Download Advances in Natural Language Processing: 7th International by Hrafn Loftsson, Eirikur Rögnvaldsson, Sigrun Helgadottir PDF

By Hrafn Loftsson, Eirikur Rögnvaldsson, Sigrun Helgadottir

This e-book constitutes the lawsuits of the seventh overseas convention on Advances in normal Language Processing held in Reykjavik, Iceland, in August 2010.

Show description

Read or Download Advances in Natural Language Processing: 7th International Conference on NLP, IceTAL 2010, Reykjavik, Iceland, August 16-18, 2010, Proceedings PDF

Similar compilers books

Visual Language Theory

Kim Marriott Bernd Meyer conversation is without doubt one of the hallmarks of people. once we give some thought to hu­ guy communique, most folks first contemplate spoken and written lan­ guages. those are related in that symbols within the language are encountered and processed sequentially, both temporally as they're spoken or as char­ acters are learn throughout a web page.

Construction and Evolution of Code Generators: A Model-Driven and Service-Oriented Approach

Computerized code new release is a necessary cornerstone of model-driven ways to software program improvement. at present, plenty of innovations can be found that help the specification and implementation of code turbines, corresponding to engines in keeping with templates or rule-based differences. All these ideas have in universal that code turbines are both without delay programmed or defined through textual requirements.

Additional resources for Advances in Natural Language Processing: 7th International Conference on NLP, IceTAL 2010, Reykjavik, Iceland, August 16-18, 2010, Proceedings

Sample text

We must not forget that experts marked the blanks without being instructed. Therefore the blanks did not have to correspond with words related to the specific domain. 6 Conclusions and Future Work The article presents a study about automatic distractor generation for domain specific texts. The system implements different methods to find distractors semantically similar to the key. It uses context-words to compute the similarity deploying LSA. We have used a balanced part of a specialised corpus to build the model.

All of them tend to obtain better results with the second text and tend to create easy items. The results of item discrimination take into account the responses of the high-scoring and low-scoring students. The high-scoring group is the top 1/3 of the class, and the low-scoring group comprises students with test scores in the bottom 1/3 of the class. Regarding item discrimination, the corrected pointbiserial index with its standard deviation as well as the percentage of items with negative values are shown in the table.

It uses context-words to compute the similarity deploying LSA. We have used a balanced part of a specialised corpus to build the model. In the near future we will make use of the whole specialised corpus to model it. In this approach, we have explored item difficulty, item discrimination and distractors’ evaluation based on Classical Test Theory. The results shown for the manually generated test were promising, and there were not significant differences among the methods. The item discrimination measure led us to study the option-by-option responses of high-scored and low-scored groups and we finished the study with the evaluation of the distractors.

Download PDF sample

Rated 4.07 of 5 – based on 34 votes