Knjižnica Filozofskog fakulteta
Sveučilišta u Zagrebu
Faculty of Humanities and Social Sciences Institutional Repository

Automatsko sažimanje teksta


Downloads per month over past year

Cimperšak, Mislav. (2011). Automatsko sažimanje teksta. Diploma Thesis. Filozofski fakultet u Zagrebu, Department of Information Science. [mentor Boras, Damir and Ljubešić, Nikola].

[img] PDF (Croatian) - Registered users only
Download (1MB) | Request a copy


Automatic text summarization is a very attractive scientific area with connections to other areas such as information retrieval, natural language processing and machine learning. In spite many trials the problem of automatic text summarization is far from being solved. The process of summarization can be separated into three phases: text analyzing, reformatting the text into a condensed version of it, creating the wanted summarized form. The given work explains the History of automatic summarization methods alongside examples of systems using those methods. Processes used in newer systems are very similar to classic approaches, but always customized to specific systems and their goals. In areas such as information retrieval and machine translation evaluation is a key component in area’s advancement. In automatic text summarization the problem of evaluation is still not solved and thus the whole scientific area suffers from the problem of objective assessment of summarized texts and not knowing in which direction should the area advance. Automatic method ROUGE was explained and later used for the evaluation of CroWebSum system. CroWebSum was developed in 2010 on Faculty of humanities and social sciences, University of Zagreb. It is a tool intended for summarization of newspaper articles. System creates extracts which help the reader to decide whether to read the whole text or not. Summary is fully created from existing sentences which were selected as being the most important within the original text. System scores each sentence and creates extract from the ones which received highest scores. CroWebSum online was created in 2011 and it is used as a web interface for the original CroWebSum. CroWebSum online was built using Django framework and it is available on

Item Type: Diploma Thesis
Uncontrolled Keywords: automatic text summarization, summary evaluation, ROUGE, web application
Subjects: Information sciences > Social-humanistic informatics
Departments: Department of Information Science
Supervisor: Boras, Damir and Ljubešić, Nikola
Date Deposited: 11 Jun 2014 12:26
Last Modified: 10 Jul 2014 13:57

Actions (login required)

View Item View Item