RKRoman Klapaukh

Roman Klapaukh

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Roman Klapaukh has a total of 27 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    CensorCheck: A Tool for Evaluating Protected Health Information Detection Systems

    by Roman Klapaukh, Douglas Boyle

    Identifying protected health identifiers (PHIs) in free text is core to anonymising clinical notes for research. There is no standard metric for the quality of a PHI detection algorithm, resulting in papers reporting heterogeneous results. CensorCheck is presented, which computes specialised metrics for evaluating PHI detection taking into account the peculiarities of anonymisation tasks.


  • Open Access

    Deployment of a Free-Text Analytics Platform at a UK National Health Service Research Hospital: CogStack at University College London Hospitals

    by Kawsar Noor, Lukasz Roguski, Xi Bai, Alex Handy, Roman Klapaukh, Amos Folarin, Luis Romao, Joshua Matteson, Nathan Lea, Leilei Zhu, et al.

    Background: As more health care organizations transition to using electronic health record (EHR) systems, it is important for these organizations to maximize the secondary use of their data to support service improvement and clinical research. These organizations will find it challenging to have systems capable of harnessing the unstructured data fields in the record (clinical notes, letters, etc) and more practically have such systems interact with all of the hospital data systems (legacy and current). Objective: We describe the deployment of the EHR interfacing information extraction and retrieval platform CogStack at University College London Hospitals (UCLH). Methods: At UCLH, we have deployed the CogStack platform, an information retrieval platform with natural language processing capabilities. The platform addresses the problem of data ingestion and harmonization from multiple data sources using the Apache NiFi module for managing complex data flows. The platform also facilitates the extraction of structured data from free-text records through use of the MedCAT natural language processing library. Finally, data science tools are made available to support data scientists and the development of downstream applications dependent upon data ingested and analyzed by CogStack. Results: The platform has been deployed at the hospital, and in particular, it has facilitated a number of research and service evaluation projects. To date, we have processed over 30 million records, and the insights produced from CogStack have informed a number of clinical research use cases at the hospital. Conclusions: The CogStack platform can be configured to handle the data ingestion and harmonization challenges faced by a hospital. More importantly, the platform enables the hospital to unlock important clinical information from the unstructured portion of the record using natural language processing technology.


  • Going round the twist—an empirical analysis of shell coiling in helicospiral gastropods

    by Katie S. Collins, Roman Klapaukh, James S. Crampton, Michael F. Gazley, C. Ian Schipper, Anton Maksimenko, Benjamin R. Hines

    The logarithmic helicospiral has been the most widely accepted model of regularly coiled molluscan form since it was proposed by Moseley and popularized by Thompson and Raup. It is based on an explicit assumption that shells are isometric and grow exponentially, and an implicit assumption that the external form of the shell follows the internal shape, which implies that the parameters of the spiral could be reconstructed from the external whorl profile. In this contribution, we show that these assumptions fail on all 25 gastropod species we examine. Using a dataset of 176 fossil and modern gastropod shells, we construct an empirical morphospace of coiling using the parameters of whorl expansion rate, translation rate, and rate of increasing distance from coiling axis, plus rate of aperture shape change, from their best-fit models. We present a case study of change in shell form through geologic time in the austral family Struthiolariidae to demonstrate the utility of our approach for evolutionary paleobiology. We fit various functions to the shell-coiling parameters to demonstrate that the best morphological model is not the same for each parameter. We present a set of R routines that will calculate helicospiral parameters from sagittal sections through coiled shells and allow workers to compare models and choose appropriate sets of parameters for their own datasets. Shell-form parameters in the Struthiolariidae highlight a hitherto neglected hypothesis of relationship between Antarctic Perissodonta and the enigmatic Australian genus Tylospira that fits the biogeographic and stratigraphic distribution of both genera.


  • Open Access

    Intelligently learning from data

    by Edward Palmer, The Inform-Lab, Roman Klapaukh, Steve Harris, Mervyn Singer


  • Open Access

    Comparing Graph Layouts for Vertex Selection Tasks

    by Roman Klapaukh, David J. Pearce, Stuart Marshall

    Different graph layouts can affect a user's ability to complete both passive understanding and active interaction tasks. While most research exploring the effects of graph layout looks at a user's ability to accomplish a passive understanding task, this paper's novel contribution is looking at their ability to complete a selection task. Specifically we compare two graph layout algorithms with respect to their suitability for free-form multi-selection. The two layout algorithms are drawn from our previous work which established that they have significantly different understandability metric scores. A motivation for this choice was to explore whether graphs with significantly different metric scores will also have significantly different performance for selection tasks. We carried out our comparison by means of a user experiment that followed a within-subjects design, where 74 users were given a PlayStation Move controller to select vertices in 20 pairs of graphs. We found that while there was no difference in the speed of interaction there was a difference in the number of errors users made between the two layout algorithms.