Sanda Martinčić-Ipšić gave a talk "Generative artificial intelligence: technology, challenges and economic benefits" and participated in the panel "Will AI replace developers?" while PhD student Vlatka Davidović presented the results on "Seq-to-Seq Neural Model for Croatian Abstractive Summarization", at CASE 2024 conference in Zagreb 26.2.2024.
PhD student Andrija Poleksić participated in Multi3genration COST Training School: Representation mediated multimodality- grounded representation, reasoning, and learning for interactive interpretation where he presented his research "Automatic knowledge graph construction from scientific publications", University of Malta, programme Valleta, 30.1.2024. - 2.2.2024.
At the DSC Europe 23 Andrija Poleksić held technical tutorial on "Relation Extraction From Texts" and Sanda Martinčić-Ipšić modarated the panel on "LLM Use In Business Transformation". She also participated in the Data Science Summit at Data Science Conference Europe 23 Belgrade, 20-24 November 2023.
PhD student Andrija Poleksić gave the talk "Extraction of relations for automatic knowledge graph construction" and Vlatka Davidović presented her work on "Report on Sequence-to-Sequence Neural Model for Croatian Abstractive Summarization" at 2nd INFCON23 Rijeka, Croatia, Nevember 11-12, 2023.
PhD student Andrija Poleksić gave the talk "Relation extraction for knowledge graph construction on scientific corpora " at 7th International PhD Conference, Ljubljana - Graz - Rijeka, programme Rijeka, Croatia, October 5-6, 2023.
Student Students Andrija Poleksić presented the research paper "Effects of Pretraining Corpora on Scientific Relation Extraction using BERT and SciBERT" at 2nd NLP4KGC: Natural Language Processing for Knowledge Graph Construction Workshop, Semantics 2023 Conference in Leipzig, 20-22. September. 2023.
PhD student Vlatka Davidović presented the research paper "Towards Sequence-to-Sequence Neural Model for Croatian Abstractive Summarization" at 34th Central European Conference on Information and Intelligent Systems CECIIS Conference in Dubrovnik, September 22nd 2023.
Sanda Martinčić-Ipšić participated at MC meeting of CA18231 - Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation (Multi3Generation) , project, Warsaw, 20-21.9.2023.
Students Andrija Poleksić and Armin Krivićić presented the research papers "Sentiment of the Tweets on Russo-Ukrainian War: the Social Network Analysis" and "Analyzing Sentiment of Reddit Posts for the Russia-Ukraine War" at 46. IEEE MIPRO Conference in Opatija, May 25th and 26th 2023.
Langnet memeber Ljupčo Todorovski gave an keynot talk "Data Science for Computational Scientific Discovery" at Data Science Adria Conference in Rijeka. Faculty of Informatics and Digital Technologies, was coorganizer of the event, and prof. Todorovski is a member of our PHD doctoral study council. 19.05.2023. program
Sanda Martinčić-Ipšić participated in round table Who & how will educate future AI experts & Data Scientist at Data Science Adria Conference, Zagreb, 17.05.2023. program
Sanda Martinčić-Ipšić participated in round table on Artificial Intelligence organized by AIRI Center, Rijeka 18.5.2023.
Sanda Martinčić-Ipšić participated in Science in Media Communication Workshop at Moise Palace in Cres. The workshop was organized by PhD School, University of Rijeka Foundation and Bernays, Cres 20-23.5.2023.
LangNet members Slobodan Beliga and Sanda Rartinčić-Ipšić received the award, for the excellence in teaching by University of Rijeka, 15th May 2023, Rijeka. CONGRATULATIONS!
Sanda Martinčić-Ipšić gave the talk at MESOC Final Conference: Exploring and assessing the societal impacts of culture on MESOC Toolkit and Document Repository. MESOC's research has sought to develop innovative methodologies to capture and measure the wider societal value of culture that includes, but is not limited to its economic impact. It also encompasses personal health and well-being, urban renovation and regeneration and people's engagement and participation. MESOC provides improved cultural value measurements, including by the use of AI, and case studies in support of effective and inclusive policies and institutional frameworks which offer a convincing vision for citizens to face current cultural and societal transformations. Furthermore, the project uses hundreds of academic and grey literature documents, delved into dozens of city case studies, and interviewed scores of experts and stakeholders. We have also invested in the technological development needed to exploit the knowledge and resources, building toolkits for a new vision to develop policies oriented to the cultural and creative sector. The program. Paris, 9-10.3.2023.
PhD students Anto Čabraja and Vlatka Davidović participated with a talk at SemTech Research Class- The programm is availovle at link, 16.12.2022.
LangNet member Slobodan Beliga received the award, for the excellence in scence as young scientist by University of Rijeka Foundation, 25th November 2022, Rijeka. CONGRATULATIONS!
Professor Marko Robnik Šikonja, Faculty of Computer and Information Science, University of Ljubljana held a Research Class talk "Cross-lingual transfer with large language models" Cross-lingual transfer with large language models2, 24th November 2022.
Sanda Martinčić-Ipšić participated in the organization of the Data Science Summit at Data Science Conference, DSC Europe 22 where She moderated the round table discussion on collaboiration between industry and academia, 16th November 2022. At the DSC Europe 22 she gave invited talk on "BI Toolkit Powered by Natural Language Processing" in the Data Visualisation Track, 17th Novembre 2022.
Sanda Martinčić-Ipšić held an invited lecture "Natural Language Processing: the Case of Integration in Business Intelligence" at Faculty of Organizational Sciences, University of Belgrade, 16.11.2022.
LangNet PhD students Karlo Babić, Anto Čabraja, Marko Pribisalić, Vlatka Davidović and Dino Aljević presented current research resuts on INFCON22 Conference , FICIT, 11.-12.11.2022.
Sanda Martinčić-Ipšić participated at MC meeting of CA18231 - Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation (Multi3Generation) , project, Venice, 19.9.2022.
Petar Kristijan Bogović presented the paper "The NLP Powered BI Toolkit: The Case of MESOC" co-authored with Dino Aljević, Božidar Kovačić and Sanda Martinčić-Ipšić IEEE MIPRO 2022 Conference in DE-DS Section (MIPRO-DE-DS 2022) , 27.05.2022.
LangNet member Sanda Martinčić-Ipšić is one of the co-authors of the paper " Neural Natural Language Generation: A Survey on Multilinguality, Multimodality, Controllability and Learning" in Journal of Artificial Intelligence Research (JAIR), Q2 WOS, April 2022.
The paper is published as the result of COST Multi3Generation project.
Wa are shotlisted to the best 6 Open Science practices in the YERUN Open Science Awards in the researcher category. For this pilot call, 30 applications are recieved in total: 24 for the researchers category and 6 for the staff category. Following the Phase 1 of evaluation, out of these 30 applications, 8 have been shortlisted (6 for the researchers category and 2 for the staff category). 14.2.2022. Link
LangNet project recieved the UNIRI Excellence Awards in Science as the highely ranked project in social sciences for results in 2021. Congratulations to all current and former LangNet members. 1.2.2022. Link
Ana Meštrović, Slobodan Beliga, Mihaela Matešeć and Sanda Martinčić-Ipšić gave the interview to Novi List, 12.12.2021.
Ana Meštrović, Mihaela Matešeć and Sanda Martinčić-Ipšić collaborated on the mobile application ČITATA with Croatian syllabification algorithm. This application was developed as part of the reading aid at Study of Logopedy at the University of Rijeka. Čitata, 9.12.2021.
Sanda Martinčić Ipšić gave the invited talk "Multidimensional Knowledge Network: Modelling of Factual, Conceptual, Procedural, and Metacognitive Knowledge" at 12th International Conference on Information Technologies and Information Society, 5.11.2021. ITIS2021
Petar Kristijan Bogović presented the paper "Topic Modelling of Croatian News During COVID-19 Pandemic" co-authored with Ana Meštović, Slobodan Beliga andSanda Martinčić-Ipšić IEEE MIPRO 2021 Conference in CIS Section (MIPRO-CIS 2021) , 30.09.2021.
Dino Aljević presented the paper "Extractive Text Summarization Based on Selectivity Ranking" co-authored with Sanda Martinčić-Ipšić and Ljupčo Todorovski at IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA'21) , 27.8.2021.
LangNet member Sanda Martinčić-Ipšić published the paper " A Framework for the Application of Shipboard Energy Efficiency Monitoring, Operational Data Prediction and Reporting." in Pomorstvo-Scientific Journal of Maritime Research 35(1), 3-15. 2021. https://doi.org/10.31217/p.35.1.1 (WOS-Emerging sources Q4, IF 0.22) July 2021.
LangNet member Sanda Martinčić-Ipšić published the paper " Predicting Seagoing Ship Energy Efficiency from the Operational Data" in Sensors journal, Q1 WOS, April 2021.
Sanda Martinčić-Ipšić and Ana Meštović presented the current research results and projects of LangNet group included in the Laboratory for Natural Speech and Language Processing at the Centre for Artificial Intelligence and Cybersecurity, University of Rijeka, 31st March 2021. AIRI Centre
video
Sanda Martinčić-Ipšić gave an invited talk „Natural Language Processing 4 for Policy Texts Analytics“ at “Artificial Intelligence for Policy Making” workshop at Social & Creative project funded by the Interreg MED programme. Workshop on exploring trends of Artificial Intelligence at the service of policy making was held online on 22nd of March 2021. Link Social & Creative project
Workshop video
Sanda Martinčić-Ipšić presentet current work in LangNet group at WG2 meeting of the CA18231 - Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation (Multi3Generation) , Online, 10.2.2021.
LangNet project recieved the UNIRI Excellence Awards in Science as the highely ranked project in social sciences for results in 2020. Congratulations to all current and former LangNet members. 11.12.2020. Link
COSTNET videos are summary of the action reserach results.
COSTNET 1 – Causation
COSTNET 2 – Monte Carlo
COSTNET 3 – Radicalisation
COSTNET 4 – Evolution
COSTNET 5 – COVID
COSTNET 6 – SOAM
On Wednesday, December 9, 2020, starting at 10:00 am, as part of the teaching of the course Search Methods and Classification of Information, the invited lecture will be given by Dr. sc. Slobodan Beliga on the topic "Keyword extraction in computer analysis of natural language" Link
Sanda Martinčić-Ipšić talked about the prospects for studying Data Science at the University of Rijeka and collaboration with the Center for Artificial Intelligence and Cybersecurity at Data Science Conference 2020 Link 19.11.2020.
LangNet team member Tajan Ban Kirigin co-editored the book "Logic, Language, and Security" in LNCS Springer. This LNCS Festschrift was published in honour of Andre Scedrov on the occasion of his 65th birthday. The 11 technical papers and 3 short papers included in this volume show the many transformative discoveries made by Andre Scedrov in the areas of linear logic and structural proof theory; formal reasoning for networked systems; and foundations of information security emphasizing cryptographic protocols. These papers are authored by researchers around the world, including North America, Russia, Europe, and Japan, that have been directly or indirectly impacted by Andre Scedrov. Logic, Language, and Security
Members of LangNet team Karlo Babić, Sanda Martinčić-Ipšić and Ana Meštrović published the paper " Survey of Neural Text Representation Models" in Information journal, October 2020.
Croatian Science Foundation granted financing of "Multilayer Framework for the Information Spreading Characterization in Social Media during the COVID-19 Crisis (InfoCoV)" project lead by Ana Meštrović. InfoCoV project is renked in the top 11 projects that will be financed in the next 18 months.2.7.2020. Results
Ana Meštrović presented the project proposal "Multilayer Framework for the Information Spreading Characterization in Social Media during the COVID-19 Crisis" at the ICT-20 @ COVID-19 - Online conference. 27.5.2020. LINK
Sanda Martinčić-Ipšić talked about LangNet research and Data Science in Prometej science show on Croatian National TV, 26.3.2020.
Members of LangNet team Sanda Martinčić-Ipšić and Ana Meštrović published the paper " Structural Analysis of Factual, Conceptual, Procedural, and Metacognitive Knowledge in a Multidimensional Knowledge Network"
in Complexity, Hindawi journal (IF 2.591, Q1).
Sanda Martinčić-Ipšić participated in the round table on past, present and future of data science at Women in Data Science Conference 2020 in Zagreb from 04-05.03.2020.
Ana Meštrović participated in the HTV show "Kod nas doma". The topic was dedicated to the collaboration between researchers and OmoLab in the development of the mobile application for dyslexia.
Sanda Martinčić-Ipšić participated at COST Multi3Generation MC and WG Meeting at University of Copenhagen, Copenhagen, Denmark, 21-23.01.2020.
Scientists from the University of Rijeka and members of the Langnet team Ana Meštrovic, and sanda Martinčić-Ipšić, and Miahela Matešić (Faculty of Humanities and Social Sciences) received the University of Rijeka Foundation Award for their achievements and transfer of knowledge from academia to industry (news paper). Congratulations!
Sanda Martinčić-Ipšić participated in the organization of the Data Science Summit at Data Science Conference 5.0 Change the World Through Data) in Belgrade from 19-21.11.2019. She moderated discussion between industry and academia in two round tables with topics in Natural Langauge Processing and Graph Analytics.
LangNet members A. Meštrović, M Matešić and S. Martinčić-Ipšić received "1st Technology Transfer" award, by University of Rijeka Foundation and company Alarm automatika, for the SyllAssist work on the algorithm for Croatian syllables segmentation used in OmoLab's application OmoReader for reading support in dyslexia, 15 November 2019, Rijeka.
LangNet member, prof. Ljupčo Todorovski on the perspectives of Artificial Intelligence on Slovene National TV (video) 07.11.2019.
Slobodan Beliga, member of LangNet group defended PhD thesis "Keyword Extraction Based on Structural Properties of Language Complex Networks" at Department of Informatics, University of Rijeka. Congratulations!
Karlo Babić and Sanda Martinčić-Ipšić participated at COST COSTNET19 Conference with poster "The Representation of Text in a Multilayer Complex Network for Deep Learning" in Bilbao, Spain, October 2019.
Sanda Martinčić-Ipšić participated at kick-off meeting of the CA18231 - Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation (Multi3Generation) COST project in Bruxelles, Belgium, 9.9.2019.
Sanda Martinčić-Ipšić participated at International Summer School on Deep Learning, Warsaw, Poland, July 22-26, 2019.
Marko Pribisalić participated at 32nd Bled eConference Humanizing Technology for a Sustainable Society whwer he presented the paper Selecting a Business Intelligence Solution that is Fit for Business Requirements, coathored with Sanda Martinčić-Ipšić and Igor Jugo. Sanda Martinčić-Ipšić is the member of organizing committee and a chair of Data Science / Big Data track.
SemText project as the spinof form formare LangNet project, had a kick-off meeting on June 11th 2019. SemText we wish you a lot of scucess!!
Lider published the article on speech and language technologies including the LangNet research plans pdf, June 2019.
The kick-off meeting of the project "Keyword Extraction and Summarization Based on Language Networks - LangNet" was on the 29th of March 2019 at the Department of informatics.
The team members discussed the working plans and showed the ongoing progress of the current research activities.
Members of LangNet team Sanda Martinčić-Ipšić, Tanja Miličić and Ljupčo Todorovski published the paper
" The Influence of Feature Representation of Text on the Performance of Document Classification"
in Applied Sciences journal (IF 1.689).
LangNet group recieved University of Rijeka grant for research in "Extraction and Summarization Based on Language Networks". The goal of this research is to develop new methods for keyword extraction and extractive summarization enabled by the formalisms of complex networks and neural networks - project abstract.
Karlo Babić participated at Early Career Training Event of COSTNET project , the Department of Statistics, Ludwig-Maximilians-University, Munich, Germany, 11-13 February 2019. All talks were in the networks domain, some were from the statistics perspective, others from physics, social, and computer science perspective. Two main speakers were Tiago de Paula Peixoto and Tom A. B. Snijders. Karlo Babić talked about methods for embedding networks into vector representations.
OmoReader is a free mobile application that helps people with dyslexia to read text comfortably. The author of the idea is Petar Reić from OmoLab and Locastic companies developed the app. The algorithm used in the process of separating the word into the syllables is the result of the original scientific work "Syllabification based on maximal onset principle for Croatian", published in the Speech journal. The authors of the paper are Prof. Ana Meštrović, Prof. Sanda Martinčić-Ipšić from the Department of Informatics and Asist. Prof. Mihaela Matešić from the Faculty of Philosophy in Rijeka. You can read more about developing an app in the newspaper article. The application can be downloaded here. The paper is here.
Sanda Martinčić-Ipšić participated at International Winter School on Big Data, Cambridge, UK, January, 2019.