｔｉｎｇｉｓｏｎｅｏｆｔｈｅｍｏｓｔｐｏｐｕｌａｒｓｏｒｔｉｎｇａｌｇｏｒｉｔｈｍｓ.Ｈｏｗｅｖｅｒ,ｍｏｓｔｐａｒａｌｌｅｌｍｅｒｇｅ＿ｓｏｒｔｉｎｇａｌｇｏｒｉｔｈｍｓａｒｅｂａｓｅｄｏｎ2＿ｗａｙｍｅｒｇｉｎｇ,ｓｕｃｈａｓｏｄｄ＿ｅｖｅｎｍｅｒｇｅｎｅｔｗｏｒｋ[3]ａｎｄｂｉｔｏｎｉｃｍｅｒｇｅｎｅｔｗｏｒｋ[4].Ｌｅｉｇｈｔｏｎ[10]ｇａｖｅｔｉｇｈｔｂｏｕｎｄｓｆｏｒｔｈｅｃｏｍｐｌｅｘｉ…     

一个带有可变结构总线的常数排序处理机阵列

本文提出一个具有N1+ε台处理机和带有N1+ε个1-位(二进制位)的可变结构的总线系统的SIMD阵列处理机系统.其中,ε是任意小的正常数.它具有常数步排序能力.先前具有常数步排序能力的系统[7]其处理机台数为N3,所带可变结构总线为N2个s-二进制位的可变结构总线,其中,s=log2N.     

统一语言学中的类几何变换函数

在统一语言学语义单元理论的基础上,从类几何变换的角度对语言所蕴涵的语义知识进行分析和描述,指出不同的自然语言之间的翻译实质上是求解同一语义在不同语言上的不同表示之间的变换.以疑问句的疑问目的变换为例讨论了统一语言学中的类几何变换问题.并将疑问句根据对象和类型重新进行二维分类,同时定义六元素形式系统的变换描述语言TDL描述其可逆变换函数及双向变换算法.最后讨论了该变换函数在多语互译机器翻译系统中的应用.     

多值逻辑中的命题相关性与逻辑运算研究

命题的属性包括结构属性和值属性.命题的结构决定了命题之间的关系,决定了命题之间的逻辑运算.命题的真值只是一个由命题的结构决定的值属性,并不能代表整个命题.逻辑运算是命题的运算,不是真值的运算.多值逻辑中,命题逻辑运算结果由命题的关系决定,真值相同的不同命题,逻辑运算结果的真值不一定相同,逻辑运算不是处处同态于某一个或某一簇真值函数(算子),有时复合命题的真值不能被它的成分命题的真值完全确定,所以多值逻辑的联结词并不总能定义成真值函数(算子)的形式.多值逻辑的命题公式不能再看作真值函数,命题公式是关于命题的函数.