The FFG funded GENTIO project has ended with MODUL Technology contributing advances in AI-based NLP and NER/NEL approaches, development of its Semantic Knowledge Base, and new classification and summarization algorithms. State of the art AI has improved our Natural Language Processing, Named Entity Recognition and Named Entity Linking components. Good NER combined with relation extraction can feed our Semantic Knowledge Base with new entities and relations automatically identified in news articles. To explore new possibilities to tackle this, we started looking at fine tuning Large Language Models (LLMs) with promising early results, the subject of a conference presentation. Transformers models were compared for the task of text summarization.
A significant impact was made in the task of news article classification, following the IPTC News categories. Here, we decided on fine tuning on top of BERTopic, a neural model for topic classification. The initial topic classification is fed through a Transformer model to align topics to the IPTC News categories. We could show higher accuracy scores on annotated news corpora than just using a language model like RoBERTa to classify articles directly, also published this year at a conference. GENTIO has enabled MODUL Technology to advance in its deployment of state of the art AI (e.g. Transformers models, language models) for core text understanding tasks than underlie ‘Web intelligence’, i.e. extracting trends and insights from Web data.

The FFG funded project EPOCH, coordinated by MODUL Technology, demonstrated the groundbreaking use of machine learning/AI approaches to time series forecasting combined with Web intelligence – the analysis of topics and trends in online news and social media over time. Supported by webLyzard technology’s Web intelligence platform and wood price data from KPMG, MODUL Technology’s AI experts combined statistical forecasting models with features extracted from the online news and social media based on mentions of terms relevant to the wood market and combined sentiment of documents mentioning the terms (e.g. ‘Borkenkäfer’, the bark beetle which damages trees).  Four categories of terms were created, and were tested for their predictive power (the extent to which they improve the accuracy of price prediction compared to pure statistical forecasting). We compared a number of models (Random Forest, XGBoost, LSTM, GRU) with four different datasets (hardwood, softwood, nadelholz (coniferous), rohpapier (raw paper)). The rolling mean sentiment feature was found to be the most predictive for future wood prices, especially for the terms related to weather and pests. We provide a slideset explaining the wood price prediction experiments and a Web interface which allows you to visualise the accuracy and feature importance in prediction using different models and datasets. The experiment code is also available to test the models with your own price data. 

• Slides
• Web interface
• Experiment code

This groundbreaking work was extended to look at predictions of prices in other markets and make use of recent Transformers models such as FinBERT to generate the sentiment features from a news corpus. The developed models outperformed classical models, showcasing the fact that combining basic economic logic (e.g., supply and demand) with affective and background knowledge can lead to significant improvements. MODUL Technology will continue to bring this advance in price prediction to other domains through future projects. 

Journal Articles

• Kaplan, H., Weichselbraun, A., Braşoveanu, A.M.P. (2023) Integrating Economic Theory, Domain Knowledge and Social Knowledge into Hybrid Sentiment Models for Predicting the Crude Oil Markets. Cognitive Computation. Springer Nature. DOI: 10.1007/s12559-023-10129-4.
• Weichselbraun, A., Steixner, J., Brasoveanu, A.M.P., Scharl, A., Gobel, M & Nixon, L.B. (2022). Automatic Expansion of Domain-Specific Affective Models for Web Intelligence Applications. Cognitive Computation 14(1), pp. 228-245. DOI: 10.1007/s12559-021-09839-4.

Book Chapters

• Braşoveanu, A M.P., & Andonie, R. (2022). Visualizing and Explaining Language Models. In Integrating Artificial Intelligence and Visualization for Visual Knowledge Discovery, pp. 213-237. Cham: Springer International Publishing. DOI: 10.1007/978-3-030-93119-3_8.

Modul Technology is excited to start working on the new project AI-CENTIVE, funded by the FFG via the ICT of the Future program. The goal of the 3 year project is to build and manage a complex mobility data ecosystem which will support changing citizens’ mobility behaviour towards choosing more sustainable options. It is an interdisciplinary project that combines data science, deep learning, weather prediction, information visualisation and tourism management. The project is implemented in collaboration with webLyzard technology, Zentralanstalt für Meteorologie und Geodynamik (ZAMG), ummadum, Data Intelligence Initiative (DIO) and BOKU. 

Nowadays the data on mobility choices and the context in which they are taken is split. The aim of the project is to turn it into actionable information thanks to efficient knowledge extraction. A common mobility data ecosystem will merge this data, allowing new AI models that learn to anticipate mobility choices based on varying contexts such as the availability of more environmentally friendly options. Customised incentives leveraging the AI predictions will motivate citizens to make more sustainable mobility choices. As such, AI-CENTIVE will contribute to Austria’s achievement of the UN 2030 Agenda for Sustainable Development. 

Stay tuned … we will very soon go online with the project’s website!

MODUL Technology is excited to announce that its participation in the TRANSMIXR project has begun,  backed by EU Horizon Europe funding of 9 million euros.

With the maturity of eXtended Reality (XR) and Artificial Intelligence (AI) technologies, a unique window of opportunity for the European Creative and Cultural Sector (CCS) exists to reimagine digital co-creation, interaction and engagement possibilities. The TRANSMIXR platform will provide (i) a distributed XR Creation Environment that supports remote collaboration practices, and (ii) an XR Media Experience Environment for the delivery and consumption of evocative and immersive media experiences. Ground-breaking AI techniques for the understanding and processing of complex media content will enable the reuse of heterogeneous assets across immersive content delivery platforms. Using the Living Labs methodology, TRANSMIXR will develop and evaluate four pilots that bring the vision of future media experiences to life in four important CCS domains: news media, broadcasting, performing arts, and cultural heritage.

TRANSMIXR Project Coordinator, Dr. Niall Murray of TUS and investigator in the SFI Adapt Centre said “TRANSMIXR is a very exciting project that will create a suite of user centric technologies to support the creation, consumption and understanding of new media experiences in distributed, collaborative and immersive ways. Underpinned by the convergence of AI and XR, the design of these new systems will be informed by and be evaluated with real end users. A key strength of the TRANSMIXR consortium is its interdisciplinary nature, bringing complimentary technical, methodological and domain expertise together to create impactful solutions for the creative and culture sectors.”

MODUL Technology CTO and project lead, Dr. Lyndon Nixon, added “TRANSMIXR will benefit immediately from the multilingual Web data collection and annotation pipeline of MODUL Technology, while we look forward to the opportunity to further innovate through AI in terms of content understanding and semantic search across heterogeneous, multimodal data collections.”

Website: https://www.transmixr.eu/

The EU’s Horizon Europe programme funding research and innovation is coming to an end, and one of the last Horizon Europe projects supporting the media industry also recently came to a successful conclusion. ReTV stood for “re-inventing television for the interactive age”. With dropping viewing time for traditional broadcast TV and an audience shifting to online streaming, Europe’s media industry needs to adapt to distributing their videos in multiple digital channels like Instagram and TikTok and needs tools to help them publish the right video at the right time to get the attention of an online audience overwhelmed by content availability.

It was led by MODUL Technology – the research spin-out of MODUL University Vienna where faculty and researchers collaborate to innovate in new media technologies. A cross-European consortium developed online services and software that help media organizations such as broadcasters to identify the topics of heightened interest to their audiences, find media assets in their collections related to those topics, adapt media content for publication on digital channels such as social networks and schedule the optimal time to publish. MODUL brought data science and AI innovation to the project work, backed by its access to a huge scale data collection and annotation pipeline which provides millions of Webpage texts and social media posts for analysis.

Firstly, we have developed a large knowledge base of global and local events. Media organizations benefit from knowing about future events relevant to their audiences, and we can configure event listings to the interests of each audience. Secondly, we provide advanced components for data analysis that can extract the most significant keywords, topics and stories from the stream of online content. Thirdly, we combine our knowledge of events and topics over time with the latest techniques in machine learning and deep learning to provide topic prediction capabilities, i.e. suggest the topics that will be popular with the audience at a future date. All of our knowledge modelling, knowledge extraction and predictive analytics capabilities continue to be used in other ongoing projects in diverse areas such as brand communication, mobility and digital marketing.

Finally, MODUL Technology demonstrated its capacity to professionally co-ordinate a large EU project through to its successful conclusion. As is written in the project’s final review report, a group of experts evaluated the results of ReTV and stated: “The project has delivered exceptional results with significant immediate or potential impact on relevant stakeholders, namely, content producers, distributors, and broadcasters. The results can bring innovation and competitiveness advantage to those stakeholders by providing them the ability to identify trends and relevant content in a semi-automated or assisted manner, consequently producing and publishing content through the best channels and meeting users’ expectations.“

While we are proud of the results of ReTV and the successful co-ordination by MODUL Technology, now it is time for the team behind this research spin-out of MODUL University Vienna to continue to innovate in data science and AI, and look forward to new challenges funded by the soon-to-be-launched EU Horizon Europe funding program.

Contact for ReTV and MODUL Technology: Asst.Prof. Dr. Lyndon Nixon, nixon AT modultech DOT eu