README.md

In JCRascal/opiniontext:

opiniontext

Overview

opiniontext is an ongoing project to build a data package containing text of US Supreme Court opinions available publicly at SupremeCourt.gov. Text data is to be made readily available in the form of text documents, csv, and R objects. Csv and R objects outputted will contain the following columns:

• Case Name
• Date
• Opinion Type
• Author
• Text

Installation

You can install the development version of opiniontext from GitHub with:

# install.packages("devtools")
devtools::install_github("JCRascal/opiniontext")

Sample Usage

library(opiniontext)
library(tidytext)
library(dplyr)
library(ggplot2)

case_words <- opinions_2019 %>%
  filter(author != "Per Curiam") %>%
  filter(author != "Syllabus") %>%
  unnest_tokens(word, text) %>%
  anti_join(get_stopwords(), by = "word") %>%
  count(author, word, sort = TRUE)

case_words <- case_words %>%
  group_by(author) %>%
  summarize(total = sum(n)) %>%
  right_join(case_words) %>%
  filter(nchar(word) > 2)

case_plot <- case_words %>%
  bind_tf_idf(word, author, n) %>%
  group_by(author) %>%
  slice_max(tf_idf, n = 10) %>%
  ungroup() %>%
  mutate(word = reorder_within(word, tf_idf, author))

ggplot(case_plot, aes(word, tf_idf, fill = author)) +
  geom_col(show.legend = FALSE) +
  labs(x = NULL, y = "tf-idf") +
  facet_wrap(~author, ncol = 4, scales = "free") +
  coord_flip() +
  scale_x_reordered()

Data Collection Plan

• ~~Phase 1: Collect all available slip opinions for the 2019 session~~
• Phase 2: Collect all available slip opinions from 2014 session to present
• Phase 3: Collect official bound volumes

JCRascal/opiniontext documentation built on Jan. 19, 2021, 6:30 p.m.