Uncorrected Author Proof Journal of Intelligent & Fuzzy Systems xx (20xx) x–xx DOI:10.3233/JIFS-179902 IOS Press 1 Determining the importance of sentence position for automatic text summarization 1 2 Griselda Areli Matias Mendoza, Yulia Ledeneva ∗ and Rene Arnulfo Garc´ ıa-Hern´ andez 3 Universidad Aut´ onoma del Estado de M´ exico, Unidad Acad´ emica Profesional Tianguistenco, Instituto Literario, Toluca, Edo. Mex, M´ exico 4 5 Abstract. The methods of Automatic Extractive Summarization (AES) uses the features of the sentences of the original text to extract the most important information that will be considered in summary. It is known that the first sentences of the text are more relevant than the rest of the text (this heuristic is called baseline), so the position of the sentence (in reverse order) is used to determine its relevance, which means that the last sentences have practically no possibility of being selected. In this paper, we present a way to soften the importance of sentences according to the position. The comprehensive tests were done on one of the best AES methods using the bag of words and n-grams models with the with DUC02 and DUC01 data sets to determine the importance of sentences. 6 7 8 9 10 11 12 Keywords: Automatic Text Summarization, n-gram Model, bag of words model, slope calculation, genetic algorithm 13 1. Introduction 14 Currently, information is exponentially growing 15 and thus, the necessary time available for process- 16 ing. Therefore, it is essential to have methods that 17 allow Automatic Extractive Summarization (AES). 18 The purpose of the methods AES is to generate sum- 19 maries more similar to those generated by the human. 20 Presently, summaries can be used in different areas. 21 There are employed to summarize information, for 22 example, for videos [1], newspapers [2–4], scientific 23 papers [5] and social networks as Twitter [6, 7] or 24 blog [8], where information rapidly changes and tech- 25 nologies are required to access real-time information 26 represented in reduced form. 27 According to Ladda Saunmali [9], the purpose 28 of the text summary is to present the most impor- 29 tant information in a shorter version of the original 30 text, maintaining its main content and helping the 31 ∗ Corresponding author. Yulia Ledeneva, Autonomous Univer- sity of the State of Mexico, Instituto Literario No. 100, CP 50000, Toluca, State of Mexico, Mexico. E-mail: yledeneva@yahoo.com. user to quickly understand the large volume of infor- 32 mation. According to Alfonseca, Berker, Da Cunha 33 Fanego among others [9–17], the summaries are clas- 34 sified according to their strategy of condensation in 35 abstractive and extractive summaries. The abstrac- 36 tive summaries are those summaries generated from 37 understanding the document and describe the content 38 with words or sentences that sometimes are not in the 39 original text. Instead, extractive summaries are gen- 40 erated from the selection of key phrases, sentences, 41 or paragraphs considered essential for the original 42 text; so, they do not require the understanding of the 43 document. 44 Among the methods proposed for AES are those 45 that need a large number of language resources 46 [18–23], so they have a high dependence on lan- 47 guage or require sophisticated processes to generate 48 a summary. There are also methods that only use the 49 structure and distribution of the original text, so they 50 are less dependent on language [2, 3, 24–29]. The 51 language-dependent methods may show better results 52 than language-independent ones. However, research 53 in language-independent methods has grown because 54 ISSN 1064-1246/20/$35.00 © 2020 – IOS Press and the authors. All rights reserved