Conceptual Graphs as Framework for Summarizing Short Texts

Abstract

In this paper, a conceptual graph-based framework for summarizing short texts is proposed. A semantic representation is implemented through conceptual graph structures that consist of concepts and conceptual relations that stand for texts. To summarize conceptual graphs, the most important nodes are selected using a set of operations: generalization, association, ranking, and pruning, which are described. The importance of nodes on weighted conceptual graphs is measured using a modified version of HITS algorithm. In addition, some heuristic rules are used to keep coherent structures based on information from WordNet (hierarchy of concepts) and VerbNet (semantic patterns of verbs). The experimental results show that this approach is effective in summarizing short texts.

Towards Personalized Summaries in Spanish based on Learning Styles Theory

“…Current research proposes several and diverse methods for automatic text summarization such as statistical [22], machine learning [23,24], text connectivity [25,26], conceptual graphs [27,28,29], algebraic reduction [30], clustering and probabilistic models [31,32,33] and methods adapted to the reader [34,35].…”

“…Conceptual graphs are structures for knowledge representation based on first-order logic. Graphs are a natural, simple, and fine-grained semantic representation which can be used to describe texts [27].…”

Automatic Generation of Summary Obfuscation Corpus for Plagiarism Detection

“…is not a simple concatenation of sentences or excerpts from an original document; often, it is a complete paraphrasing of the text using several operations to abstract the text [15]. According to the results of the performance of systems, this approach presents great challenges to systems, when the text is a sort of plagiarized version of an author's ideas.…”

Geração de linguagem natural por meio de representações semânticas abstratas para o português do Brasil

“…However, its generation needs a deeper automatic analysis of the source text to determine what information is more important, how it should be organized, which sentences should be generated as a union (or separation) from others, and how they should be described since these summaries must accomplish linguistic requirements such as grammaticality and meaning similarity. Thus, the generation of natural language becomes extremely important to deal with this kind of As mentioned in the previous example, Data-to-Text approach tends to be important in applications such as abstractive summarization, in which intermediate semantic representations are used to represent the source text(s) and the abstract to be generated and, then, a text is generated from it (MIRANDA-JIMÉNEZ; GELBUKH; SIDOROV, 2014;LIU et al, 2015) 4 .…”