library(lubridate)
library(tidyverse)

dt <- readxl::read_excel('reviews.xlsx')
dt <- dt %>% mutate(date = ymd(date), roundDat= round_date(date,'month'))

Total reviews:

nrow(dt)

## [1] 2473

ggplot(data = dt, aes(x = round_date(date, '4 months'), y = rating)) + 
  stat_summary(fun.y="mean", geom="line") +
  ylim(0,5) +
  ggtitle("NINA: Mean score (4-months)") +
  scale_x_date(name = 'date', date_labels = '%b %y', breaks = '8 month')

ggplot(data = dt) +
  geom_histogram(aes(x = date), bins = 120) +
  scale_x_date(date_labels = '%b %y', breaks = '8 month') +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  ggtitle('Number of reviews by date')

We can also extract the review for people who reviewed more than once

dt_double <- dt %>% count(names) %>% arrange(desc(n)) %>% filter(between(n , 2, 5)) %>% left_join(dt, by = 'names')
dt_double

## # A tibble: 25 x 7
##    names        n date       review          helpfulVotes rating roundDat  
##    <chr>    <int> <date>     <chr>           <chr>         <dbl> <date>    
##  1 Michael…     3 2016-07-27 Suche Möglichk… 0                 4 2016-08-01
##  2 Michael…     3 2016-07-22 Kostet die Kom… 28                1 2016-08-01
##  3 Michael…     3 2016-01-22 Eigentlich ein… 3                 2 2016-02-01
##  4 B. S.        2 2017-02-05 Die App unters… 0                 1 2017-02-01
##  5 B. S.        2 2016-07-22 Keine Verbindu… 1                 1 2016-08-01
##  6 Christi…     2 2018-01-03 Warnt viel zu … 3                 1 2018-01-01
##  7 Christi…     2 2015-09-20 Testwarnung ha… 0                 1 2015-10-01
##  8 Christi…     2 2017-07-23 Grösste Gewitt… 1                 1 2017-08-01
##  9 Christi…     2 2016-07-23 Pushnachrichte… 0                 2 2016-08-01
## 10 Denis S…     2 2019-05-12 Top. Informier… 2                 5 2019-05-01
## # … with 15 more rows