Aspect Based Sentiment Analysis identifies the aspect terms and the sentiment polarity associated with each aspect term in the given review. In real-time, a business can be improved with the customer feedbacks about the features in the product like “The only thing I don't understand is that the resolution of the screen isn't high enough for some pages, such as Yahoo! Mail” based on these reviews the company can get the pitfall in the product which can be fixed for more profit.
Author has just used above sample to test if function is working fine. As we browse more in code , we can see that 20 most common aspect are identified from Corpus and used further .
vineetvermaml
commented
2 years ago
Hello, I am trying to understand your code. However, I am unable to understand how did you find the prd : prd = [0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]. and the "common words": common_words = ['OS', 'Vista', 'Windows', 'Windows 7', 'applications', 'battery', 'battery life', 'carry', 'charge', 'cost', 'design', 'display', 'extended warranty', 'features', 'games', 'gaming', 'graphics', 'hard drive', 'keyboard', 'keys', 'look', 'memory', 'motherboard', 'mouse', 'operating system', 'performance', 'power', 'power supply', 'price', 'processor', 'program', 'programs', 'quality', 'runs', 'screen', 'service', 'shipping', 'size', 'software', 'speakers', 'speed', 'system', 'use', 'value', 'warranty', 'warrenty', 'weight', 'windows', 'work', 'works']
I would appreciate if you can give me a hands on that. Thanks, Emrul