Preface: The internet and your privacy.
Before we dive into DeepText, I think it is important for internet users to understand that in the connected world there are businesses that are built on collecting and analyzing various user data. This can include user data about search history, social engagement, websites visited, location data and those are only a few examples.
These businesses recognize that this data is extremely valuable. In fact, two of the largest companies in the world, Facebook and Google, generate most of their revenue from selling advertisement on their platforms. These marketing channels are extremely valuable because they allow marketers to target users that are most likely interested in the products or services offered. These targeting capabilities are only possible because user data is collected and analyzed to understand and segment users.
If you are the type of person who is uncomfortable with businesses collecting data about you, you should stop using the internet and social media. The fact of the matter is that this data is collected and it’s not going to stop anytime soon. The discussion on whether data collection is ethical is the topic for a different blog post. For now, just realize that various user data is collected across the entire internet.
What is DeepText?
Now that we understand why user data is collected, let’s discuss Facebook’s DeepText. If it’s the first time you’ve heard about it, I’ll simply say that space age technology is upon us.
DeepText is a deep-learning based text interpretation engine that has the ability to interpret more than 20 languages with near-human accuracy. Deep-learning means the system can teach itself by interpreting and analyzing text data by using multiple processing layers with complex structures. DeepText will allow Facebook to interpret thousands of posts per second and understand language, including slang, at a very high-level. To learn more about the technology, check out Facebook’s post on how DeepText works.
An investment in user experience.
Now it’s time to ask ourselves why text interpretation is important to Facebook. DeepText is part of their initiative to better understand its user’s interests and behaviors. By analyzing this data, Facebook will have a wealth of information that will be used to create a better experience on its platform. They will also be able to develop better targeting features that will serve advertising to users more effectively.
Here are a few examples of how Facebook plans to use this information to create a better experience for users:
- Populate news feeds with posts about topics that user interacts with most.
- In messenger, Facebook will be able to provide users with advertising that addresses a need discussed in the conversation.
- Filter out spam content to provide users with higher quality content.
- Classify user’s interest and behaviors to deliver more relevant advertising.
- Find the most relevant comments on posts to improve comment quality.
As we’ve mentioned in other blog posts, a better user experience on Facebook means more engagement and more engagement leads to more advertising revenue.
How will DeepText affect business users and advertisers?
It’s safe to assume that advertisers will continue to see new avenues for advertising rolled out each year. These new advertising features will most likely help marketers catch users at various stages in the sales funnel. It wouldn’t be surprising to see Facebook take a deeper look at need indication and develop new ways for advertisers to serve ads to users the moment that user indicates a need for their products and services.
By collecting and analyzing text information, Facebook can also use that data to more effectively segment users and categorize interests and behaviors. Ad targeting features will most likely improve and the rate of engagement/conversion for ads will probably increase as well.
It will be interesting to see how Facebook uses this new data to further shape their platform and provide their users with a better experience.