Talk to Books - browse passages from books using experimental AI
If you're looking for a new way to find books, try Talk to Books. It's powered by AI and lets you search books using natural language. With Talk to Books, you can use whole sentences instead of just keywords. Google has created this tool with billions of lines of training data, so you can discover books in a whole new way.
In Google AI's Talk to Books, when you type in a question or a statement, the model looks at every sentence in over 100,000 books to find the responses that would most likely come next in a conversation. The response sentence is shown in bold, along with some of the text that appeared next to the sentence for context.
Mastering Talk to Books may take some experimentation. Although it has a search box, its objectives and underlying technology are fundamentally different than those of a more traditional search experience. It's simply a demonstration of research that enables an AI to find statements that look like probable responses to your input rather than a finely polished tool that would take into account the wide range of standard quality signals. You may need to play around with it to get the most out of it.
Not a traditional search
Try our sample queries to get a feel for how Talk to Books works. Then play around with your own ideas. Use it to explore topics you are interested in. Part of the fun is coming up with queries that help you discover interesting perspectives and books you may want to read.
Play with it
Talk to Books is more of a creative tool than a way to find specific answers. In this experiment, we don't take into account whether the book is authoritative or on-topic. The model just looks at how well each sentence pairs up with your query. Sometimes it finds responses that miss the mark or are taken completely out of context.
Use natural language
If Talk to Books isn't finding responses you like, you may get better results by using different words or simply more words. It often does better with full sentences rather than just keywords or short phrases.
Talk to Books is a great way to explore books in a new and different way. With its natural language processing, you can use whole sentences to find interesting books to read. Although it's not a traditional search engine, Talk to Books can be a lot of fun to play with. Give it a try and see what you discover!
Background information on Machine Learning, Deep Learning and Semantic Experiences
Overall, machine learning is a method of teaching computers to learn from data, without being explicitly programmed. In the context of language, this typically means training a computer on a large corpus of text, in order to get it to learn the statistical patterns that occur in language. This approach has been used successfully for tasks such as machine translation, where a system is trained on billions of sentences in order to learn how to translate between languages.
At Google, they're using machine learning to build systems that can understand the meaning of natural language utterances. We hope that these systems will eventually be able to help people in their everyday lives, by providing them with information and assistance when they need it.
To train our machine learning models, Google use a technique called Deep Learning. Deep Learning is a subset of machine learning that uses neural networks - algorithms that are inspired by the structure of the brain - to learn from data.
Deep Learning has been shown to be effective for a variety of tasks, including image recognition, speech recognition, and machine translation. We believe that Deep Learning will also be key to building systems that can understand the meaning of natural language utterances.
In the past, neural networks were difficult to train because they required a lot of data and computing power. However, recent advances in Deep Learning have made it possible to train neural networks that are much larger and more powerful.
At Google, they have been using Deep Learning to build systems that can better understand the meaning of natural language utterances. In particular, we have been focusing on building systems that can answer questions about real-world knowledge.
To do this, we have been using a technique called Semantic Experiences. Semantic Experiences are a type of machine learning where the computer is given a set of experiences (or "training data") that are labelled with semantic information. For example, if we want to build a system that can answer questions about restaurants, we would give the computer a set of experiences that are labelled with information about restaurants.
These experiences could be things like reviews of restaurants, menus from restaurants, or even photos of restaurants. The important thing is that they are labeled with semantic information.
Once the computer has seen enough Semantic Experiences, it can start to answer questions about restaurants. For example, if you ask the computer "Where is the best restaurant in town?", it can use its Semantic Experience to find the restaurant that is most popular among reviewers.
They believe that Semantic Experiences will be key to building systems that can understand and respond to natural language utterances. In particular, they think that they will be important for building systems that can provide people with information and assistance when they need it.
How It Works: Semantic Experiences - Experiments in understanding language
How does a computer understand you when you talk to it using everyday language?
Google's approach was to use billions of lines of dialogue to teach an AI how real human conversations flow.
Once the AI has learned from that data, it is then able to predict how likely one statement would follow another as a response. In these demos, the AI is simply considering what you type to be an opening statement and looking across a pool of many possible responses to find the ones that would most likely follow.
The technique we're using to teach computers language is called machine learning. Google's Machine Learning Glossary defines machine learning as:
"...a program or system that builds (trains) a predictive model from input data."
What does that mean for Google?
Input data: The input data is a billion pairs of statements, where the second statement is a response to the first one.
Predicting: Google is predicting the response to a question or a statement. After seeing all those pairs of sentences and responses, the AI learns to identify what a good response might look like.
Model: The trained system that is used for making predictions. After training, Google's model is able to pick the most likely response from a pool of options.