Bots   9536

« earlier    

Research: "Don't @ Me: Hunting Twitter Bots at Scale"
New research from Duo Labs focusing on identifying Twitter bots at scale discusses ways of effectively identifying bots on Twitter, with Duo researchers applying their approach in a "case study" analysis focusing on a botnet spreading a cryptocurrency scam. Duo first identified a dataset of 88 million public Twitter accounts for the study, and then utilized "practical data science techniques" to find the bot networks.  Duo says that "by monitoring the botnet over time, we discover ways the bots evolve to evade detection," adding that "after finding initial bots using the tools and techniques described in this paper, a thread can be followed that can result in the discovery and unraveling of an entire botnet." Duo notes that their research is focused purely on identifying automated Twitter accounts, as opposed to automated accounts that are "necessarily malicious."

Duo made the data collection code they used open source; it's available on GitHub.

Read the full technical paper (pdf) here.
research  twitter  social  bot  botnet  bots  duo 
yesterday by dmcdev
Threads raises $20M for its luxury goods ’boutique’ that exists only in messaging apps | TechCrunch
When you think of e-commerce marketplaces, chances are that the first things that come to mind are storefronts built on websites and apps. But today an e-commerce startup that has never had either — and never plans to — has raised a fistful of cash to continue building out its shopping experience on the platform has been its growth engine: messaging apps.

London-based Threads has raised $20 million in funding for an operation that courts high-end, millennial, mostly female customers with tailored selections of luxury fashion, which it then sells to them on services like WeChat, WhatsApp, Snapchat, Instagram and Apple’s iMessage for their primary interactions with a team of human (not AI) shopping assistants.
clothing  new-companies  bots  ecommerce 
yesterday by dancall
New Google AI Assistant Helps Users Stay Fit |
According to Venture Beat, there are rumors the company is set to launch Google Coach, which would deliver health and fitness insights proactively, using calendar appointments, reminders, and logged activities. For example, if you skip a scheduled gym day, it might remind you to find another time. And if you’re not meeting a fitness goal, it could offer up workouts and routines that would help you get closer to achieving it.

Google Coach will also keep track of your eating habits, as well as other signals (like location) to make recommendations when you’re dining out or shopping at the grocery store. Google Coach’s suggestions will come in the form of notifications and messages.
google  health  ai  bots 
2 days ago by dancall
Report: Humanizing Digital Interactions - Customer Experience Matters®
" The Human Conversational Model, which is made up of seven elements – Intent Decoding, Contextual Framing, Empathetic Agility, Supportive Feedback, Basic Manners, Self-Awareness, and Emotional Reflection"
ai  conversational_ai  bots 
2 days ago by luceh
Bots didn’t flop; they just became invisible - The Verge
In other words, the bots never really went away; they just became invisible. More automated messaging can be found on companies’ websites and apps than ever before. The work continues. And as Intercom’s own story has shown, businesses’ appetites for the automation they enable is only increasing. (Intercom released a tool to let businesses build custom chat bots earlier this month.)
ee  bots 
3 days ago by osi_info_program
The Flourishing Business of Fake YouTube Views - The New York Times
Piles of interesting here. For a start, the latest Horowitz brat, Russia Today, and Fox all use the same botherd. A classic example of D^2's principle that if you respond to attempted fraud just as an administrative issue, fraudsters will experiment and learn how to steal from you reliably. Finally, algo kitsch.
advertising  fraud  clickfraud  bots  youtube  security  to_blog 
3 days ago by yorksranter
Tips For Requesters On Mechanical Turk: The BOT problem on Mturk
In an academic setting, 95% is an A. In a professional setting, 95% is abysmal. Many studies are published on Mturk using 80-95% and some using no qualifications at all. Your study WILL fill using 99% and above if you pay a fair wage.
research  methods  academia  mechanical-turk  amazon  mturk  bots  automation 
7 days ago by jomc

« earlier    

related tags

%on_github  %product_hunt  2018  a  a11y  aaa  academia  accessibility  advertising  affect  after  aggregation  ai  ai_-_assistants_...  alexa  alexa_skills  algorithms  alt  alttext  amazon  amazon’s  and  anti-spam  api  are  art  assistants  automation  autonomous  aws  backlash  bad.ux  bookmarked_on_site  bot  botkit  botnet  bots  business  change  chat  chatbot  chatbots  civility  clickfraud  clothing  code  collaboration  content  conversational_ai  copywriting  cx  dark.patterns  dark.ux  dataviz  delay  design  development  digitalhealth  disinformation  drawing  drupal  duo  ecommerce  ee  facebook  facebooks  fail  fakenews  faq  faqbot  finally  form  framework  fraud  funny  future  gamification  gcp  glitch  google  guardian  hackers  health  home  hometech  honeypot  hospital  how-to  howto  hsklikes  humans  im  imagery  images  instagram  instructions  interactivepublishing  it  javascript  job  journalism  kim  long  machines  madeline_lamo  marketing  mechanical-turk  mediainnovation  medienatlas  messaging  messengers  methods  microsoft’s  misinfocon  moderation  mturk  navigation  netpolicynotes  network  new-companies  news  newsroom  nlp  node  opensource  openstates  overview  ovum  package  payment  policy  politics  pp  predictions  press  prevention  privacy  programming  projects  propaganda  prototypes  publicbydefault  publicity  purge  python  quartz  r-project  reference  research  robotics  robots  rss  russia  ryan_calo  saas  sales  sample-data  schedule  schweiz  security  social+networks  social-media  social  social_media  socialmedia  spam  startups  talking  tax  technology  telegram  text  to_blog  todo  tools  toread  trends  trolls  tutorial  twitter  venmo  verification  voice  washingtonpost  youtube 

Copy this bookmark: