English Generation From Interlingua by Example-Based Method

ENGLISH GENERATION FROM INTERLINGUA BY EXAMPI~E-I~ASI~I) METHOI) Abstract This paper describes the experiment of the English generation from interlingua by the example-based method. The generator is implemented by using English Word Dictionary ,and Concept Dictionary developed in EDR. How to construct examples lind how to define the similarities are main problems. The results of experi- ments are shown. 1. Introduction Eiji Komatsu*, Jin Cui**, lliroshi Yasuhara** (*) Oki Electric htdustry Co. Ltd. Meltimedia I.aboratory 11-22, Shibaura 4-Chome, Minato-ku, Tokyo 108 Japan e-mail : komatsu @okilab.oki.co.jp (**) Japan Electronic l)ictionary Research Institute l.td. (EDR) 6th Laboratory Mita-Kokusai-Bldg. 4-28, Mira l-Chome, Minato-kv, Tokyo 108 Japan e-mail : sai@edrrr.edr.co.jp, yasuhara@edr6r.edr.co.jp to mean example data of the example-based method. And the terms "interlingua" and "syntactic tree" are used to mean sets, elements m~d fragments of elentents. 2. Input anti Output The generator t,anslates an interlingt, a to a syntactic tree. Fig.2.1 shows a sample of input interlingnae and Fig.2.2, a sample of output syntactic trees. Both samples correspond to the same sentence "My brother will take the medicine". This paper describes the generator that is originally implemented to correct and evah,ate English Word Dic- tionary and Concept Dictionary being developed in EDR (El)R,1993). To evaluate Concept Dictionary, as the first strategy, interlingua method was introduced. As the num- ber o1' concepts is very large and they are elements of complex hierarchy, it is difficult to make roles and on the other hand the example-based method was expected to be more effective than the rule-based method. So, as the sec- ond strategy, the example-bused method was also intro- duced. The example-based method is usually used in MT by the transfer method (Nagao, 1984; Sato, 1991; Stnnita, 1992), though one by Sadler (1989) is by the interlingua method. In this generator, the example-based method co- exists with the interlingua method because of above rea- sons, but the combination of the example-based method and tim interlingua method is not intportant, because l'rom another point of view, the generation from interlingua is recognized as a translation from one hm- guage i.e. interlingua to another i.e. English and the gen- eration from interlingua can be seen similar as transla- tions in above MT systems. So in this experiment, how to apply the example-based method to various natural hm- guage processing and lbr which parts the method are suitable are the main interests. For this purpose, the gen- erator is designed to execute the generation with maxi- mum usage of the example-based method. In this experiment, the coverage of the generation is not complete, that is, some elements st, ch as articles and conjunctions are not generated. Below, section 2 describes the input and ot, tput of the generator, section 3, examples used in this system, sec- tion 4, the similarities used to retrieve examples and to select words, section 5, the generation algorithm, section 6, the experiments for verb selections and section 7, the conclusion. The examples, similarities and the generation algo- rithm are decided a priori then modilied …

Content truncated to 500 words