Ijaz summary system
The single-document and multi-document summarizing system project of Ijaz was commissioned by the Information Technology Organization of Iran and was carried out by the Web Technology Laboratory of Ferdowsi University of Mashhad. In this big project, a set of tools needed to produce summarizing systems was produced. Also, the web version of the single-document and multi-document summary generator was also produced, which can be seen on the main page of the site. This system has the ability to produce summaries for Persian and English single-document and multi-document texts. Various criteria have been used to produce this system.
Project LinkMore Details
Also, for the first time in the country, a large corpus of Persian summarization was produced for the evaluation of summarizing systems using the necessary standards and spending more than 2000 person-hours of time. The “Response” format (the standard format of summarizing systems) is presented in two single-document and multi-document models. The body of a single document contains 100 different topics of various types of news, which have been selected from the most popular news agencies in Iran. Each of these topics has 5 abstracts and extracts produced by trained experts. The multi-document body “Answer” also includes 50 topics, each topic contains 20 documents, and each topic includes 5 human summaries and abstracts.
Also, for the first time in the country, an evaluation tool for summarizing systems was also produced. This tool is able to evaluate summarization systems by using various criteria and using human summaries produced in the “answer” body. This tool can be downloaded in the “site tools” section. Other tools have been developed for natural language preprocessing, which can be downloaded.
Research Field
NLP
Implementation Date
2012
Project Members
- Automatic summarization of multiple documents based on concept extraction, Asef Pour Masoumi, 2011-09-21, Master’s Thesis
- A new method of semantic weighting of words in text processing applications, Hossein Kamiyar, 2011-09-21, Master’s Thesis
- Abstract summarization based on the similarity of sentences, Fatemeh Pourgholam Ali, 2010-02-20, Master’s Thesis
- Semanticism in the automatic evaluation of English and Persian machine summarizers using the vocabulary network, Ahmad Estiri, 2012-09-21, Master’s Thesis