• Publicationsclose
• 2019-2023close
• Open Access close
• Article close
• Publikationer från KTH close

We present an online API to access a number of Natural Language Processing services developed at KTH. The services work on Swedish text. They include tokenization, part-of-speech tagging, shallow parsing, compound word analysis, word inflection, lemmatization, spelling error detection and correction, grammar checking, and more. The services can be accessed in several ways, including a RESTful interface, direct socket communication, and premade Web forms. The services are open to anyone. The source code is also freely available making it possible to set up another server or run the tools locally. We have also evaluated the performance of several of the services and compared them to other available systems. Both the precision and the recall for the Granska grammar checker are higher than for both Microsoft Word and Google Docs. The evaluation also shows that the recall is greatly improved when combining all the grammar checking services in the API, compared to any one method, and combining services is made easy by the API. QC 20230328

FluidDyn is a project to foster open-science and open-source in the fluid dynamics community. It is thought of as a research project to channel open-source dynamics, methods and tools to do science. We propose a set of Python packages forming a framework to study fluid dynamics with different methods, in particular laboratory experiments (package fluidlab), simulations (packages fluidfft, fluidsim and fluidfoam) and data processing (package fluidimage). In the present article, we give an overview of the specialized packages of the project and then focus on the base package called fluiddyn, which contains common code used in the specialized packages. Packages fluidfft and fluidsim are described with greater detail in two companion papers [4, 5]. With the project FluidDyn, we demonstrate that specialized scientific code can be written with methods and good practices of the open-source community. The Mercurial repositories are available in Bitbucket (https://bitbucket.org/fluiddyn/). All codes are documented using Sphinx and Read the Docs, and tested with continuous integration run on Bitbucket Pipelines and Travis. To improve the reuse potential, the codes are as modular as possible, leveraging the simple object-oriented programming model of Python. All codes are also written to be highly efficient, using C++, Cython and Pythran to speedup the performance of critical functions. Funding statement: This project has indirectly benefited from funding from the foundation Simone et Cino Del Duca de l’Institut de France, the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No 647018-WATU and Euhit consortium) and the Swedish Research Council (Vetenskapsrådet): 2013-5191. We have also been able to use supercomputers of CIMENT/GRICAD, CINES/GENCI (grant 2018-A0040107567) and the Swedish National Infrastructure for Computing (SNIC).

This paper studies the problem of parameter estimation in resonant, acoustic fluid-structure interaction problems over a wide frequency range. Problems with multiple resonances are known to be subjected to local minima, which represents a major challenge in the field of parameter identification. We propose a stepwise approach consisting in subdividing the frequency spectrum such that the solution to a low-frequency subproblem serves as the starting point for the immediately higher frequency range. In the current work, two different inversion frameworks are used. The first approach is a gradient-based deterministic procedure that seeks the model parameters by minimising a cost function in the least squares sense and the second approach is a Bayesian inversion framework. The latter provides a potential way to assess the validity of the least squares estimate. In addition, it presents several advantages by providing invaluable information on the uncertainty and correlation between the estimated parameters. The methodology is illustrated on synthetic measurements with known design variables and controlled noise levels. The model problem is deliberately kept simple to allow for extensive numerical experiments to be conducted in order to investigate the nature of the local minima in full spectrum analyses and to assess the potential of the proposed method to overcome these. Numerical experiments suggest that the proposed methods may present an efficient approach to find material parameters and their uncertainty estimates with acceptable accuracy. QC 20190228

Microphysiological systems mimic the in vivo cellular ensemble and microenvironment with the goal of providing more human-like models for biopharmaceutical research. In this study, the first such model of the blood-brain barrier (BBB-on-chip) featuring both isogenic human induced pluripotent stem cell (hiPSC)-derived cells and continuous barrier integrity monitoring with <2 min temporal resolution is reported. Its capabilities are showcased in the first microphysiological study of nitrosative stress and antioxidant prophylaxis. Relying on off-stoichiometry thiol–ene–epoxy (OSTE+) for fabrication greatly facilitates assembly and sensor integration compared to the prevalent polydimethylsiloxane devices. The integrated cell–substrate endothelial resistance monitoring allows for capturing the formation and breakdown of the BBB model, which consists of cocultured hiPSC-derived endothelial-like and astrocyte-like cells. Clear cellular disruption is observed when exposing the BBB-on-chip to the nitrosative stressor linsidomine, and the barrier permeability and barrier-protective effects of the antioxidant N-acetylcysteine amide are reported. Using metabolomic network analysis reveals further drug-induced changes consistent with prior literature regarding, e.g., cysteine and glutathione involvement. A model like this opens new possibilities for drug screening studies and personalized medicine, relying solely on isogenic human-derived cells and providing high-resolution temporal readouts that can help in pharmacodynamic studies. QC 20220330

The action research methodology and several of its methods have previously been highlighted and described by the authors as a fitting and rigorous framework approach for complex waste management systems. This was in response to criticism of the ex ante selection of traditional empiric systems analysis tools to provide decision support and ‘sustainable improvement’ in such complex systems which often involve strong human and political factors. Several of the action research methods described have recently been utilized in a case study around mineral (aggregate) construction and demolition waste in the Stockholm region. These methods were integrated through a series of workshops and work areas undergone together with project members from several private and public sectors. Leaving the problem fuzzy (loosely defined) in the beginning; utilizing convergent interviewing, rich pictures and focus groups allowed the researchers and partner stakeholders to identify not one but several problem areas within the system of focus. Indicator creation and a dialectic processes were then used to identify qualitative and quantitative aspects of salience around these problem areas. These resulting indicators were strengthened through a process of verification. Each indicator was then analyzed by what was deemed to be appropriate and transparent means. It is argued that this approach may create better communication, transparency, and understanding by the stakeholders. These factors in turn allowing stronger stakeholder ownership of the process and assisting in more informed decisions and help to provide stability for desired change. However the process was not without its drawbacks such as intense communication and time requirements. QC 20120110 Project BRA

Inter-operator spectrum sharing in millimeter-wave bands has the potential of substantially increasing the spectrum utilization and providing a larger bandwidth to individual user equipment at the expense of increasing inter-operator interference. Unfortunately, traditional model-based spectrum sharing schemes make idealistic assumptions about inter-operator coordination mechanisms in terms of latency and protocol overhead, while being sensitive to missing channel state information. In this paper, we propose hybrid model-based and data-driven multi-operator spectrum sharing mechanisms, which incorporate model-based beamforming and user association complemented by data-driven model refinements. Our solution has the same computational complexity as a model-based approach but has the major advantage of having substantially less signaling overhead. We discuss how limited channel state information and quantized codebook-based beamforming affect the learning and the spectrum sharing performance. We show that the proposed hybrid sharing scheme significantly improves spectrum utilization under realistic assumptions on inter-operator coordination and channel state information acquisition.

Discharge of organic load from treated wastewater may cause environmental eutrophication. Recently, fungi have gained much attention due to their removal of pharmaceutical substances by enzymatic degradation and adsorption. However, the fungal effect in removing nutrients is less investigated. Therefore, two fungal species, the white-rot fungus T. versicolor as a laboratory strain and the mold A. luchuensis as an environmental isolate from the municipal wastewater treatment plant, were studied to determine the fungal potential for phosphorus, nitrogen, and the total organic carbon removal from municipal wastewater, carrying out a batch scale experiment to a fluidized bed pelleted bioreactor. During the batch scale experiment, the total removal (99.9 %) of phosphorus by T. versicolor was attained after a 6 hours-long incubation period while the maximal removal efficiency (99.9 %) for phosphorus from A. luchuensis was gained after an incubation period of 24 hours. Furthermore, both fungi showed that the pH adjustment to 5.5 kept the concentration of nitrogen constant and stabilized the total organic carbon reduction process for the entire incubation period. The results from the fluidized bed bioreactor demonstrated opposite tendencies on a nutrient removal comparing to a batch experiment where no significant effect on phosphorus, nitrogen, and total organics carbon reduction was observed. The obtained results from this study of batch and fluidized bed bioreactor experiments are a promising starting point for a successful fungal treatment optimization and application to wastewater treatment. QC 20210107

We envision PDS2, a decentralized data marketplace in which consumers submit their tasks to be run within the platform, on the data of willing providers. The goal of PDS2is to ensure that users maintain full control on their data and do not compromise their privacy, while being rewarded for the value that their data generates. In order to achieve this, our marketplace architecture employs blockchain technology, privacy-preserving computation and decentralized machine learning. We then compare different potential solutions and identify the Ethereum blockchain, trusted execution environments and gossip learning as the most suitable for the implementation of PDS2. We also discuss the main open challenges that are left to tackle and possible directions for future work.

Significant amounts of biomass residues were generated in Indonesia. While untreated, residues emit greenhouse gases during the decomposition process. On the other hand, if efficiently utilized, these residues could be used to produce value-added products. This study investigates opportunities for harnessing the full potential of palm oil residues (i.e., empty fruit bunches, kernel shells, fiber, and mill effluent). As far as we are aware, the study is the first attempt to model the palm oil supply chain in a geographically explicit way while considering regional infrastructures in Sumatra Island, Indonesia. The BeWhere model, a mixed integer linear programming model for energy system optimization, was used to assess the costs and benefits of optimizing the regional palm oil supply chain. Different scenarios were investigated, considering current policies and new practices leading to improved yields in small-scale plantations and power grid connectivity. The study shows that a more efficient palm oil supply chain can pave the way for the country to meet up to 50% of its national bioenergy targets by 2025, and emission reductions of up to 40 MtCO2eq/year. As much as 50% of the electricity demand in Sumatra could be met if residues are efficiently used and grid connections are available. We recommend that system improvements be done in stages. In the short to medium term, improving the smallholder plantation yield is the most optimal way to maximize regional economic gains from the palm oil industry. In the medium to long term, improving electricity grid connection to palm oil mills could bring higher economic value as excess electricity is commercialized. QC 20190130

QC 20220426