Customer Sentiment Analysis

The following document summarizes and provides insight into a Customer_Sentiment csv file.

library(readr)
customer_sentiment <- read_csv("Customer_Sentiment.csv")

library(tidyverse)

Structure

str(customer_sentiment)
## spc_tbl_ [25,000 × 13] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
##  $ customer_id         : num [1:25000] 1 2 3 4 5 6 7 8 9 10 ...
##  $ gender              : chr [1:25000] "male" "other" "female" "female" ...
##  $ age_group           : chr [1:25000] "60+" "46-60" "36-45" "18-25" ...
##  $ region              : chr [1:25000] "north" "central" "east" "central" ...
##  $ product_category    : chr [1:25000] "automobile" "books" "sports" "groceries" ...
##  $ purchase_channel    : chr [1:25000] "online" "online" "online" "online" ...
##  $ platform            : chr [1:25000] "flipkart" "swiggy instamart" "facebook marketplace" "zepto" ...
##  $ customer_rating     : num [1:25000] 1 5 1 2 3 5 4 5 3 5 ...
##  $ review_text         : chr [1:25000] "very disappointed with the quality." "fast delivery and great packaging." "very disappointed with the quality." "product stopped working after few days." ...
##  $ sentiment           : chr [1:25000] "negative" "positive" "negative" "negative" ...
##  $ response_time_hours : num [1:25000] 46 5 38 16 15 10 38 53 7 56 ...
##  $ issue_resolved      : chr [1:25000] "yes" "yes" "yes" "yes" ...
##  $ complaint_registered: chr [1:25000] "yes" "no" "yes" "yes" ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   customer_id = col_double(),
##   ..   gender = col_character(),
##   ..   age_group = col_character(),
##   ..   region = col_character(),
##   ..   product_category = col_character(),
##   ..   purchase_channel = col_character(),
##   ..   platform = col_character(),
##   ..   customer_rating = col_double(),
##   ..   review_text = col_character(),
##   ..   sentiment = col_character(),
##   ..   response_time_hours = col_double(),
##   ..   issue_resolved = col_character(),
##   ..   complaint_registered = col_character()
##   .. )
##  - attr(*, "problems")=<pointer: 0x151e1f780>

Dimension

dim(customer_sentiment)
## [1] 25000    13

Names of Variables

names(customer_sentiment)
##  [1] "customer_id"          "gender"               "age_group"           
##  [4] "region"               "product_category"     "purchase_channel"    
##  [7] "platform"             "customer_rating"      "review_text"         
## [10] "sentiment"            "response_time_hours"  "issue_resolved"      
## [13] "complaint_registered"

Summary of Customer Sentiment

summary(customer_sentiment)
##   customer_id          gender          age_group           region     
##  Min.   :    1   Length   :25000   Length   :25000   Length   :25000  
##  1st Qu.: 6251   N.unique :    3   N.unique :    5   N.unique :    5  
##  Median :12500   N.blank  :    0   N.blank  :    0   N.blank  :    0  
##  Mean   :12500   Min.nchar:    4   Min.nchar:    3   Min.nchar:    4  
##  3rd Qu.:18750   Max.nchar:    6   Max.nchar:    5   Max.nchar:    7  
##  Max.   :25000                                                        
##   product_category  purchase_channel      platform     customer_rating
##  Length   :25000   Length   :25000   Length   :25000   Min.   :1.000  
##  N.unique :    9   N.unique :    1   N.unique :   20   1st Qu.:2.000  
##  N.blank  :    0   N.blank  :    0   N.blank  :    0   Median :3.000  
##  Min.nchar:    5   Min.nchar:    6   Min.nchar:    4   Mean   :3.002  
##  Max.nchar:   14   Max.nchar:    6   Max.nchar:   20   3rd Qu.:4.000  
##                                                        Max.   :5.000  
##     review_text        sentiment     response_time_hours   issue_resolved 
##  Length   :25000   Length   :25000   Min.   : 1.00       Length   :25000  
##  N.unique :   15   N.unique :    3   1st Qu.:18.00       N.unique :    2  
##  N.blank  :    0   N.blank  :    0   Median :36.00       N.blank  :    0  
##  Min.nchar:   20   Min.nchar:    7   Mean   :36.02       Min.nchar:    2  
##  Max.nchar:   41   Max.nchar:    8   3rd Qu.:54.00       Max.nchar:    3  
##                                      Max.   :71.00                        
##  complaint_registered
##  Length   :25000     
##  N.unique :    2     
##  N.blank  :    0     
##  Min.nchar:    2     
##  Max.nchar:    3     
##

Factors of Product Categories

product_factor <- as.factor(customer_sentiment$product_category)
levels(product_factor)
## [1] "automobile"     "beauty"         "books"          "electronics"   
## [5] "fashion"        "groceries"      "home & kitchen" "sports"        
## [9] "travel"
summary(product_factor)
##     automobile         beauty          books    electronics        fashion 
##           2833           2690           2812           2725           2782 
##      groceries home & kitchen         sports         travel 
##           2858           2726           2763           2811

Frequency of Compliants for Fashion

customer_sentiment %>%
        filter(sentiment == "negative" & gender == "female") %>%
        ggplot(aes(x = product_category)) +
        geom_bar(fill = "#CD6839") +
        labs(title = "Compliants Registered by Product Category",
             x = "Product Category",
             y = "Number of Compliants"
             )

Scatterplot

ggplot(
       customer_sentiment,
       aes(x = customer_sentiment$customer_rating, fill = sentiment)
) +
        geom_bar() + 
        labs(
                title = "Customer Sentiment based on Customer Satisfaction Rating",
                x = "Customer Rating",
                y = "Customer Count"
        )