With over 326 million monthly active users, Twitter has surely become a part and parcel of our daily lives. From celebrity feuds to trending news, Twitter offers a range of entertainment options to its users. Not only this, even businesses can use the platform to promote events, market, and spread awareness about their brand.
No wonder, the platform is using artificial intelligence to improve the experience of the users who regularly visit Twitter. Let’s see how Twitter is utilizing artificial intelligence and what other businesses can learn from it.
How Twitter Leverages the Power Of AI?
One of the major challenges that Twitter faced is showing relevant tweet suggestions to every user. While a tweet related to fashion may be relevant to one user, it may not mean anything to another. Artificial intelligence is a tool that helped Twitter achieve this. Now, the AI algorithms are used to scan and then, score various tweets to rank them according to every individual’s preferences.
We all know that the UK and Germany government are becoming strict by the day when it is about racist comments, fake news, illegal content, and hate speech on the internet. Hence, Twitter is also reducing racist and other offensive contents on the platform with AI.
It is amazing to note that in just the first 6 months of AI usage, Twitter was successful in eliminating 300,000 Twitter accounts related to terrorists. Most of these terrorist accounts or the accounts promoting such activities were identified with the help of AI algorithms.
Twitter is using artificial intelligence in various other ways to improve the user experience. For instance, the company is now using AI for image cropping to offer users a nearly accurate crop.
Technical Aspects: How Twitter’s Algorithms work?
Twitter collects a huge amount of data and then run it through its deep neural network systems to find out which users want to see what type of content. Every tweet is ranked in order to figure out whether a user would want to see this tweet in their feed. The ranking model considers various things such as:
- The tweet and its content
- Number of likes and retweets
- The connection of the users with the author of the tweet
- Whether the user has had previous interactions with this author
If the relevancy score obtained from all these factors is high, the probability of showing a tweet to the user increases.
There is no stopping the AI wave now. Twitter is constantly improving its algorithms for enhancing its services. They are willing to speed up the process and improve the quality of the algorithms for advanced, real-time updates. Moreover, Twitter is using IBM Watson to improve the content quality on the platform by removing offensive and abusive tweets.
The company is constantly using advanced technology and tools along with knowledge from experts to improve the quality of tweet shown to every user. They are looking forward to enhancing the overall user experience.