VTT Research System
Not a member yet
161399 research outputs found
Sort by
Automating Customer Feedback Analysis in E-commerce: A Multi-Model Approach
No full textUnderstanding customer satisfaction in e-commerce is crucial for businesses to remain competitive. While traditional feedback analysis methods are labour-intensive and subjective, machine learning advances have enabled more efficient and scalable sentiment analysis. However, existing models struggle with aspect-based sentiment analysis (ABSA), particularly in detecting implicit aspects and handling mixed sentiments. This paper presents a multi-model machine learning pipeline designed to enhance ABSA by integrating fine-tuned Large Language Models (LLMs) with BERT and RoBERTa-based models. The pipeline consists of an LLM-generated synthesized annotated feedback model, a BERT-based aspect detection model, a RoBERTa-based ABSA model, and an LLM-based ABSA model for handling implicit aspects and mixed sentiments. Additionally, a RoBERTa-based model is employed for overall sentiment detection. By leveraging both manually annotated and synthetic data, the pipeline improves sentiment classification accuracy and aspect coverage, even in data-scarce environments. The results demonstrate that combining multiple models enhances detection accuracy compared to single-model approaches. This study provides a scalable and effective solution for e-commerce feedback analysis, offering businesses valuable insights for improving customer experience and decision-making
Thermoelectric and electronic transport properties of thermal and plasma-enhanced ALD grown titanium nitride thin films
No full textTitanium nitride (TiN) thin films demonstrate high electrical conductivity and thermal stability up to 400 °C in ambient conditions, with stability extending to 600–800 °C under inert or vacuum environments. Unlike many metals and transition metal nitrides, TiN combines high carrier mobility with moderate carrier concentration, making it ideal for thermal management and power-efficient applications in nanoelectronics and energy harvesting. This study systematically investigates the thermoelectric and electronic transport properties of TiN films grown by plasma-enhanced atomic layer deposition (PEALD), comparing them to those produced using traditional thermal atomic layer deposition (thermal ALD). These properties are studied as a function of growth temperature and the number of growth cycles. In particular, TiN films deposited by PEALD at 400 °C for 2000 ALD cycles exhibited a remarkable power factor of 512 µW m−1K−2at room temperature compared to a power factor of 4.95 µW m−1K−2measured for thermal ALD films fabricated under the same deposition conditions. Additionally, thermal conductivity was also measured for thicker TiN films (86 nm), yielding values of 26.96 W m−1K−1for PEALD and 7.01 W m−1K−1for thermal ALD, marking the first such report for ALD-grown TiN. These values offer an upper estimate of the thermal behavior in thinner films. Based on these measured properties, the thermoelectric figure of merit (zT) at room temperature was calculated to be 0.0056 for PEALD TiN films which is significantly higher than the value of 0.0002 obtained for thermal ALD TiN films. Our findings provide critical insights into transport properties of TiN, offering guidance for the development of conductive nanolayers in thermoelectric, nanoelectronic, and on-chip cooling applications, where precise control over thermal and electronic behavior is vital, thereby expanding the relevance of ALD TiN in high-performance applications.</p
Extraction of brominated flame retardants from acrylonitrile butadiene styrene (ABS) using supercritical carbon dioxide
Get PDFElectrical and electronic equipment, as well as construction materials, commonly contain brominated flame retardants (BFRs), which are harmful to human health and the environment. These additives can cause major issues during disposal or recycling phases and, hence, methods to remove them are highly demanded. This study focuses on removing three BFRs, namely Tetrabromobisphenol A (TBBPA), Decabromodiphenyl Ether (decaBDE), and Hexabromocyclododecane (HBCD) from acrylonitrile butadiene styrene (ABS) plastic using supercritical carbon dioxide (scCO2) extraction. Virgin ABS compounds with known BFR concentrations were prepared by melt extrusion and injection moulding, and the scCO2 extraction parameters, including extractor configuration, extraction time, pressure, temperature, type of co-solvents and physical dimensions of the sample, were optimized. To further remove the BFRs, samples were then cleaned in a pressure extractor with isopropanol (hybrid approach), achieving high BFR removal efficiency. The elemental bromine and BFR concentrations in the samples, both before and after the scCO2 extraction, were examined by X-ray fluorescence (XRF) analysis and mass spectrometry, respectively. In a semi-continuous configuration for the scCO2 extractor at 100 °C and ethanol co-solvent we effectively removed all three BFRs achieving a maximum bromine extraction of 84.3% from ABS-TBBPA samples during 240 min. In the hybrid approach, the extraction increased to 89.4% by treating the samples in the pressure extractor
MITE: the Minimum Information about a Tailoring Enzyme database for capturing specialized metabolite biosynthesis
No full textSecondary or specialized metabolites show extraordinary structural diversity and potent biological activities relevant for clinical and industrial applications. The biosynthesis of these metabolites usually starts with the assembly of a core ‘scaffold’, which is subsequently modified by tailoring enzymes to define the molecule’s final structure and, in turn, its biological activity profile. Knowledge about reaction and substrate specificity of tailoring enzymes is essential for understanding and computationally predicting metabolite biosynthesis, but this information is usually scattered in the literature. Here, we present MITE, the Minimum Information about a Tailoring Enzyme database. MITE employs a comprehensive set of parameters to annotate tailoring enzymes, defining substrate and reaction specificity by the expressive reaction SMARTS (Simplified Molecular Input Line Entry System Arbitrary Target Specification) chemical pattern language. Both human and machine readable, MITE can be used as a knowledge base, for in silico biosynthesis, or to train machine-learning applications, and tightly integrates with existing resources. Designed as a community-driven and open resource, MITE employs a rolling release model of data curation and expert review. MITE is freely accessible at https://mite.bioinformatics.nl/
Attention-driven refinement network for continuity-preserving airway segmentation in class-imbalanced CT
No full textTubular airway segmentation is a prerequisite for bronchoscopic intervention in treating pulmonary diseases. Training convolutional neural networks (CNNs) for airway segmentation remains a clinical challenge due to the local discontinuities and distal small airway leakages caused by low resolutions and severe data imbalances. To address these issues, we propose an attention-driven refinement network, based on the degree of feature contribution, to improve the performance of fine-grained airway segmentation. A pointwise feature recalibration (PWFR) module is first designed to implement a differential feature treatment strategy by emphasizing competitive features and continuously suppressing redundant features, highlighting the prominence of airways in the learning task. Furthermore, a novel attention-driven knowledge distillation (AttdKD) module is developed to fully integrate the spatial and channel knowledge at various stages of the network, which strengthens the focus on distal small airways under conditions of class imbalance and mitigates the local discontinuity problem. The segmentation visualization results indicate that our refinement network effectively improves the thin airway recognition rate and improves the overall continuity of the airways under the guidance of the PWFR and AttdKD modules. The branches detected (BD) and tree length detected (TD) achieved scores of 93.96 %/81.7 % and 92.71 %/79.9 % on the ATM’22 and EXACT’09 datasets, respectively, and obtained scores of 92.72 %/92.44 % and 92.16 %/91.97 % on the abnormal case test sets of COVID-19 and Fibrosis, respectively. Extensive experiments demonstrate that our proposed method exhibits excellent sensitivity to distal small airways and achieves notable overall segmentation performance compared to the state-of-the-art (SOTA) baselines.</p
Integrating nuclear Small Modular Reactors into low-carbon energy systems:an illustration using a recent European R&D initiative
No full textThe race to develop Small Modular Reactors (SMRs) is in full swing around the world. SMRs are nuclear reactors with a power output of a few hundred MWe incorporating high modularisation and standardisation by design, thus facilitating economies of in-series production. SMR technologies have the potential to strongly contribute to decarbonisation of the energy sector but are yet to be deployed. Considered at a local or regional scale, SMRs can be fully integrated in innovative hybrid energy systems (HES), including variable renewables and nuclear energy in the form of electricity, heat or hydrogen, energy storage systems, heat networks, and power grids. These systems must operate flexibly to ensure the stability of energy networks. These integrated energy systems are currently under development, however, in Europe, studies on such systems remain limited. In this context, a European Industrial Alliance on Small Modular Reactors, launched by the European Commission in 2024, pointed out significant R&D gaps to be tackled to make these energy systems ready for deployment. Therefore, TANDEM, a Euratom-funded project was carried out between 2022 and 2025 to help fill these gaps. The project has delivered methodologies and tools for the assessment of HES and validated and demonstrated them on case studies for decarbonisation. The project enabled first evaluation considerations of the technical performance and economic viability of such systems. It then covered nuclear safety aspects and environmental impact. Finally, it investigated citizen engagement and Education & Training needs to prepare the workforce required for developing and deploying these energy systems.</p
Enabling cryogenic technologies for superconducting quantum devices
No full textLow-temperature refrigerators cool systems down to cryogenic temperatures near absolute zero, where thermal noise and decoherence are suppressed. This allows quantum phases, such as superconductivity, to emerge in certain materials and enables the harnessing of individual quantum states for scientific and high-performance applications. However, the refrigerators used for these purposes are large and rely on cryoliquids, such as scarce and expensive 3He, which can be a limiting factor depending on the technological application. To enable more scalable, costeffective cryogenic platforms, new refrigeration technologies must be developed. To this end, chip-scale coolers based on superconducting tunnel junctions have been envisioned to provide a fully solid-state alternative. Proof-of-principle operation of these coolers has been demonstrated at temperatures below 1.5 K, but to link them with commercially available 4He pulse tube cryocoolers, stage operating above 2.0 K is required. Additionally, thermally isolating and electrically conducting methods are needed to cascade coolers operating at different temperature ranges. In this thesis, the fundamental components of a multi-stage chip-scale cooler operating at temperatures compatible with 4He pulse tube cryocoolers are developed. A superconducting flipchip assembly fabricated with In-bumps was characterized in the sub-kelvin temperature range, and the inter-chip thermal resistance was found to be suitable for chip-scale cooling applications. A through-chip signal routing method utilizing ALD TiN-based TSVs was developed, and the demonstrated critical temperature of 2.0 K enables dissipationless DC transport for multi-chip assemblies, such as cascaded coolers. Additionally, ALD MoCx was shown to exhibit a superconducting transition temperature up to 4.4 K and high conformality, showing promise as a TSV-compatible material. The key achievement of electronic cooling of Al thin film from a bath temperature of 2.4 K down to 1.6 K was demonstrated using Nb-based superconducting tunnel junctions, probed by an onchip junction thermometer. Thermal model calculations highlighted the emergence of superconductivity in the Al beneath the cooler junctions, persisting up to a bath temperature of 2.4 K: one kelvin higher than the nominal critical temperature of the Al thin film. The single-stage cooler operating above 2.0 K enables solid-state on-chip cooling from 4He pulse-tube compatible temperature without the use of magnetic fields. Additionally, Al- and V-based tunnel junctions were fabricated at the wafer scale using degenerately doped Si as the normal electrode. The junctions exhibited suitable low-temperature electrical characteristics for cooling applications. From superconducting interconnects to tunnel-junction components supporting high cooling power density above 1 K, the achievements presented in this thesis enable modular design of chip-scale cascade coolers. This technology is envisioned to support the scaling of several superconducting quantum devices from proof-of-principle to multi-component systems beyond experimental lab environments
Exploring reference standards for the measurement of respiratory rate in children under 5 years of age:a scoping review
No full textBackground Respiratory rate is an important part of assessing the clinical state of children, and various methods exist to measure it. However, there is a lack of a universally accepted reference standard to validate the performance of these methods. Aim To identify different reference standards that have been used to evaluate respiratory rate measurement methods in children under 5 years of age and describe their perceived strengths and limitations. Methods MEDLINE and Web of Science were searched for studies in English. Studies of children under 5 years of age, published between 2013 and 2024, in which a method for measuring respiratory rate was compared against a reference standard, were included. Deductive content analysis was used to map perceived strengths and limitations of each standard, and a forest plot analysis was used to compare agreement between the reference standard and the index tests. Results From 992 retrieved studies, 56 were included. The most common reference standard was impedance pneumography (22/56), primarily used in high-income settings, followed by manual counting (19/56), mostly employed in low- and middle-income settings, and capnography (9/56). Child age, clinical condition, setting, training of personnel and the ease of implementation were all important factors in which the reference standard was used and how it performed. Conclusion Three different reference standards were used for most studies; however, their relative performance to each other is unclear. There is a need for research that directly compares the performance of these reference standards across different age strata and settings in order to confidently recommend a reference standard for respiratory rate measurement methods.</p
An integrated approach to structure, texture and nutritional quality in high-moisture extruded meat analogues from faba protein concentrate and single-cell proteins
No full textThis study evaluated meat analogues using high-moisture extrusion (HME) using faba protein concentrate (FPC) alone (Control) and blends with single-cell proteins (SCPs): microalgae Chlorella vulgaris (SCP1) and bacteria Xanthobacter spp. (SCP2). Three blends were formulated via linear programming based on the beneficial nutrients content in meat (beef, pork and chicken): Blend1 (60% FPC + 40% SCP1), Blend2 (22.5% FPC + 77.5% SCP2), and Blend3 (13.5% FPC + 11% SCP1 + 75.5% SCP2). Composition, texture, phytic acid and in vitro digestibility analyses assessed protein quality and mineral bioaccessibility. Samples were oven cooked before assays to simulate typical consumption. Cooking caused minor structural changes, without significantly affecting protein denaturation or phytic acid levels, as extrusion was the dominant thermal process. Protein digestibility was high (close to 100%) across all samples and generally unaffected by cooking. SCP inclusion significantly improved amino acid profiles, with Blend1 and Blend2 classified as excellent sources and Control and Blend3 as good sources of essential amino acids. Minerals such as manganese and potassium showed enhanced bioaccessibility linked to reduced phytic acid levels due to SCP incorporation and extrusion. Compared to average meat and dietary reference values, extruded blends demonstrated promising nutritional equivalency, supporting their potential as sustainable, nutrient-dense meat analogues. This study highlights the benefit of combining alternative protein blends with high-impact extrusion to enhance meat substitute nutritional quality
Decoding acceptance of driver monitoring systems:Evaluating alternative measurement models, cross-country variations, and behavioural intention
No full textDriver monitoring systems (DMS) demonstrate significant potential for enhancing road safety. It is imperative to comprehend potential users’ attitudes towards DMS to optimise their benefits and increase public acceptance. This study investigates potential users’ acceptance of DMS in conditionally automated driving systems (SAE level 3) by evaluating alternative measurement models and assessing cross-country variations across nine countries (i.e., Germany, Spain, France, Japan, Poland, Sweden, the United Kingdom, the United States, and China). Utilising survey data from 9025 drivers, we compared the principal component analysis and the four models (a single-factor model, a six factors model, a two higher-order factors model, and a two lower-order factors model) via structural equation modelling. A model with two correlated factors, General Acceptance and Concerns, emerged as the optimal solution with high reliability across constructs. Significant cross-country differences in all constructs were found, although only 0.3% of the variance in behavioural intention was attributable to country-level differences. A linear mixed model demonstrated that the general acceptance factor positively related to behavioural intention, whereas concerns had a small but significant negative effect. The implications for research and practice suggest that while individual-level perceptions are paramount, country context also plays a role, albeit a modest one, in shaping users’ willingness to adopt DMS technologies.</p