Research of the LDA Algorithm Results for Patents Texts Processing
Alla Grigorievna Kravets, Sergey S. Vasiliev, Dmitriy Shabanov · 2018
The aim of the article is to study the LDA algorithm results for different classes of patents. We studied similarity of extractable topics from “chemical” (C08 and A6l) and “physical” (H01 and G06) classes of patents. The optimal number of topics can be selected from the interpretation of the resulting topics for the coherence of words in the topic and the reflection of the general discourse. In presented datasets only general topics are known, is not possible to suggest which sub-topics can discover. In the course of the research, the dynamics of the change in the models' quality with the change of parameters, according to which relatively optimal parameters are chosen., is considered. However., the question of model optimization requires more detailed consideration.