Facebook said Wednesday that personal data for up to 87 million people--tens of millions more than originally thought--may have been "improperly shared" with Cambridge Analytica, the data analytics firm that worked for Donald Trump's 2016 campaign. Most of those affected were in the United States, the company said. Facebook included the disclosure in the second-to-last paragraph of a company statement that also described new measures to restrict third-party access to user data. Recent stories in the New York Times and the British Observer cited a former Cambridge Analytica employee, Christopher Wylie, who said that the Facebook data of more than 50 million people had been harvested and provided to Cambridge Analytica in 2014. The data was acquired, Wylie said, in the hopes of building personality-based models to target and influence voters in US elections.

I'm using this as way to familiarize myself with bookdown so I don't make as many mistakes with my web scraping field guide book. That book is out of distribution and much of the content is in Matthew's "Mining the Social Web" book. There will be many similarities between his "21 Recipes" book and this book on purpose. I am not claiming originality in this work, just making an R-centric version of the cookbook. As he states in his tome, "this intentionally terse recipe collection provides you with 21 easily adaptable Twitter mining recipes".

This article originally appeared in the Motley Fool. In 2014, Apple (NASDAQ:AAPL) CEO Tim Cook penned a missive regarding Apple's approach to privacy. In the letter Cook famously noted, "A few years ago, users of Internet services began to realize that when an online service is free, you're not the customer, you're the product." Cook was mostly looking to contrast the business model of Apple, which involves mostly making money on device sales, versus that of Alphabet (NASDAQ:GOOG) (NASDAQ:GOOGL), which often sells devices at breakeven to make money from advertising and data collection. However, Cook drew the ire of Facebook's (NASDAQ:FB) CEO Mark Zuckerberg, who responded in an interview with Time magazine.

Code-mixed data is an important challenge of natural language processing because its characteristics completely vary from the traditional structures of standard languages. In this paper, we propose a novel approach called Sentiment Analysis of Code-Mixed Text (SACMT) to classify sentences into their corresponding sentiment - positive, negative or neutral, using contrastive learning. We utilize the shared parameters of siamese networks to map the sentences of code-mixed and standard languages to a common sentiment space. Also, we introduce a basic clustering based preprocessing method to capture variations of code-mixed transliterated words. Our experiments reveal that SACMT outperforms the state-of-the-art approaches in sentiment analysis for code-mixed text by 7.6% in accuracy and 10.1% in F-score.

Legislators, the Federal Trade Commission and other agencies now are considering rules to protect the privacy of users of social networks like Facebook. While those efforts remain in the early stages, even tech companies say privately they expect some regulation to happen down the road. Yet some law-enforcement agencies, including the Federal Bureau of Investigation, and national-security advocates point to a tradeoff, noting that too much privacy can be as bad as too little. Bad actors take advantage of both extremes, abusing access to individuals on networks that are too open or freely conspiring on systems that are too closed. Law-enforcement agencies rely on access to user data as an important tool for tracking criminals or preventing terrorist attacks.

Beyond enhancing robotics and personal devices, emotion AI can be applied in customer experience initiatives, such as VoC programs. A fleet of vendors already offer sentiment analysis by mining billions of data points on social media platforms and user forums.

USA TODAY's Jefferson Graham explains how you can tell Google, Facebook and Amazon to stop following you around the web. LOS ANGELES -- Like many, I downloaded the data this week that Facebook compiled about me over the years, and frankly, it was spooky. I love the social network for showing off my latest photos, staying in contact with old friends and catching up on the latest news. This happened when I joined and Facebook asked if I wanted to connect with other friends by importing my contacts from my computer, which happened to have their phone numbers and sometimes addresses. Facebook has lived with this info since I joined in 2007.

You can request a folder of the things you have uploaded to or shared on Facebook from all your sessions on various devices--plus other curious information--to save on your computer. Once you unzip the folder, open the "index.htm" Right now, you'll see every Facebook friend you've ever had--plus the ones you've gotten rid of. You'll see all your videos, pictures, events, liked pages, private messages and timeline posts. Soon, Facebook will have an improved download tool, and include more information such as posts you've liked, your reactions and comments, even your search and location history. What isn't in this single download is a lot of the behind-the-scenes data that Facebook may use to increase engagement and target ads.

Political data analytics firm Cambridge Analytica still has not deleted the entirety of the Facebook user data it acquired through questionable means in 2015 despite requests that it do so, Channel 4 News reported. According to a new report from the British broadcaster, Cambridge Analytica still maintains a database of personal information and interests pertaining to 136,000 individuals located in the state of Colorado. Cambridge Analytica reportedly has not completely deleted the Facebook data it acquired. It is unclear as to why the political firm is still hanging onto the sizable chunk of data, which purportedly contains a detailed personality and psychological profile for each person. Cambridge Analytica typically uses such information for political advertising campaigns.

Top Stories, Mar 19-25: 5 Things You Need to Know about Sentiment Analysis and Classification; Top 12 Essential Command Line Tools for Data Scientists Top KDnuggets tweets, Mar 14-20: Introduction to Markov Chains "What are Markov chains, when to use them, and how they work"