In this paper the research techniques of complex network are introduced into the complement of missing data in text and a new method of text mining is put forward basing on the text structure of large-scale texts. First the GRE word net is constructed by using lots of relative articles specially for experiment, then the static characters of this network are analyzed and the context relationship...

Natural language texts frequently contain related information in different positions in discourse. As human readers, we can recognize such information across sentence boundaries and correctly infer relations between them. Given this inference capability, we understand texts that describe complex dependencies even if central aspects are not repeated in every sentence. In linguistics, certain omi...

In this paper we analyse the selectivity measure calculated from the complex network in the task of the automatic keyword extraction. Texts, collected from different web sources (portals, forums), are represented as directed and weighted co-occurrence complex networks of words. Words are nodes and links are established between two nodes if they are directly co-occurring within the sentence. We ...

syntax In order to analyze the above introduced phe nomena it is su cient to work on a restricted language fragment de ned in this section The usual way to formally de ne a fragment of the lan guage is to provide a grammar Since the consid ered phenomena occur in many natural languages almost every western language has the syntactic constructs necessary for expressing the previous utterances I ...

Although children's knowledge of the sound patterns of words has been a focus of debate for many years, little is known about the lexical representations very young children use in word recognition. In particular, researchers have questioned the degree of speci®city encoded in early lexical representations. The current study addressed this issue by presenting 18±23month-olds with object labels ...

This study examined how skilled Japanese readers activate semantic information when reading kanji compound words at both the lexical and sentence levels. Experiment 1 used a lexical decision task for two-kanji compound words and nonwords. When nonwords were composed of kanji that were semantically similar to the kanji of real words, reaction times were longer and error rates were higher than wh...

One of the methods to enhance bio-accessibility of drugs like Acridone and its oxime is toform inclusion complexes with β-cyclodextrin. The formation of such complexes has beenconfirmed by changes in spectral characteristics and melting point data. The aqueous phasesolubility studies reveal 1:1 stoichiometry between the compound and, β-cyclodextrin. The studyof thermodynamic parameters like ΔG,...