Medical Recommendation System using Co-Occurrence Graphs

Yanakorn Ruamsuk, Withawin Tirasopitlert, Anirach Mingkhwan, Herwig Unger

Abstract


Medical diagnosis [1] is the process of determining which disease explain a person’s symptoms and signs. Correct diagnosis depend on the knowledge and experience of doctors, huge amount of knowledge is distributed in textbooks and documents. it takes a huge amount of time to learn a lot of information for diagnosis, moreover, diagnosis is often challenging, because many signs and symptoms are nonspecific or some symptoms can be many of diseases. Therefor the knowledge processing must be automated. A co-occurrence graph is incorporated from sets of documents that represent knowledge, which is similar to human brain, when people read a book, the book content will become to knowledge. For co-occurrence graph, it created by information extraction from documents, the information that extract from document is read sentence by sentence and words of each sentence will create relation to each other, then the co-occurrence graph will appear and represent to brain knowledge. In this paper will presents an experiment to bring the co-occurrence graph theory to apply with medical information and create the system that contribute the diagnosis from symptoms.

Full Text:

PDF

References


Medical diagnosis, https://en.wikipedia.org/wiki/Medical_diagnosis, last accessed 2020/1/5.

Kubek, Mario & Böhme, Thomas & Unger, Herwig. (2017). Empiric Experiments with Text Representing Centroids. Lecture Notes on Information Theory. 5. 23-28. 10.18178/lnit.5.1.23-28.

L. R. Dice. Measures of the amount of ecologic association between species. Ecology, 26(3):297-302, July 1945.

Kubek, Mario & Böhme, Thomas & Unger, Herwig. (2017). Empiric Experiments with Text Representing Centroids. Lecture Notes on Information Theory. 5. 23-28. 10.18178/lnit.5.1.23-28.

A. K. Jain, M. N. Murty, and P. J. Flynn, "Data Clutering: a review" ACM Computing Survey, pp. 264-323, 1999.

Kubek, Mario & Böhme, Thomas & Unger, Herwig. (2017). Spreading activation: a fast calculation method for text centroids. 24-27. 10.1145/3162957.3163014.

Kubek, M. & Unger, H., (2012). Detecting source topics by analysing directed co-occurrence graphs. In: Eichler, G., Wienhofen, L. W., Kofod-Petersen, A. & Unger, H. (Hrsg.), 12th International Conference on Innovative Internet Community Services (I2CS 2012). Bonn: Gesellschaft für Informatik e.V.. (S. 202-211).


Refbacks

  • There are currently no refbacks.