Shosaku Tanaka

Towards Generating Multiple-Choice Tests for Evaluating Comprehension of Arbitrary English Texts

We develop a method for generating multiple-choice test for evaluating comprehension of an arbitrary English text and its answer to support more extensive reading in English. Our test consists of one sentence (a correct optional sentence) that is consistent with the selected English text and several sentences (distractor optional sentences) that are inconsistent with this text. Learners select one sentence consistent with the text from multiple optional sentences on reading the text. In our proposed method, the system extracts several important sentences from the text, and replaces at least one word in each of these sentences with its synonym (if possible). One of these sentences is then selected as the correct response, while further changes to the polarities or nouns in the remaining sentences are carried out to generate distractor sentences for the multiple-choice test. Our method has potential to improve the learning effect associated with extensive reading.

Estimating satisfactoriness of selectional restriction from corpus without a thesaurus

A selectional restriction specifies what combinations of words are semantically valid in a particular syntactic construction. This is one of the basic and important pieces of knowledge in natural language processing and has been used for syntactic and word sense disambiguation. In the case of acquiring the selectional restriction for many combinations of words from a corpus, it is necessary to estimate whether or not a word combination that is not observed in the corpus satisfies the selectional restriction. This paper proposes a new method for estimating the degree of satisfaction of the selectional restriction for a word combination from a tagged corpus, based on the multiple regression model. The independent variables of this model correspond to modifiers. Unlike a conventional multiple regression analysis, the independent variables are also parameters to be learned. We experiment on estimating the degree of satisfaction of the selectional restriction for Japanese word combinations ⟨noun, postpositional-particle, verb⟩. The experimental results indicate that our method estimates the degree of satisfaction of a word combination not very well observed in the corpus, and that the accuracy of syntactic disambiguation using the co-occurrencies estimated by our method is higher than using co-occurrence probabilities smoothed by previous methods.