Automatic Building of a Semantic Model of Disease Symptoms Based on Text Corpus
Journal Title: Computer Science and Mathematical Modelling - Year 2014, Vol 0, Issue 14
Abstract
The research described in the article refers to the study of data from the domain of medicine. The diagnostic test results are recorded in different ways. They may take the form of tables, graphs or images. Regardless of the original data format, it is possible to draw up their verbal description, which focuses on the description of the observed symptoms. Such descriptions make up the text corpora concerning individual diagnostic technologies. Knowledge on disease entities is stored in a similar manner. It has the form of text corpora, which contain descriptions of symptoms specific to individual diseases. By using natural language processing tools semantic models can be automatically extracted from the texts to describe particular diagnostic technologies and diseases. One of the obstacles is the fact that medical knowledge can be written in a natural language in many ways. The application of the semantic format allows the elimination of record ambiguities. Ultimately, we get a unified model of medical knowledge, both from the results of diagnostic technologies describing the state of the patient and knowledge of disease entities. This gives the possibility of merging data from different sources (heterogeneous data) to a homogeneous form. The article presents a method of generating a semantic model of medical knowledge, using lexical analysis of text corpora.
Authors and Affiliations
Grażyna Szostek, Marek Jaszuk, Andrzej Walczak
Keywords
Related Articles
- EP ID EP63206
- DOI 10.5604/15084183.1152151
- Views 117
- Downloads 0
Multiple criteria mechanisms supporting clinical decisions making in the model of the repository based on the patterns
In this paper opportunities of usage diagnostics conclusion mechanisms based on patterns defined in multi criteria patient’s medical data space were introduced. Such mechanisms might be used in medical resolve’s supporti...
Inżynieria wymagań w metodach Agile
„Zwinne” wytwarzanie oprogramowania (Agile Software Development) stało się bardzo popularne na przestrzeni kilku ostatnich lat. Metody Agile zostały wymyślone w celu szybszego dostarczenia działającego oprogramowania do...
Własności bezsiatkowej metody symulacji dyfuzji ze źródła skończonego
W pracy przedstawiono analizę właściwości algorytmu numerycznego symulacji dyfuzji ze źródła skończonego opartego o bezsiatkową metodę numeryczną. Wykazano zależności czasowe symulacji oraz przedstawiono błędy symulacji....
Ensemble Methods for Improving Classification of Data Produced by Latent Dirichlet Allocation
Topic models are very popular methods of text analysis. The most popular algorithm for topic modelling is LDA (Latent Dirichlet Allocation). Recently, many new methods were proposed, that enable the usage of this model...
Problem eliminowania fałszywych alarmów w komputerowych systemach ochrony peryferyjnej
W artykule rozpatruje się problem ochrony obiektów powierzchniowych za pomocą komputerowego systemu ochrony peryferyjnej, sterującego ochroną utworzoną w postaci barykady z wmontowanymi w nią czujkami naciągowymi. Analiz...