Using Elements Of Semantic Parsing In E-Learning Environments

The package provides flexible and simple mechanism for support interaction between HTTP server and data, which have a tabular presentation. 2.

The package allows a cross-platform data transfer. The formats of templates of the text documents and the formats of the tabular data do not depend on the choice of the operation system. It is also possible to transfer the processor of the text document templates between various operation systems.

The set of the resulting documents can be easily extended by mean ether changing the tabular data or extending the set of document templates. Some intermediate project were already implemented using the conceptual model of the I&R system and the software package those have been discussed above. These projects were published as small and middle WWW sites in Internet and Intranet networks.

A lot of researches are devoted to the possibilities of applying artificial intelligence tools in educational systems. The analysis shows that much attention is paid to semantic analysis. In particular there are many tools that use semantic analysis in automated assessment of student’s complete answer to some question. Investigations conducted by the Modeling and Software Department of Kryvyi Rih Technical University show that possibilities of semantic analysis application in educational environments are considerably wider. In particular at the stage of distance courses design using semantic parsing it is possible to estimate the correspondence of text materials to teaching aims, correspondence of test tasks to theoretical materials used. During the process of training the elements of semantic parsing can be used for estimation of richness of messages content. The present article is devoted to the realization of these and other possibilities.

The tools of key information extraction from the text and analysis of its semantic context lay in the foundation of the methods proposed by us. A lot of investigations (among them there are many works by Peter Turney [Turney, 1997;Turney, 2000]) have been devoted to key information extractions tools from the text, compiling of short summaries and annotations on the basis of this information. The elements of such investigations are widely implemented in searching web-systems, electronic libraries catalogue tolls, but almost no investigations have been conducted for implementations of similar tools in the distance learning systems. At the same time information exchange in the interactive modules of educational environments takes place using text information and its automatic analysis could considerably increase the efficiency of educational process.

Several tasks that can by solved using elements of semantic parsing have been distinguished in our investigations. Here are some of them.

Firstly, the distinguishing of the key information is in itself an important tool and can be widely used in the distance learning organization. The present method enables, for example, to automate the process of keywords distinguishing in educational texts and glossary building, the creating of concise context of each chapter and of the whole educational course. And the most important is that the method of key information distinguishing in the educational text is a basis for all other directions of our investigations and before considering the other possibilities of semantic parsing application we concentrated our efforts on the development and perfection of this very module.

Secondly, key information distinguishing in the educational text allows creating the analyzer of educational material correspondence to the set educational aims. This will allow the developer of a distance course, firstly, to pay attention to the aims to be achieved in the distance course. Strictly determined aims enable the correspondence of educational materials to the set aims to be traced automatically, to indicate to the completeness of one or another question disclosure and to conduct search for the necessary information in the proposed database even in automatic mode. A collection of text messages, files or World Wide Web resources can be such a database. Thus the semantic parsing becomes an indispensable assistant during designing and creating the distance course in the educational environment.

Thirdly, the conditions for correspondence analysis of test tasks to the aims of education and to the contents of theoretical materials are being created. There are some well-known facts when much attention has been paid to insignificant aspects considered in the theoretical material during composing the test tasks. This, of course, decreased the efficiency of testing and did not give the objective estimation of students’ knowledge. On the basis of the key information di