We describe the construction of two corpora in the domain of Swiss legal texts: The DS21 corpus is based on the Collection of Swiss Law Sources and contains historical legal texts from the early Middle Ages up to 1798; the Swiss Legislation Corpus (SLC) is based on the Classified Compilation of Swiss Federal Legislation and contains all current Swiss federal laws. The paper summarizes the key p...

This paper deals with a complex system of processing raw Czech texts. Several modules were implemented which perform different levels of processing. These modules can easily be incorporated into many other linguistic applications and some of them are already exploited in this way. The first level of processing raw texts represents a reliable morphological analysis – we give a survey of the effe...

Discourse markers foreshadow the message thrust of texts and saliently guide their rhetorical structure which are important for content filtering and text abstraction. This paper reports on efforts to automatically identify and classify discourse markers in Chinese texts using heuristic-based and corpus-based data-mining methods, as an integral part of automatic text summarization via rhetorica...

This paper considers improving the performance of text classification, when summaries of the texts, as well as the texts themselves, are available during learning. Summaries can be more accurately classified than texts, so the question is how to effectively use the summaries in learning. This paper proposes a new method for addressing the problem, using a technique referred to as ’stochastic ke...

Rather than using length-based or translation-based criterion to align bilingual texts, this paper proposes a part-of-speech-based (POS-based) criterion. The postulation is that bilingual texts should share the same concepts, ideas, entities and events. In addition, these are usually represented by some critical POSes. Thus, the numbers of critical POSes in a language pair of a bead are close. ...

Natural language is a common type of input for data processing systems. Therefore, it is often required to have a large testing data set of this type. In this context, the task to automatically generate natural language texts, which maintain the properties of real texts is desirable. However, current synthetic data generators do not capture natural language text data sufficiently. In this paper...

Determination and early detection of emerging trends can be retrieved from numeric data as well as from texts. Using texts for trend mining brings advances for the recognition process. The systematic integration of informaion descriptions and metadata schemes enable the additional semantic analysis of the available information. In this paper, we introduce the issue of trend recognition in infor...

In this paper, we propose a new text recognition model based on measuring the visual similarity of text and predicting the content of the unlabeled texts. First a Siamese network is trained with deep supervision on a labeled training dataset. This network projects texts into a similarity manifold. The Deeply Supervised Siamese network learns visual similarity of texts. Then a K-nearest neighbor...