Text Mining and Natural Language Processing in R

Learn to read online csv data into R, handle metadata, skip top rows, set headers, and access data frames before indexing and subsetting.

Read json data in r, parse world bank json files, and extract country ids and iso codes using lapply and custom functions to access regional attributes.

Scrape IMDB pages with rvest and the Selecter gadget to extract rankings, titles, runtimes, and genres, build a data frame, and gain insights from 2017 film data.

Discover another way to access and inspect elements on a dynamic New Zealand tourism page, inspecting popular cities via linked HTML pages.

Use the Guardian API in R to extract headlines and text on a topic, register for a developer key, and clean the data by removing encodings, links, and HTML tags.

Learn to extract data from Twitter in R by configuring credentials, connecting to the Twitter API, and retrieving tweets with date ranges and geolocation such as the Barcelona area.

Learn to extract tweets with the rtweet package in R by authenticating with app keys and querying English tweets on a topic for text mining.

Learn how to extract geolocated tweets with rtweet, stream London tweets for 60 seconds, convert to a data frame, and analyze text, coordinates, and trends.

Authenticate with the tweet package, search 500 users tweeting the hashtag, and plot the top locations from the location column, notably Washington, D.C.

Register your app on Foursquare developers, obtain a client id and client secret, install the orkun package from GitHub, and authenticate to access venue tips via the API.

Learn to extract venue reviews and check-ins from the Foursquare API using R, focusing on Indian restaurants in Copenhagen, and analyze user tips and comments.

Explore tweet data from Hillary Clinton and Donald Trump using a preprocessed dataset to reveal original versus retweeted content and reply activity.

Explore tidytext basics by converting Jane Austen texts to tidy data, tokenizing into words or sentences with unnest_tokens, detecting chapters via regex, and examining Pride and Prejudice as an example.

Explore and visualize text from Pride and Prejudice using tidytext in R, including tokenization, stop-word removal, word clouds, and sentiment analysis with the bing lexicon.

Explore how word clouds visualize tweet sentiments about India's 2016 demonetization, using pre-processing, corpus construction, and frequency-based sentiment analysis in R.

Create word clouds with quanteda by building a corpus and document frequency matrix, then clean text with stemming and stop-word removal for immigration manifestos and tweets.

Learn to compute word frequency from text data in R, cleaning and converting to a corpus, then analyze Twitter data with API keys and plot the most frequent terms.

Explore text polarity by calculating negative, neutral, and positive content with the cued up library on Mugabe tweets, revealing overall negativity and extracting positive and negative keywords.

Download four public-domain novels, tokenize chapters, and build a four-topic lda model in R to distinguish the books by themes.

Cluster tweets with quanteda by building a document-feature matrix, selecting the top 50 words via idf, and performing hierarchical clustering to reveal word groupings like fake news and media.

Apply supervised classification in R to distinguish ham from spam emails, building a document-term matrix from cleaned text and achieving about 94 percent accuracy.

Use Artex tools to create a document matrix from email text and classify ham versus spam with a linear support vector machine. Split data 75/25 and predict unseen emails.

Explore multiclass classification on text data using R, including encoding removal, tokenization, TF-IDF vectorization, and training a support vector machine with tenfold cross-validation and evaluation via a confusion matrix.

Extract 250 tweets for a hashtag using Social Media Lab and Margaretha, build an actor network and semantic term network, identify communities and clusters, and visualize user-hashtag associations.

Posit lets you deploy and share data science projects from your browser, with no install, using RStudio or Jupyter, and supports R or Python apps like Shiny, Streamlit, and Dash.