Learning English Verb Selection Rules from Hand-made Rules and Translation Examples
I did not write a summary because the paper text is empty. Only the page markers <!-- с. 1 --> … <!-- с. 15 --> came through, with no content. The PDF is most likely a scan with no text layer, so it needs OCR. Another possible cause is a failure during text extraction.
From memory I can say only this, in general terms. The paper seems to come from NTT and to deal with Japanese–English machine translation. In it, rules for choosing the English verb are learned from translation examples and hand-written rules. This is probably done with decision-tree induction, using semantic categories from a thesaurus as features. This is uncertain, and I have not checked it against the text. I do not have the exact details (sample sizes, accuracy, system names), and they must not be made up for the library.
Please send the text after OCR or another source for the PDF. I will then prepare the summary in exactly the "## Summary / ## Key points" format.
Cite
@inproceedings{akba1995learning,
title = {{Learning English Verb Selection Rules from Hand-made Rules and Translation Examples}},
author = {Akba, Yasuhiro and Ishii, Megumi and Almuallim, Hussein and Kaneda, Shigeo},
booktitle = {Proceedings of the Sixth Conference on Theoretical and Methodological Issues in Machine Translation of Natural Languages},
year = {1995},
url = {https://aclanthology.org/1995.tmi-1.16/}
}This page in Markdown → https://md.research.elamur.ai/free/papers/akba-1995-learning-english-verb-selection-rules/