Nová automatická morfologická analýza češtiny

Author(s): Klára Osolsobě,Jaroslava Hlaváčová,Vladimír Petkevič,Martin Svášek,Josef Šimandl / Language(s): Czech Issue: 4/2017

A detailed morphological description of word forms in any language is one of the necessary conditions for the successful automatic processing of linguistic data. The aim of this paper is to present a project aimed at a new description of Czech morphology, especially the planned changes in the tagset. The key changes are as follows:1) the unambiguous description of variants;2) the concept of a multiple lemma;3) the revision of part-of-speech definitions.

Příčestí minulé činné na -(v)ší v dnešních českých publicistických textech

Author(s): Markus Giger / Language(s): Czech Issue: 2/2010

Among the results of Russian influence on Czech in the 19th century was the emergence of an active past participle in -(v)ší in Czech. Although not welcomed by all grammarians, this participle continued its existence in Czech until today, becoming mainly a device of archaic and bookish style. In the actual work, the occurence oft the active past participle in -(v)ší in the largest partial corpus of the Czech National Corpus containing journalistic texts is studied. A main result of the study is that apart from a large number of examples from different verbs which show the active past participle on -(v)ší in the studied corpus once or twice and where it is indeed a device of archaic and bookish style, sometimes even of irony and humor, there is a small group of (mainly intransitive) verbs, where this participle functions with considerable frequency in stylistically more neutral contexts of written Standard Czech as the only participle (sometimes as a – stylistically more marked – variant of a more numerous active past participle in -l). In theses cases, it remains overwhelmingly a syntactically unextended direct attribute of a noun. Such active past participle in -(v)ší is to be found most often in sports coverage where it is built from a set of verbs with terminological function.

Niedrigfrequente grammatische Phänomene als sprachliche Zweifelsfälle

Author(s): Marek Konopka / Language(s): German Issue: 2/2010

Some grammatical phenomena that only seldom appear in the corpora of written language often coincide with speakers’ uncertainty about a given form’s grammatical status. Such display of uncertainty is often subject to prescriptive criticism, which pays little attention to actual usage. However, thorough and discriminating corpus analyses can help in a proper description of various low-frequency phenomena and in situating them more adequately in the grammatical system, against the background of different contexts, communicative situations, and language varieties. To exemplify this potential, this study examines three linguistic phenomena in German, using a corpus--based approach: the dative singular ending -e, the construction aus aller Herren Länder, which lacks the dative plural ending -n, and the non-standard preterite form frug. The results can be seen as a contribution to a more precise grammatical description on the one hand and, on the other, as a basis for an improved, more usage-oriented approach in providing practical advice to language users.

Předložky v(e), na a jejich konkurence ve statických prostorově-relačních významech

Author(s): Radek Ocelák / Language(s): Czech Issue: 2/2010

The article provides a classification of the static spatial relations expressed by the prepositions v(e) and na (in most contexts translated as in and on) in contemporary Czech, based on the data of The Czech National Corpus and the internet. Both prepositions, used with a nominal phrase in the locative, may refer to several distinct relations between the determined and the determining spatial object (usually, but not necessarily, various types of localization). For each of the prepositions, these relations are classified with special attention to possible variation between the two (and some other) prepositions in describing the same situation. Such detailed analysis of prepositional meanings and variation seems useful e. g. in advanced learning of Czech as a foreign language. The description of the static spatial meanings of the prepositions is a proper basis for describing their other spatial (dynamic) as well as non-spatial (ab-stract) meanings.

Slovo editora

Author(s): František Štícha / Language(s): English,Czech Issue: 1/2010

Dydaktyka przekładu pisemnego w dobie nowoczesnych technologii tłumaczeniowych

Author(s): Marcin Trojszczak / Language(s): Polish Issue: 3/2019

The present paper aims to discuss the influence of modern translation technologies such as online dictionaries, bilingual corpora, and machine translation software on translation didactics and training. More specifically, the paper discusses how the introduction of smartphones, computers, and machine translation software to the traditional translation classroom can impact the basic interactions between teachers and students, in particular with regard to their knowledge asymmetry. Moreover, it addresses the issue of ecological resemblance of traditional classroom augmented with modern translation technologies with respect to the work mode of contemporary commercial business and technical translators. Besides discussing the impact of technological factors on teacher-student interactions and ecological resemblance, the paper also touches upon a revolutionary potential of machine translation software whose improving quality coupled with its extended use in the classroom as a source of raw translations could lead to reshaping traditional translation classes into post-editing classes.

АВТОМАТИЧЕН МОРФОЛОГИЧЕН АНАЛИЗ. ЕЛЕКТРОНЕН МОРФОЛОГИЧЕН РЕЧНИК

Author(s): Elisaveta Balabanova / Language(s): Bulgarian Issue: 2/2020

The paper deals with the process of automatic morphological analysis. The issues of morphology and the operations in it are briefly discussed. The morphological dictionary in electronic form is discussed and also the process of automatic morphological analysis in NLP. The problems, encountered when doing automatic morphological analysis, are reviewed. The processes of automatic normalization and lemmatization are also discussed briefly.

Developing Online Czech Proofreader Tool: Achievements, Limitations and Pitfalls

Author(s): Dana Hlaváčková,Hana Žižková,Klára Dvořáková,Markéta Pravdová / Language(s): English Issue: 1/2022

This paper deals with achievements, limitations and pitfalls of developing the Online Czech Proofreader Tool (OCPT). The Tool has been developed in cooperation with the Department of Czech language of the Faculty of Arts of Masaryk University, the Institute of Theoretical and Computational Linguistics of the Faculty of Arts of Charles University, the Czech Language Institute of the Czech Academy of Sciences and Seznam.cz since 2019. The article describes the linguistic data used and tools and modules that constitute the OCPT and indicates the limitations of using an online web-based proofreader tool, especially in areas where mere application of formal rules for language error detection is not sufficient. The article also brings up the drawbacks of developing the OCPT which include occurrence of false-positives.

Reception of literature generated by artificial neural networks

Author(s): Zuzana Husárová,Karel Piorecký / Language(s): English Issue: 1/2022

This article introduces the phenomenon of literature generated by artificial neural networks, with specific examples of texts created in the Czech and Slovak cultural environment. It follows the historical background connected with generative and combinatory poetics and later describes the principles of data processing used by neural networks work; it also presents the parameters of their machine learning. The focus lies on the reception of these artificial texts in the media and in literary studies, leading to the proposition of two reading types specialized for these works: “reading of artificiality” and “literary metareading”, while rehabilitating Mathauser’s term of “metahability”. The study concludes by suggesting “literary metareading of artificiality” as a term that would combine the aforementioned approaches into a new reception of neural network literature.

ROY YOUDALE: Using Computers in the Translation of Literary Style: Challenges and Opportunities

Author(s): Marián Kabát / Language(s): English Issue: 1/2022

Review of: MARIÁN KABÁT - ROY YOUDALE: Using Computers in the Translation of Literary Style: Challenges and Opportunities New York and London: Routledge, 2020, 259 pp. ISBN 978-0-429-03034-5 DOI: https://doi.org/10.4324/9780429030345

THE FEASIBILITY OF ARTIFICIAL INTELLIGENCE IN EMULATING HUMAN BEHAVIOR: AN ANALYSIS

Author(s): Andrei Zamfira / Language(s): English Issue: 01/2020

The main purpose of this paper is to conduct a study and make analysis regarding the techniques, technologies and systems that have been evolved in the recent decades as branches of the broadest domain, which is Artificial Intelligence (AI) having as main objective to emulate the similar processes found in nature at living beings, especially those human cognitive. We will survey techniques such as pattern matching, learning and evolving from stimulus that make computer programs behave like humans. We'll state what the Turing test is and who was the first computer program to pass it, ELIZA, that relied mainly on pattern matching. This is only the beginning of our research in the domain previously stated, and we will focus here on natural language processing, the grammatical constraints that determine how natural language is generated and understood. The desire to explain language comes from ancient world, but not even today, after centuries of thought, a complete understanding of language seems to be a long way off. We'll see that engaging a computer program into a dialog is a very complex process that requires multiple types of knowledge from the following categories: phonetics and phonology, morphology, syntax, (compositional) semantics, pragmatics and politeness (discourse). The grammatical constraints that determine the understanding and generation of natural language phrases are expressed using a technique called X-bar Schema, which breaks a phrase into groups based on their word types (noun, verb, adjective, preposition, pronoun etc.) and represents them as a tree structure composed of maximal projection nodes and intermediate nodes. We will show what are the rules of this type of representation in order for the phrase to remain correct and give concrete examples of trees representations. We proposed to connect our examples with artificial agents from three Hollywood-ian masterpieces , that are HAL from 2001: A Space Odyssey, Joi from Blade Runner 2049, Ava from Ex. Machina, and many more, all these with the goal to emphasize what does it mean for an intelligent computer program to imitate human behavior that we do daily without being aware of the complexity of processes it involves.

Booking.com: A Comparative Study From A Website Localization Perspective

Author(s): Cristian Lako / Language(s): English Issue: 17/2014

The paper analyzes the level of localization of two booking.com hotel pages. The study focuses on applying website localization theories from a theoretical perspective. The languages/cultures considered are American English, British English and Romanian.

Semasiology and Onomasiology in Web Content Localization

Author(s): Cristian Lako / Language(s): English Issue: 18/2015

The article examines the validity of employing methods from linguistics to web content localization. The two proposed approaches, elicit two distinctive, yet compatible, methodologies that can be successfully applied to web content translation.

On The (Benefic) Effects of Evolution Upon Translation

Author(s): Bianca-Oana Han / Language(s): English Issue: 19/2015

The article deals with the idea that a translator should become aware of and benefit from the technology-based tools that aim at supporting his work and not to annul it, since these software programmes are such designed as to provide valuable assistance to the translator.

AJALOOLISTE TEKSTIDE NORMALISEERIMINE

Author(s): Gerth Jaanimäe / Language(s): Estonian Issue: 17/2021

Normalizing historical texts or in other words converting them to modern spelling enables us to analyze them with tools designed for contemporary language. It also makes it possible to search the texts for different keywords and automatically compare the old spelling to contemporary spelling. This article gives a general overview of normalizing, different methods, previously performed experiments and the main problems in the context of the old Estonian texts from the second half of the 19th century.

QUOTE EXTRACTION FROM ESTONIAN MEDIA: ANALYSIS AND TOOLS

Author(s): Dage Särg,Karmen Kink,Karl-Oskar Masing / Language(s): English Issue: 17/2021

This paper describes the identification, adaptation and creation of tools that are needed for creating a quote extractor for Estonian media texts that would be able to properly extract both direct and indirect quotes and attribute them to the correct person identified by full name and profession. This includes named entity recognition and resolution as well as grammar-based extraction of direct and indirect quotes. To get a further understanding of indirect speech in Estonian media, we also performed a corpus linguistic analysis of the quotes extracted with our tools from one week of Estonian news.

NEW PATTERNS AND APPROACHES TO COMPARATIVE LINGUISTICS

Author(s): Violeta Negrea / Language(s): English Issue: 1/2016

The article focuses on the practical relationship between the historical analysis of language facts and other scientific domains. References to the historical evolution of the name of the field are followed by the major steps that it took to shape new questions that skewed its historiography. In the third part the spotlight falls on the new paths by which comparative linguistics widens its field and associates the historical study of languages to other domains, making place to brand-new reinterpretation and recontextualization of old information through advanced processes of reasoning that reveal different objects of inquiry. The new research paradigms yield insights into the development stages of culture and civilization phenomena and exploit hypothesis of their future evolution.

Using Computational Linguistics to Analyse Main Research Directions in Economy of Regions

Author(s): Ivan Valeryevich Savin,Nikita Sergeevich Teplyakov / Language(s): English Issue: 2/2022

Over the past decades, the process of knowledge generation has accelerated, producing a lot of scientific publications, which makes reviewing even a relatively narrow subject area very demanding, if not impossible. However, recent text data mining tools can assist researchers in conducting such analysis in an objective and time -efficient way. We conduct such a literature review on 1307 articles published in the journal Economy of Regions from 2010 to 2021 using advanced topic modelling techniques. This analysis aims to describe the main research areas in the journal over time, the dynamics of their popularity and the relationship with key quantitative indicators. We identified 22 topics ranging from "Agriculture" and "Economic Geography" to "Fiscal Policy" and "Entrepreneurship". We estimate how popularity of these topics was changing over time and find topics that gained the most popularity from 2010 to 2021 (+17,61 %, "Spatial Economics") or lost it (-14, 58 %, "Economics of Innovation"). The topic of enviromental economic collects the largest number of citations per article (3.64, on average), and the topics on monetary policy and poverty are the most popular among manuscripts in English, which also true for articles written by authors with foreign affiliation. Papers with third-party funding are concentrated the most in "Spatial Economics" (around 11 %), and the least - in "Agriculture". Our results can help to understand the evolution in scope of research of Economy of Regions and serve researchers to find promising directions for future studies.

Strategy for Improving Crowdfunding Investments in Startup Business

Author(s): Yanti Pasmawati,Alva Edy Tontowi,Budi Hartono,Titis Wijayanto / Language(s): English Issue: 1/2022

This research was conducted to analyze the extent to which online customer reviews (OCRs) can stimulate investment backers as a strategy to increase crowdfunding investment. Design / Method / Approach: This research is quantitative. Natural language processing (NLP) processes review text documents based on linguistic study, a lexicon-based method is used for sentiment analysis classification based on polarity score (pros and cons), while Multiple linear regression forms a model or relationship between online customer reviews and crowdfunding investments. OCRs consisting of numeric and text features were collected from one hundred technology products (3D printing, drones, cameras, wearables) on Kickstarter.com. Findings: The study results show that, in addition to positive reviews, the number of comments and the number of sentiment reviews can increase consumer interest in investing in technology products on the crowdfunding platform. Moreover, positive reviews have the most positive effect on crowdfunding investments. Practical Implications: The study results are expected to be used for startup business, especially technology products as a strategy to increase funding investment on a reward-based crowdfunding platform. Startups can take advantage of online customer reviews as one of important factors in stimulating potential backers and backers to invest. Social implications: The strategy of utilizing online customer reviews can be used especially for technology product-based startup business to get funding support as a resource in completing a product development stage. Originality / Value: The novelty of this research is that it focuses on a technological product development stage, product campaigns on a reward-based crowdfunding platform, considering online customer reviews through sentimental (online reviews) and numerical characteristics (number of comments, number of sentiment reviews) simultaneously as a strategy to increase investment. Research Limitations / Future Research: This study has some limitations as it used only online customer reviews as an attribute that affects crowdfunding investment. Future research is expected to explore online customer reviews to determine important attributes (unique words) as consideration for strategies to increase crowdfunding investment. Paper type: Empirical

Zwischen sozialem Bedarf und Potenzial der Sprache: Zu Realisierungsmöglichkeiten der Äquivalenz von weiblichen Personen- und Berufsbezeichnungen im Polnischen und Deutschen – ein Update

Author(s): Paweł Bąk / Language(s): German Issue: 6/2022

The socioculturally determined need for corresponding forms of expression is often different in various cultural circles, which is in turn influenced by social factors. However, the socially determined use of language is only possible within the framework of the currently available linguistic means. This paper presents the recent results of a survey the problem formulated in the title conducted among women in Germany and Poland. The aim of the study was to examine, 10 years after the last similarly designed survey, the development and possible changes in the linguistic realisation of female personal names in German and Polish language use. The study differs from other research in that here the use of the language is presented in relation to the language users themselves. A certain asymmetry in the formation and use of feminine personal names is confirmed between German and Polish. However, it is not merely an expression of attitudes, and cannot serve as the sole criterion for an evaluation of the mentality of a particular cultural group; it is also partly conditioned by the possibilities of the language system itself.