INTRO:

Summer is around the corner which means I won’t be able to enjoy my sense of vision without some type of spectacles. I don’t want just any pair of sunglasses, I need a pair that is durable yet has great lenses. AKA I want nothing to do with those damn Panama Jack gas station sunglasses bullsh**. Lol. No, I want oakley, or I want ray ban and nothing inbetween. These sunglass manufacturers are some of the best. My intention is to understand which customer base is generally more content with their pair. I am able to find this data through gathering tweets @ or mentioning these products. Once the tweets are gathered, I can divide the tweets by individual words and combining each word with sentiment data. More in depth analysis follows below.

Packages Used: httpuv, rtweet, tidyverse, tidytext, and textdata

Which sentiment for each company was the most prominent?

If I am going to spend probably at least $100 on sunglasses, I am curious to know which brand is more trusted, feared or joyed, ect. By extracting tweets and combinging with the nrc data, I was able to pipe the input into a bar graph. The bar chart has the count of words relating to the nrc dictionary on the y axis, and the sentiment on the x axis. I filled the senitments with company so I was easily able to compare oakley to ray ban. The sentiments that I am mostly interested in are trust, anger, positive, and negative. I think it is important for someone to trust the brand they buy and not to be angry. I know if I’m angry with sunglasses enough to take it to twitter, I probaly wouldn’t make a purchase from this brand again. From the chart, I can see that ray ban scores higher with anger and less than trust. It also has less positive and more negative words than oakley. Just with this information I can tell ray ban is going to have to make up a lot of ground to get me to buy their brand.

Since people most likely wear sunglasses in the summer, I want insight on which brand gets more positive feedback during the sunny and bright months.

This side by side bar chart was very interesting. From this data I assumed that more people wear ray bans in the summer. If you recall from the previous graph, ray ban scored lower on all the variables that were analyzed. However, in the summer months, ray ban has more negative and positive feedback than oakley; the difference from postive feedback is greater than the difference from negative feedback between the two brands. Extracting these tweets and piping them into bing sentiment analysis bar chart put me in a pickle. I was leaning heavy toward buying oakley but now it seems that ray ban get more positive feedback in the summer.

This chart shows the whole year and positive/negative feedback. It still seems as if ray ban gets more positive. However, the negative gap increased as there is now more negative feedback from rayban. From this I can conclude that rayban gets positive feedback more often in the summer months and gets more negative feedback through the entire year. This can help me reach an answer because I wear sunglasses anytime of year, especially when I drive, but knowing the difference between nonsummer and summer tells me more about each brand.

## [[1]]
## # A tibble: 86 x 4
## # Groups:   word [73]
##    word   company     n sentiment
##    <chr>  <chr>   <int> <chr>    
##  1 bad    oakley      6 anger    
##  2 bad    rayban      5 anger    
##  3 bang   rayban      1 anger    
##  4 belt   rayban      2 anger    
##  5 bloody oakley      4 anger    
##  6 bloody rayban      1 anger    
##  7 bout   rayban      2 anger    
##  8 broken rayban      1 anger    
##  9 cancer oakley      1 anger    
## 10 court  oakley      1 anger    
## # … with 76 more rows

I want to see the words people were using when angry with their brand.

This list goes more in depth on the anger sentiment. I chose this because I want to see which words people were using when angry with their brand. For oakley, the words: damn, bad, bloody, and disapointed were some of the most reocurring. For ray ban the words most commonly used with anger were: bad, damn, shit, money, hate, hell, and losing. These types of words help me in choosing a pair because I can interpret what people are getting angry about. For ray ban example, when someone is angry and uses the word ‘losing’, they are probably angry with themsleves for losing their glasses. On the other hand, when people are using the word ‘disappointed’, like for oakley, this tells me that a lot of people were expecting the product to be better. From this analysis, I can conclude that more people were angry with rayban than oakley, and they used more angry words to describe. Since disappointed was only used twice, that word was not even that significant. This list helped get me to my result.

RESULT:

I am going to buy a pair of oakley sunglasses. My analysis, especially as of the last week or so, has proven that people trust oakley more. Others also give oakley more positive feedback and are less angry with this brand. I am also a somewhat active person and oakley makes durable sunglasses for playying sports and other outdoor activities. I probably prefered oakley before this analysis but my results reinforced my original thoughts. In the future, I would like to see the senitment analysis comparison between oakley and spy since these brands are a little more similar.