NLP for the Wine Connesuer
The content of your blog post goes here.Opinions start, revolve around, (and ultimately hopes to build upon) word choice. Is it possible to really know what someone is thinking based on the words they use? Can you teach a computer to understand sentiment? What if personal vocabulary labels your speech as an outlier? Is there such a thing? This project hopes to explore some of these questions using wine reviews and seek to answer the question if wine taste can be quantified.
<img src="https://imgur.com/pDqPdx1.png" title="Close Words" width=70%/>
These reviews are always one sentence in length, describing the subtleties of wine flavors. The vocabulary is such in the dataset, that a word like ‘brimstone’ would be most closely associated with earthy tastes like; ‘peat’, ‘chamomile’, ‘beeswax’, and ‘granite’. These reviews also give some clue as to how the wine tasters will score the wine with words like ‘excellent’ to describe a high scoring wine or universally bad descriptors like ‘abrasive’ for the under performing wines.
Wine critics are an interesting case study to me, because they have to power to polarize people towards or against something using only their words. These critics are of the fairly toothless variety. They might give a scathing review but still offer a score of 80/100 for the wine. Wine Enthusiast is, after all, selling ad space in the magazine. With this in mind we are justifed in normalizing the target variable and transforming that data to a bivariate outcome split down the mean average. Scores were either a 0 "bad" or a 1 "good".
Sol-Sim - Solar Energy Ethos
Sol-Sim is a sunlight simulation that was developed to clear the cloudy skies obscuring the high skies of solar energy. This educational tool should help people gain more confidence in an eventual commitment or simply educate more people about solar energy. Sol-Sim is an honest attempt to simulate the output of a solar panel without any expectations that the user invest and so hopefully can give an unbiased look into some of the key factors which can determine the effectiveness of a solar array.
currently you can view this simulator here
Northwind Traders:
Hypothesis Testing Across the 7 Seas
Dartmouth Wilderness Trout Tagging
The following notebook is a work in progress for the data collected form trout spawning waters in nothern New Hampshire.
RegEx For Text
Regular Expression, or RegEx, is a sequence of charaters that form a search pattern.
Regular Expression can be used to check if a string contains the specified search pattern. RegEx is like a code or a group of symbols used to filter through strings and this brief tutorial will not even cover the depth of it, but it will serve as an introduction to the concept.