Toine Pieters - Academia.edu (original) (raw)

Papers by Toine Pieters

Information

We adapt previous literature on search tasks for developing a domain-specific search engine that ... more We adapt previous literature on search tasks for developing a domain-specific search engine that supports the search tasks of policy workers. To characterise the search tasks we conducted two rounds of interviews with policy workers at the municipality of Utrecht, and found that they face different challenges depending on the complexity of the task. During simple tasks, policy workers face information overload and time pressures, especially during web-based searches. For complex tasks, users prefer finding domain experts within their organisation to obtain the necessary information, which requires a different type of search functionality. To support simple tasks, we developed a web search engine that indexes web pages from authoritative sources only. We tested the hypothesis that users prefer expert search over web search for complex tasks and found that supporting complex tasks requires integrating functionality that enables finding internal experts within the broader web search en...

Springer eBooks, 2022

Domain specialists such as council members may benefit from specialised search functionality, but... more Domain specialists such as council members may benefit from specialised search functionality, but it is unclear how to formalise the search requirements when developing a search system. We adapt a faceted task model for the purpose of characterising the tasks of a target user group. We first identify which task facets council members use to describe their tasks, then characterise council member tasks based on those facets. Finally, we discuss the design implications of these tasks for the development of a search engine. Based on two studies at the same municipality we identified a set of task facets and used these to characterise the tasks of council members. By coding how council members describe their tasks we identified five task facets: the task objective, topic aspect, information source, retrieval unit, and task specificity. We then performed a third study at a second municipality where we found our results were consistent. We then discuss design implications of these tasks because the task model has implications for 1) how information should be modelled, and 2) how information can be presented in context, and it provides implicit suggestions for 3) how users want to interact with information. Our work is a step towards better understanding the search requirements of target user groups within an organisation. A task model enables organisations developing search systems to better prioritise where they should invest in new technology.

Zenodo (CERN European Organization for Nuclear Research), May 4, 2022

In this paper we present the currently running PDI-SSH project Homo Medicinalis (HoMed), in which... more In this paper we present the currently running PDI-SSH project Homo Medicinalis (HoMed), in which we use machine learning to build an Automatic Speech Recognition (ASR) infrastructure for disclosing privacy-sensitive doctor-patient consultation recordings. ASR using machine learning has been described as being in a 'golden age' (Vipperla 2020: 9). Advances in ASR technology make it possible to search culturally significant digitized audiovisual data archives on a spoken word level (Ordelman and Van Hessen, 2018). This has obvious benefits for the study of any topic over time in a changing sociocultural context: the digital turn (Nicholson, 2013) that made it possible to study explore millions of newspaper pages using distant reading, might now be followed by a cross-media turn as our audiovisual cultural heritage data archives can be explored in the same manner using ASR (Van der Molen, 2022: 202). However, ASR performance decreases when, acoustically (recording conditions) and/or semantically, training and target domain do not match. Most of the current out-of-the-box commercial systems are intended for generic use (e.g. standard-Dutch, relatively good recording conditions), targeting the natural conversational speech scenario, lacking in technical terms and jargon (Litman et. al, 2018). Therefore, there is an evident need for improving ASR performance when used on in-house data-specific scenarios (Tejedor-García et. al, 2021). In HoMed we aim to do that in our work towards enabling such research beyond the public media data domain to audiovisual recordings from the highly privacy-sensitive medical-professional domain. Our understanding of the use and reputation of medicines can then be based on its coverage in the public sphere (e.g. newspaper, radio, television), but also on the way in which it has been discussed in the professional medical sphere by both medical professionals and patients (doctor-patient recordings). In this project, the audiovisual health care related data collected over the years at the Netherlands Institute for Health Services Research (Nivel), are used to retrain the currently largest open-source general-purpose ASR system for Dutch (Kaldi_NL) for the automatic

History of Pharmacy and Pharmaceuticals

Cancer Imaging : the Official Publication of the International Cancer Imaging Society, 2004

Interferon: The Dawn of Recombinant Protein Drugs, 1999

Numerous anticancer therapies (radiotherapy, nitrogen mustard, anti-folic acid antagonists) have ... more Numerous anticancer therapies (radiotherapy, nitrogen mustard, anti-folic acid antagonists) have been developed following accidental clinical observations. The antitumor properties of other compounds have been revealed through systematic screening programs, such as the natural substances screening program of the National Cancer Institute. The history of these therapies is relatively short, and it includes original clinical or biological observations, usually (but not necessarily) tests in laboratory animals, and finally trials in patients. The story of the introduction of interleukin-2 into cancer therapy is different. It is linked with the history of immunology of cancer, an enterprise which started more than a 100 years ago and has been rich in abandoned trails. The introduction of interleukin-2 to cancer therapy was the result of a partly contingent encounter between several distinct lines of investigation. My paper follows the complex history of the attempts to develop immunotherapies of cancer, and stresses the intertwining — and the separation — of different strands of inquiry. It illustrates the early failures to transform promising observations about the putative role of immune mechanisms in the control of malignant growths into accepted therapies of human cancer. These early attempts are contrasted with the successful development of recombinant protein drugs in the 1980s. The increased scale and scope of biomedical investigations, and the consolidation of links between laboratory studies, bedside observations, large-scale clinical trials and collaboration with the industry, led to the transformation of a substance first described as an agent which promotes the growth of immunocompetent cells in the test tube into an anticancer drug.

The Routledge History of Madness and Mental Health, 2017

Nederlands Tijdschrift voor Geneeskunde, 2013

Transboundary and Emerging Diseases, 2019

This article has been accepted for publication and undergone full peer review but has not been th... more This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as

Scientific Reports, 2019

Leprosy is an infectious disease caused by Mycobacterium leprae affecting the skin and nerves. De... more Leprosy is an infectious disease caused by Mycobacterium leprae affecting the skin and nerves. Despite decades of availability of adequate treatment, transmission is unabated and transmission routes are not completely understood. Despite the general assumption that untreated M. leprae infected humans represent the major source of transmission, scarce reports indicate that environmental sources could also play a role as a reservoir. We investigated whether M. leprae DNA is present in soil of regions where leprosy is endemic or areas with possible animal reservoirs (armadillos and red squirrels). Soil samples (n = 73) were collected in Bangladesh, Suriname and the British Isles. Presence of M. leprae DNA was determined by RLEP PCR and genotypes were further identified by Sanger sequencing. M. leprae DNA was identified in 16.0% of soil from houses of leprosy patients (Bangladesh), in 10.7% from armadillos’ holes (Suriname) and in 5% from the habitat of lepromatous red squirrels (Britis...