SAB(IO): A BIOLOGICALLY PLAUSIBLE CONNECTIONIST APPROACH TO AUTOMATIC TEXT SUMMARIZATION

An implementation of a computational tool to generate new summaries from new source texts is presented, by means of the connectionist approach (artificial neural networks). Among other contributions that this work intends to bring to natural language processing research, the use of a more biologically plausible connectionist architecture and training for automatic summarization is emphasized. The choice relies on the expectation that it may bring an increase in computational efficiency when compared to the sa-called biologically implausible algorithms.

IMAGE, MAGIC AND IMAGINATION: CHALLENGES TO THE ANTHROPOLOGICAL TEXT

What different forms of engagement do image and text allow the spectator/reader? We know that text and image communicate, and that all communication depends on a relationship between those who communicate. The objective of this text is therefore to understand the new possibilities available to an anthropology of the expression of knowledge that makes use of images, such as photographs and films.

Children`s oral production of narratives and constructive memory using prior reading of pictures

With the purpose of approximating two issues, oral narrative and constructive memory, we assume that children, as well as adults, have a constructive memory. Accordingly, researchers of the constructive memory share with piagetians the vision that memory is an applied cognition. Under this perspective, understanding and coding into memory constitute a process which is considered similar to the piagetian assimilation of building an internal conceptual representation of the information (hence the term constructive memory. The objective of this study is to examine and illustrate, through examples drawn from a research about oral narrative with 5, 8 and 10 years old children, the extent to which the constructive memory is stimulated by the acquisition of the structures of knowledge or ""mental models"" (schemes of stories and scenes, scripts), and if they automatically employ them to process constructively the information in storage and rebuild them in the recovery. A sequence of five pictures from a book without text was transformed into computerized program, and the pictures were thus presented to the children. The story focuses on a misunderstanding of two characters on a different assessment about a key event. In data collection, the demands of memory were preserved, since children narrate their stories when the images were no longer viewed on the computer screen. Each narrative was produced as a monologue. The results show that this story can be told either in a descriptive level or in a more elaborated level, where intentions and beliefs are attributed to the characters. Although this study allows an assessment of the development of children`s capabilities (both cognitive and linguistic) to narrate a story, there are for sure other issues that could be exploited.

Book illustration, as (intersemiotic) translation: Pictures translating words

This article examines book illustrations through the prism of Translation Studies. It mainly suggests that the pictures in illustrated books are (intersemiotic) translations of the text and that, as such, they can be analyzed making use of the same tools applied to verbal interlingual translation. The first section deals with the theoretical bases upon which illustrations can be regarded as translations, concentrating on theories of re-creation, as illustration is viewed essentially as the re-creation of the text in visual form. One of the claims in this section is that, because illustration is carried out in very similar ways as interlingual translation itself, the term ""intersemiotic"" relates more to the (obvious) difference of medium. For this reason the word is most often referred to in parentheses. The second section discusses three particular ways through which illustrations can translate the text, namely, by reproducing the textual elements literally in the picture, by emphasizing a specific narrative element, and by adapting the pictures to a certain ideology or artistic trend. The example illustrations are extracted from different. kinds of publication and media, ranging from Virgil`s Aeneid, Lewis Carroll`s Alice in Wonderland and Mark Twain`s Adventures of Huckleberry Finn to an online comic version of Shakespeare`s Hamlet.

Improvements on the Porter`s Stemming Algorithm for Portuguese

The amount of textual information digitally stored is growing every day. However, our capability of processing and analyzing that information is not growing at the same pace. To overcome this limitation, it is important to develop semiautomatic processes to extract relevant knowledge from textual information, such as the text mining process. One of the main and most expensive stages of the text mining process is the text pre-processing stage, where the unstructured text should be transformed to structured format such as an attribute-value table. The stemming process, i.e. linguistics normalization, is usually used to find the attributes of this table. However, the stemming process is strongly dependent on the language in which the original textual information is given. Furthermore, for most languages, the stemming algorithms proposed in the literature are computationally expensive. In this work, several improvements of the well know Porter stemming algorithm for the Portuguese language, which explore the characteristics of this language, are proposed. Experimental results show that the proposed algorithm executes in far less time without affecting the quality of the generated stems.

A complex network approach to text summarization

Automatic summarization of texts is now crucial for several information retrieval tasks owing to the huge amount of information available in digital media, which has increased the demand for simple, language-independent extractive summarization strategies. In this paper, we employ concepts and metrics of complex networks to select sentences for an extractive summary. The graph or network representing one piece of text consists of nodes corresponding to sentences, while edges connect sentences that share common meaningful nouns. Because various metrics could be used, we developed a set of 14 summarizers, generically referred to as CN-Summ, employing network concepts such as node degree, length of shortest paths, d-rings and k-cores. An additional summarizer was created which selects the highest ranked sentences in the 14 systems, as in a voting system. When applied to a corpus of Brazilian Portuguese texts, some CN-Summ versions performed better than summarizers that do not employ deep linguistic knowledge, with results comparable to state-of-the-art summarizers based on expensive linguistic resources. The use of complex networks to represent texts appears therefore as suitable for automatic summarization, consistent with the belief that the metrics of such networks may capture important text features. (c) 2008 Elsevier Inc. All rights reserved.