Search

Developing a drug from scratch is a very long and expensive process that has a small probability of success. For this reason, pharmaceutical companies are devoting their efforts to find drugs that could be repositioned. When using a drug to treat a disease is necessary to consider what adverse reactions it may cause, this is why the prediction of adverse reactions is highly related to drug repositioning. We propose the detection of overlapping communities over a biological network of chemicals, diseases and genes in order to find drug-disease pairs that could be used as basis for later drug repositioning and adverse reactions prediction analysis. Of the evaluated overlapping community detection algorithms, OSLOM got the best results, producing 724 communities from which was possible to extract 215944 drug-disease pairs not present in the analyzed graph. We illustrate the usefulness of this set through examples of associations between pairs found in the scientific literature. [ABSTRACT FROM AUTHOR]

Copyright of Procesamiento del Lenguaje Natural is the property of Sociedad Espanola para el Procesamiento del Lenguaje Natural and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This paper proposes a comprehensive evaluation of supervised classification algorithms without semantic information for opinion mining on Twitter. The evaluation will assess the potential of this approach to tackle the given problem. Also, as result of evaluating, the impact of different variants of the text preprocessing is assessed, as well as different alternatives of dimension reduction techniques in the problem solution. Finally, the results obtained in the experiments offer evidence to direct the use of supervised classification to classify opinions and its valence in a domain of short messages such as Twitter. [ABSTRACT FROM AUTHOR]