Претрага

Упит dcterms_creator_txt:Ksenija A. Stojanović

1580 items

  • Rule-based Automatic Multi-word Term Extraction and Lemmatization

    Ranka Stanković, Cvetana Krstev, Ivan Obradović, Biljana Lazić, Aleksandra Trtovac (2016)
    In this paper we present a rule-based method for multi-word term extraction that relies on extensive lexical resources in the form of electronic dictionaries and finite-state transducers for modelling various syntactic structures of multi-word terms. The same technology is used for lemmatization of extracted multi-word terms, which is unavoidable for highly inflected languages in order to pass extracted data to evaluators and subsequently to terminological e-dictionaries and databases. The approach is illustrated on a corpus of Serbian texts from ...
    term extraction, terminology, multi-word units, lemmatization, finite-state transducers
    ... evaluation of results is to follow and if the approved MWT is to be entered into some kind of a dictionary or a terminological data-base. In this case we need a lemmatized MWT, that is, a MWT in the form of a dictionary head-word. The problem of lemmatization of special kind of MWUs, 507 ...
    ... number, case and animateness, belong to the AXN class. X stands for a component that does not inflect when the MWU inflects or a separator, usually a space or a hyphen. Nominal MWUs in Serbian belong to one of several tens of different general classes, but 14 of these classes account for ...
    ... lemmas. For instance, in the case of 2-component MWUs the order of precedence of graphs is: AXN, 2XN (a noun preceded by a word that does not inflect in the MWU), N2X (a noun followed by a word that does not inflect in the MWU), NXN. Thus, two MWU forms mašine taložnice and korita trake ‘belt ...

  • Old or New, We Repair, Adjust and Alter (Texts)

    Cvetana Krstev, Ranka Stanković (2020)
    U ovom radu predstavljamo kako se e-rečnici i kaskade transduktora konačnih stanja implementirani u alatu Unitex mogu koristiti za rešavanje tri problema transformacije teksta: ispravljanje tekstova nakon OCR-a, vraćanje dijakritičkih znakova i prebacivanje između različitih jezičkih varijanti.
    ispravka teksta, OCR greške, restauracija dijakritika , jezičke varijante, elektronski rečnik, transduktori konačnih stanja
    ... ‘beer’ and ниво ‘level’ would be accepted. A few examples of the application of this procedure are given in Table 1. Special attention is paid to hyphenated words. A hyphen in a Serbian OCR text can signify a word hyphenated at the end of the line or a hyphen in a multi-word. Our procedure first eliminates ...
    ... performed. Table 2. A FST that initiates a correction for a specific replacement pair (top). A generic FST that performs a certain type of replacement (bottom). two words: the original, input word ($1$Y) and the corrected, output word ($1$X).4 The application of these FSTs always leaves in a processed texts ...
    ... multiple candidates for sto (sto ‘table/hundred’ and što, a functional word that can be an adverb, a pronoun or a conjunc- tion); – A list of 50 most frequent bigrams obtained from the SrpKor in which at least one word would contain letters c, s, z or a digraph dj if diacritics were removed; for instance ...

  • Using English Baits to Catch Serbian Multi-Word Terminology

    Cvetana Krstev, Branislava Šandrih, Ranka Stanković (2018)
    In this paper we present the first results in bilingual terminology extraction. The hypothesis of our approach is that if for a source language domain terminology exists as well as a domain aligned corpus for a source and a target language, then it is possible to extract the terminology for a target language. Our approach relies on several resources and tools: aligned domain texts, domain terminology for a source language, a terminology extractor for a target language, and a ...
    aligned texts, word alignment, terminology extraction, electronic dictionaries, morphological inflection
    ... language, a terminology extractor for a target language, and a tool for word and chunk alignment. In this first experiment a source language is English, a target language is Serbian, a domain is Library and Information Science for which a bilingual terminological dictionary exists. Our term extractor ...
    ... that if for a source language domain terminology exists as well as a domain aligned corpus for a source and a target language, then it is possible to extract the terminology for a target language. Our approach relies on several resources and tools: aligned domain texts, domain terminology for a source language ...
    ... which a source part of a chunk matches a term from a list of domain terms in a source language: S(align.chunk) ∼ S(term.list), where symbol ∼ denotes the relation “match” (that is for our ex- periment defined in Subsection 4.5.); • Filtering once more previously filtered chunks to those in which a target ...

  • SASA Dictionary as the Gold Standard for Good Dictionary Examples for Serbian

    Ranka Stanković, Branislava Šandrih, Rada Stijović, Cvetana Krstev, Duško Vitas, Aleksandra Marković (2019)
    У овом раду представљамо модел за избор добрих примера за речник српског језика и развој иницијалних компоненти модела. Метода која се користи заснива се на детаљној анализи различитих лексичких и синтактичких карактеристика у корпусу састављених од примера из пет дигитализованих свезака речника САНУ. Почетни скуп функција био је инспирисан сличним приступом и за друге језике. Дистрибуција карактеристика примера из овог корпуса упоређује се са карактеристиком дистрибуције узорака реченица ексцерпираних из корпуса који садрже различите текстове. Анализа је показала да ...
    Српски, добри примери из речника, аутоматизација израде речника, издвајање својстава, Машинско учење
    ... example. If a noun has a complement that affects its meaning, the complement should be represented in the example: Tada se javila u njega velika ljubav i velika podobnost za slikarstvo (paraphrase: ‘In that moment he felt a great affection and a great talent for painting’)3. If a keyword is a verb that ...
    ... only in paper form. At the same time, a formal description of dictionary entry was produced, and a lexical database model was developed (Stanković et al., 2018). The conversion of the SASA dictionary from unstructured text into a lexical database consisted of a thorough analysis of formatting conventions ...
    ... syntactic whole – with a subject and a predicate. It is even possible to add a missing sentence constituent, but it has to be in square brackets, as a mark of this kind of editorial intervention. (Though the excerpts are in the form of full sentences, the context they provide is sometimes insufficient, and ...

  • 2D geoelectrical resistivity tomography application at the former city waste dump "Ada Huja": Eco-geological problem

    Branislav Sretenović, Filip Arnaut, Ivana Vasiljević, Vesna Cvetkov (2019)
    2D inversion, geoelectrical method, Wenner array, ecology, Ada Huja
    ... profile line I has a length of 124 m (Figure 4). A high value resistivity zone with a length of 56 m (252 m to 308 m of the profile line) can be noticed and it represents most likely sand, gravel and other construction materials. Decomposed waste has a lens-like geometry and retains a low resistivity ...
    ... exhibited data, and looking at the 2017 methane stimulated fire at the landfill “Vinča“, which posed a threat to a million and a half inhabitants of Belgrade, such a mini-plant would not be a enormous investment. Daily production of waste for the city of Belgrade reaches 1700 tones (Mitrović, 2014) ...
    ... reflection is noticed that could indicate a horizontal interface, that was not detected with the geoelectrical scanning method. Acquired borehole data indicated a presence of subsurface water at a depth of around 5 m (Figure 5, middle) which makes a good reflective surface that divides two di ...

  • An aproach to Implementation of blended learning in a university setting

    Ivan Obradović, Ranka Stanković, Olivera Kitanović, Jelena Prodanović (2011)
    ... contains a short definition of the term in both languages, but without any relations between equivalents. An example of a dictionary entry, in English, and then in Serbian, is: geodatabase (GDB), A collection of geographic datasets of various types held in a common file system folder, a Microsoft ...
    ... support system. Hence, a decision was made to introduce Moodle to the already existent blended learning system at FMG. Moodle is a free and open source (published under the GNU Public License) LMS (Learning Management System) platform. From a technical point of view, it is a web application ...
    ... motivation for introducing a blended learning system came from the fact that in traditional face-to-face learning, the students were prone to taking a basically passive role. Thus the new learning system was aimed at turning the student role in the learning process to a more active one, and enhancing ...

  • Medical Domain Document Classification via Extraction of Taxonomy Concepts from MeSH Ontology

    Mihailo Škorić, Mauro Dragoni (2019)
    This paper is a result of a task that was presented to attendants of Keyword Search in Big Linked Data summer school, that was organized by Vienna University of Technology, under the Keystone COST action in the summer of 2017. It presents a specific approach to the classification via creation of minimal document surrogates based on the US National medical library’s MeSH ontology, which is derived from the Medical Subject Headings thesaurus. In a series of previously classified medically ...
    document classification, MeSH, ontology, information extraction
    ... set of terms from document Di, and Sj is a set of terms from document Dj , then this index can be defined as a double the number of common terms divided with the total number of terms in both documents (if S is a set, |S| is a number of terms in that set). If documents do not have any common terms then ...
    ... identified with a maximum of four repetitions (if they have more than eight blocks) of a term denoting a class. After applying these steps, only class identifiers now appear in document surrogates, which should be easily counted. After the test sets have been successfully created, a simple program is ...
    ... eca.2019.19.1.3 ABSTRACT: This paper is a result of a task presented to attendants of Keyword Search in Big Linked Data summer school, that was organized by Vienna University of Technology, under the Keystone COST action in the summer of 2017. It presents a specific approach to the classification via ...

  • Asbestos-Based Pottery from Corsica: The First Fiber-Reinforced Ceramic Matrix Composite

    Philippe Colomban, Aleksandar Kremenović (2020)
    Asbestos-containing pottery shards collected in the northeast of Corsica (Cap Corse) and dating from the 19th century, or earlier, have been analyzed by SEM-EDS, XRPD, FTIR and Raman microspectroscopy. Blue (crocidolite) and white (chrysotile) asbestos fiber bundles are observed in cross-sections. Most of the asbestos is partly or totally dehydroxylated, and some transformation to forsterite is observed to occur, indicative of a firing above 800 C. Examination of freshly fractured pieces shows a nonbrittle fracture with fiber pull-out, consistent with ...
    asbestos, CMC, Corsica, fiber, pottery
    ... of a firing; above 800 "C. Examination of freshly fractured pieces shows a nonbrittle fracture with fiber pull-out, consistent with a composite material behavior, which makes these ceramics the oldest fiber-reinforced ceramic matrix composite. Residues indicate the use of this pottery as a crucible ...
    ... fracture of a ceramic gives rise to a brittle edge, and fracture initiation depends upon the defects and their distribution in the matrix. It is thus difficult to design the shape and mechanical strength of a ceramic object. However, the association of two brittle materials, i.e., a composite material ...
    ... Coarse grains (up to 5 mm) are mixed in a rather homogeneous matrix of a high open porosity (~30%). Table 1. Oxide composition measured by SEM-EDS on spots indicated in Figure 1 for samples a to c. Characteristic values in bold. Sample b Sample c Sample a Sample d Oxide «0» «» «3» «60» «l» «6» ...

  • Threshold-induced correlations in the Random Field Ising Model

    Sanja Janićević, Dragutin Jovković, Lasse Laurson, Đorđe Spasojević (2018)
    Multidisciplinary
    ... Viyy=25 > 1074 © = 1-1024, R=2.30, Vi,=50 = © = 1-1024, R=2.30, V,,—50 a A 12046, R=2.25, Vj,=102 a A 12046, R=2.25, Vy.=102 om ¥ L=4087, R=2.22, Vy,=208 = 1074 V L-4087, R=2.22, Vy,=208 ~ ~ a¥ ~ an e a <a ™ wa ~a04 a , e Q Ag A ? @ 7 e | > » ee NS YY g 3 3 = va to! 102108 . va . sr . S -8 ini T 10 ...
    ... 30, Vy,=50 a vy ine © £=1024, R=2.30, Vy.=50 a vn 4 1=2046, R=2.25, V,,=102 i904 vy 4 1=2046, R=2.25, V,,=102 = 10°4 L=4087, R=2.22, V.=208 = L=4087, R=2.22, V.=208 ~ ~ a # 1024 S 1074 & _ & FS 10% ~s = a = = gf 7s ae a 1 : 7 1044 x yoo 4 = 1044 3 ae = e t 5] 210°) = = S10! % wa <a S 10° = a ~ 10! 102108104 ...
    ... © 1=1024, R=2.30, Vy,—50 a @ 11024, R=2.30, Vy,=50 & 4 L=2046, R=2.25, Vy—102 & vee A L=2046, R=2.25, Vy,—102 Sao = vw ws 10°4 Vy L=4087, R=2.22, Vy, =208 3 ~ VY L=4087, R=2.22, Vy, =208 ~ = 1074 E “n” — eK a = ~z10"4 Aa gs ~~ = | we = 10° zg 1044 108 a Y eve % a 7 Ss = 194 A Sot = = % 7 z % & -6 | z ...

  • Softverski alati za korišćenje resursa za srpski jezik

    Ivan Obradović, Ranka Stanković (2008)
    ... of a transformation of resources such as dictionar- ies, graphs and regular expressions from a format SOFTWARE TOOLS FOR SERBIAN LEXICAL RESOURCES 52a used by Intex to a format used by NooJ. Figure 9 depicts a panel for conversion of a morphologi- cal dictionary into Unicode using a C# procedure ...
    ... words “lice” and “lik” are used in Serbian. Thus a semantic network of concepts in a particular language becomes a cor- responding semantic network of words, which is further materialized as a lexical data base of a specific structure. Development of wordnets started in 1985 in the research team ...
    ... whether it is a simple word or a compound, using SMD, and connects them by logical “or” relations. In the same way, by a simple choice of appropriate fields, the user selects whether he/ she wants to make a query in Cyrillic or Latin alphabet, or in both. WS4QE obtains semantic expansion of a query by ...

  • Vebran Web Services for Corpus Query Expansion

    Ranka Stanković, Miloš Utvić (2020)
    U ovom radu se govori o razvoju veb usluga Vebran i njihovoj primeni u poboljšanju pretraživanja korpusa. Veb-servisi Vebran koriste se za konsultovanje spoljnih leksičkih izvora za srpski jezik (uglavnom elektronski morfološki rečnici i srpski Vordnet) i proširivanje korisničkih upita radi dobijanja relevantnijih rezultata iz srpskih korpusa.
    corpus search, web service, Serbian lexical resources, query expansion
    ... only be used over a https connection, since passing it over a non-encrypted channel would make it trivial for third parties to intercept. The token endpoint is where apps make a request to get an access token for a user. After successful authorization, clients are allowed to send a request spec- ifying ...
    ... attribute word (it does not inflect as a part of MWU). 2) If SMD does not contain an MWU as a compound, each component of MWU is analysed separately. A component of MWU which has been found to be a lemma is associated with marker _L (e.g. ptica pevačica in Table 2). If a component has been found to be an ...
    ... migrated from textual e-dictionaries to a lexical database. After years of development, SMD, developed as a system of textual files, have become a large and complex lexical resource. An on-line applica- tion for dictionary development and management, based on a central lexical data repository (lexical ...

  • WS4LR - a Worksation for Lexical Resources

    Cvetana Krstev, Ranka Stanković, Duško Vitas, Ivan Obradović (2006)
    Lexical Resources, Wordnet, Serbian
    ... describes the syntactic, semantic, derivational, or other properties of a lemma. A part of speech code and an inflectional class code uniquely determine the finite transducer that generates all the forms in a lemma paradigm. A finite transducer, being capable of producing the output, adds to all ...
    ... these dictionaries are given in Appendix A. 2.2 Wordnets Roughly speaking, a wordnet, such as the Princeton WordNet (PWN) is composed of synsets, or sets of synonymous words representing a concept, with basic semantic relations between them forming a semantic network (Fellbaum, 1998). Each synset ...
    ... and runs on a personal computer under Windows 2000/XP/2003 operating system with at least 256MB of internal memory. 1 Introduction The Human Language Technology group at the Faculty of Mathematics has been developing various lexical resources over quite a long period, reaching a considerable ...

  • Extraction of Bilingual Terminology Using Graphs, Dictionaries and GIZA++

    Branislava Šandrih, Ranka Stanković (2020)
    U nauci, industriji i mnogim istraživačkim oblastima, terminologija se brzo razvija. Najčešće, jezik koji je „lingua franca“ za većinu ovih oblasti je engleski. Kao posledica toga, za mnoga polja termini domena su koncipirani na engleskom, a kasnije se prevode na druge jezike. U ovom radu predstavljamo pristup za automatsko izdvajanje dvojezične terminologije za englesko-srpski jezički par koji se oslanja na usaglašeni dvojezični korpus domena, ekstraktor terminologije za ciljni jezik i alat za usklađivanje delova. Ispitujemo performanse metode na domenu ...
    ekstrakcija terminologije, validacija terminologije, GIZA++, grafovi, Unitex, klasifikacija teksta
    ... ‘statistical yearbook of the republican institute’ where A stands for an adjective, N for a noun and PREP for a preposition. Each of these components can be a single word or a MWU. Our system was used in a mode in which all possible MWTs in a word sequence are recognised, and not only the longest one ...
    ... 119–138 ekrana (namely, a photo of a current state of the screen) or as a “skrinšot” (i.e, the word is transcribed). It is not uncommon that even experts from a certain field have difficulties while translating texts that contain domain terminology. As in the example with a “debugger”, the transcribed ...
    ... on the existence of a bilingual dictio- nary with no parallel texts and the second one requiring only the existence of a small amount of parallel data. In order to compile a bilingual lexicon for a specific domain, we combined and compared several settings. Besides using only a parallel sentence-aligned ...

  • Multi-word Expressions for Abusive Speech Detection in Serbian

    Ranka Stanković, Jelena Mitrović, Danka Jokić, Cvetana Krstev (2020)
    Ovaj rad predstavlja istraživanja na usavršavanju i unapređenju srpske verzije rečnika Hurtlex, višejezičnog leksikona uvredljivih reči. Posebnu pažnju posvećujemo dodavanju izraza sa više reči (polileksemskih jedinica) koji se mogu smatrati uvredljivim, jer su takvi leksički zapisi veoma važni za postizanje dobrih rezultata u mnoštvu zadataka otkrivanja uvredljivog jezika. Srpski morfološki rečnici se koriste kao osnova za čišćenje podataka i stvaranje rečnika. Istaknuta je veza sa drugim leksičkim i semantičkim resursima na srpskom jeziku i predviđena je izgradnja sistema za ...
    uvredljiv govor, govor mržnje, leksički izvori, višejezični leksikon, izrazi sa više reči
    ... muzara ‘dairy cow’; N PREP N (25), a noun followed by a prepositional phrase, e.g. govno od čoveka ‘shit of a man’, roba s greškom ‘dam- aged goods’; A ADV N (11), adjectives as simile figures, e.g. glup kao noć ‘stupid as night’, N CONJ N (9), two nouns connected with a conjunction, e.g. bruka i sramota ...
    ... lexicon. They took words with negative polarity as a baseline for creating a basic lexicon of 551 words, which was further enriched via machine learning into a lexicon of 2898 abusive words. Several authors used the Wiegand lexicon as a blacklist in their hate speech and abusive language detection systems ...
    ... crowdsourcing. In an abusive content detection system, a lexicon could be used in one of the following ways: (i) As a classification feature, either as a binary indicator of the abusive word occurrence in the examined text (Pamungkas and Patti, 2019), or a numerical value corresponding to the number of abusive ...

  • Part of Speech Tagging for Serbian language using Natural Language Toolkit

    Ranka Stanković, Boro Milovanović (2020)
    Dok se razvijaju složeni algoritmi za NLP (obrada prirodnog jezika), osnovni zadaci kao što je označavanje ostaju veoma važni i još uvek izazovni. NLTK (Natural Language Toolkit) je moćna Python biblioteka za razvoj programa zasnovanih na NLP-u. Pokušavamo da iskoristimo ovu biblioteku za kreiranje PoS (vrsta reči) oznake za savremeni srpski jezik. Jedanaest različitih modela je kreirano korišćenjem NLTK API-ja za označavanje. Najbolji modeli se transformišu sa Brill tagerom da bi se poboljšala tačnost. Obučili smo modele na označenom ...
    obrada prirodnog jezika, mašinsko učenje, neuronske mreže
    ... March 2020. Having a plethora of different algorithms makes this library a good choice for a research. Serbian language belongs to a group of low-resource languages so there’s a modest research on this topic. First attempts to create an automatic PoS tagger for Serbian relied on a dictionary. Delić ...
    ... all-around PoS tagger that can be a standard for a Serbian language. Best performance in this research is achieved with the Perceptron tagger, a neural network which is more than a decade old. Since then, a breakthrough with deep learning has happened, so there’s a strong belief that further improvements ...
    ... PoS (Part of Speech) tagging, a process of assigning a part of speech category to each token in the text. The program that performs tagging is called tagger. The taggers can be created in multiple ways. In this paper, we will create a tagger for Serbian with a help of a Python library NLTK (Natural ...

  • Mean-based iterative methods for finding multiple roots in nonlinear chemistry problems

    Dejan Ćebić, Nebojša M. Ralević (2021)
    нелинеарне једначине, поступци засновани на срединама, вишеструки корени,

  • INVENTS: A Hybrid Mine Ventilation Planning and Design System

    Lilić Nikola, Stanković Ranka, Obradović Ivan (2013)
    Ventilation system analysis is a complex process based on the calculation and analysis of numerous parameters. These problems can be successfully solved by the SimVent numerical package, but a full understanding and use of the obtained results require the involvement of an experienced specialist in the ventilation field. The solution was found in the creation of a hybrid system INVENTS, whose knowledge base represents a formalization of the expert knowledge in the mine ventilation field. In this paper we ...
    hybrid systems, mine ventilation, expert systems
    ... the following blocks: 1. a block for the analysis of air flow and pressure distribution in the ventilation network, 2. a block for the analysis of climate conditions in mine chambers, 3. a block for temperature and heat depression analysis in fire incidence conditions, 4. a block for gas distribution ...
    ... ventilation planing and analysis problem using a hybrid system composed of three entities: a package for aerodynamic definition of mine ventilation networks - ResNet, a package for ventilation simulation in complex mine ventilation networks - SimVent and a diagnostic expert system for ventilation system ...
    ... architecture of a hybrid system for mine ventilation planing and analysis the software implementation of this system is described, which comprises a symbolic upgrade of the existing simulation packages SimVent and ResNet. The development of a frame-based knowledge base presents a natural approach ...

  • INVENTS: a hybrid system for subsurface ventilation analysis

    Nikola Lilić, Ranka Stanković, Ivan Obradović (2000)
    Ventilation system analysis is a complex process based on the calculation and analysis of numerous parameters. These problems can be successfully solved by the SimVent numerical package, but a full understanding and use of the obtained results require the involvement of an experienced specialist in the ventilation field. The solution was found in the creation of a hybrid system INVENTS, whose knowledge base represents a formalization of the expert knowledge in the mine ventilation field. In this paper we ...
    ... the following blocks: 1. a block for the analysis of air flow and pressure distribution in the ventilation network, 2. a block for the analysis of climate conditions in mine chambers, 3. a block for temperature and heat depression analysis in fire incidence conditions, 4. a block for gas distribution ...
    ... ventilation planing and analysis problem using a hybrid system composed of three entities: a package for aerodynamic definition of mine ventilation networks - ResNet, a package for ventilation simulation in complex mine ventilation networks - SimVent and a diagnostic expert system for ventilation system ...
    ... architecture of a hybrid system for mine ventilation planing and analysis the software implementation of this system is described, which comprises a symbolic upgrade of the existing simulation packages SimVent and ResNet. The development of a frame-based knowledge base presents a natural approach ...

  • Polyhedral characteristics of the cosalite-type crystal structures

    Sabina Kovač, Predrag Dabić, Aleksandar Kremenović, Aleksandar Pačevski, Ljiljana Karanović, Marko V. Rodić, Martina Tribus (2019)
    kristalna struktura, kozalit, poliedarske karakteristike

  • A business intelligence approach to mine safety management

    Ljiljana Kolonja, Ranka Stanković, Ivan Obradović, Olivera Kitanović, Dejan Stevanović, Marija Radojičić (2016)
    ... mutual interoperability a critical issue. One of the means to secure this interoperability within a specific domain is the use of ontologies, containing a comprehensive common domain vocabulary. This means that all terms used within a domain need to be standardized, with a clear and unambiguous ...
    ... achieve this, it does not suffice simply to react to a problem when it occurs, but an analysis of weak points is needed, leading to a reduction of risk. A prerequisite for such an analysis is the creation of a comprehensive database with a history of injuries and conditions that led to an injury ...
    ... documents as well as actions that provide personalized training for a specified context, e.g. a given activity, workplace, operator type, work machine, etc. In order to make knowledge within RudOnto available to a wider circle of users, a client-server application was developed enabling web access to ...