Extraction and Classification of App Features from App Reviews

Elizaveta Yankovskaya
The number of tools for bioinformatics is constantly increasing. To organize the available information and to facilitate the search, different ontologies are used. Today annotation of new descriptions is done manually, which is time-consuming and not always correct. We proposed a new annotation method, which, based on the description of the tool, offers one or more annotation labels in accordance with the ontology. In our method, we applied modern methods of natural language processing, such as latent Dirichlet allocation and word2vec. We compared the manual annotation labels with the labels obtained by using our algorithm and the first results look auspicious.
Graduation Thesis language
Graduation Thesis type
Master - Computer Science
Sven Laur
Defence year