I have an academic project to build an application to determine the “feeling” of consumer toward brands: whether it’s positive, negative, or neutral. However, I’m stuck with no idea how to get the sample on how to “read” the English with code. For example:
Mrs. Darlie buy me an apple 🙂
Contrast to the following sentence:
I Liked Darlie so much and it wake me up every early morning
From human common sense, we can notice that the first sentence is about a “human”. First person describing the lady named Mrs. Darlie, but in my case I want to focus on getting brand names only, which means the first sentence should be dropped, focusing on the second sentence instead. From this sentence, we can determine that it’s an object while most likely describing the toothpaste we’re using every early morning.
How can I build this “common sense” with Python? (If there’s any example in PHP, I could try to convert it into Python) or is there any library available? Alternatively, please suggest some keywords for research.
Thanks to Tom De Smedt & Walter Daelemans from CLiPS Computational Linguistics Group, University of Antwerp. They solved my questioned in this Journal with example & entire source code on This web page.
The concept of this application is first crawling related sentences from social website or any public websites, then backend linked to the Wordnet.org english database to determine the “type” or “category” of the words from the sentences. Next process it with classification technique. it’s AWESOME!!
Thanks for reply. and hope this help others 🙂