Probabilistic aspect mining approach for interpretation and evaluation of drug reviews
Tushar N Tekade, M. Emmanuel · 2016
Evidence Based Medicine (EBM) is the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. Extracting Information from considerable bodies of text is useful & challenging. A generative probabilistic aspect mining model is used here to study the collection. Frequency based approach in aspect based opinion mining extracts high frequency noun phrases and relation based approach identifies aspects based on the aspect-sentiment relation in reviews. However for drug reviews the aspects are not indicated explicitly by authors and description of side effects and people's experience is diverse. Topic Modeling identifies aspects based on the co-occurrence of words in reviews. Here in this research Topic Modeling Based on Probabilistic Approach a more fine grained aspect level opinion mining is used. It is interesting to apply the model to find aspects relating to different segmentation of data such as different age groups or other attributes. It is also interesting to work with aspect interpretation as aspects are now represented by a list of keywords. If a few sentences can be extracted or generated automatically to summarize the keywords, interpretation & understanding will be improved.