Automatic abstract experimental apparatus and method

Abstract

The invention relates to an automatic summarization experimental facility and a method thereof. The steps of the method comprise: step 1, performing pretreatment to a summarization object data to obtain summarization data, step 2, generating an automatic summarization by invoking a specified summarization algorithm, step 3, utilizing a preset standard summarization and a specified evaluation tool to evaluate the automatic summarization, and returning to an evaluation result of the summarization property, step 4, optimizing the property of the specified summarization algorithm according to the evaluation result of the summarization property. The invention resolves the problem of cross-language seamless transition of the tabloid algorithm and the evaluation tool during the summarization experiment process, realizes the automation of the parameter training process of the summarization algorithm and a summarization experiment, enables technologists to be capable of allocating the summarization algorithm and a parameter dereferencing allocation file by themselves, and automatically validates the property of the algorithm and learns the parameter combination.

Description

technical field [0001] The invention relates to the field of information processing, in particular to an automatic abstract experiment device and method. Background technique [0002] With the progress of the times and the development of the economy, people's demand for information in daily life is increasing, especially with the increasing popularity of the Internet, a large amount of information is released and disseminated on the Internet every day. Taking the development of the Internet in China as an example, according to CNNIC statistics, the number of Internet users in China has increased rapidly from 1.175 million at the end of June 1998 to 253 million at the end of June 2008, and the number of WWW sites has also increased from about 3,700 to 1.919 million. In just 10 years, the number of users has increased by more than 200 times, and the number of WWW sites has increased by more than 500 times. What followed was the rapid growth of Web information resources. As o...

AI Technical Summary

Problems solved by technology

[0004]However, since the current summarization algorithms are mostly implemented in languages ​​such as C / C++ or Matlab, and the existing evaluation tools are all implemented in Linux based on Perl language, At the same time, the interaction between algorithms and evaluation tools is very poor, so that the abstract experiment process often requires a lot of time and labor costs to generate and evaluate automatic abstracts. At the same time, it is necessary to manually select and adjust abstract parameters, which leads to the failure of abstract research. Inefficiency

Images