1999 Volume 13 Issue 4
Published: 16 August 1999
-
Select all|
-
Review1999, 13(4): 2-7.Personal pronouns make up the core vocabulary of all natural languages , and whether they are processed properly is a matter of great significance in machine translation. By way of contrastive analysis , the thesis reveals the similarities and dissimilarities between the personal pronouns within English and Chinese. Aiming at a number of popularly found problems in some English - Chinese translation systems , the thesis puts forward some solutions.
-
Review1999, 13(4): 8-16.This paper briefly describes the text physical structure , logical structure and the vector space model of them . The system of text browsing is developed based on mechanism of information navigation. Text hierarchical analysis approach is put forward in text structure analysis instead of the approach of text relationship map. Each unit in text structure is given a mark information which consist of summary , keywords and weight , as a result , representations of text visualization are formed. Using the hypertext links among text , hierarchies and paragraphs , users can arbitrarily browse the text according to their interests.
-
Review1999, 13(4): 17-22.The paper uses the Log-Likelihood-Ratio-Test-Based feature words selection approach in the field of text categorization. In comparison with the traditional method , that is , each of the frequecy test , salience test and distributioness test is conducted independently , the proposed approach uses covariance matrix to coordinate the associations among the variant statistics so that all of them are integrated into a whole. The experiments show that the approach is superior to the t raditional approach.
-
Review1999, 13(4): 23-29.The main problems in Chinese page analysis are presented on the basis of the differences between Chinese and English layout . The Chinese layout comprehensive features are summarized with which a layout analysis method is built mainly based on a bottom-up approach. The results of experiment have shown that this method is able to analyse the standard Chinese layout . Compared with the existing approaches ,it is more efficient and suitable to process Chinese layout .
-
Review1999, 13(4): 30-35.In this paper , a post-processing method using HMM(Hidden Markov Model) for Chinese document recognition is proposed. HMM combines language model with single character recognition(SCR) model to make the best of SCR output . The parameters of language model are derived from corpus , while the parameters of SCR model are conditional probabilities that can be converted into posterior probabilities by theoretic analysis. On the basis of SCR output analysis , posterior probabilities of candidates are obtained by statistical method. Experiments in off - line Chinese document recognition show that post - processing performance depends on efficient evaluation of posterior probability , besides proper language model.
-
Review1999, 13(4): 36-41.In this paper , a hierarchical graph-element approach for recognition of neat handwritten Chinese characters (CCS) was proposed . We give the rules extracting these graph-elements and the subordinate degree algorithm . We proposed a hierarchical structure description method from outer-level to inner-lever to describe CCS. And a matching algorithm of character feature was presented which founded on the similarity , importance and stability of primitives. Then the result of testing on 800 high frequency CCS for non-specific-person manifest that the average recognition rate reaches 85%.
-
Review1999, 13(4): 42-49.This paper explores a new method for automatic recognixing characters and symbols in the engineering drawing scanning image which is based on the point correlation neural network technology. This method has taken into account the degenerated character samples ,and attempts to complete automatically the training process by a new learning rule ,to create an optimal feature representation of each character from a set of character sample image. The research makes the learning complexity constant ,and gets effective application in the engineering drawing character recognition.
-
Review1999, 13(4): 50-57.Full Text Retrieval is one of the most efficient techniques of information retrieving. This paper discusses the main techniques in designing a full text retrieval system based on Chinese Web document library , especially the detail of the approach of‘indexing of characters’. A practical system is introduced at the end.
-
Review1999, 13(4): 58-63.This paper mainly introduce how to constitute Mongolian automatic processing system. It includes the Mongolia phraseology library ,the Mongolian linguistry library ,the Mongolian data library ,and pilot study course of software on the Mongolian automatic processing system.
