Microsoft released Cortana, the voice-activated digital assistant built into windows 10 and Windows phones and for all Android phone users in a public beta on April 2, 2014. It was named after Cortana, who is a synthetic intelligence character in the video game franchise of Microsoft’s Halo.
Its development started in 2009 with Zig Serafin as general manager. Cortana is designed to help get things done. It is helpful in getting weather forecasts, setting up reminders, telling jokes, sending emails, finding files, searching the internet and so what not. It has eight languages tailored for 13 countries. It helps us remember things we have said we would do in our emails, without even having to ask her.
Five years ago, Cortana was going to be Microsoft's answer to Siri and Alexa, an AI-powered personal assistant capable of responding to voice commands, armed with a collection of third-party skills and a notebook filled with your personal preferences. Oh, and a fearsome collection of jokes.
Today, if you click the Cortana button on a Windows 10 PC, you still get the "Ask Cortana" prompt, and you can click the microphone button to start talking. But saying "Hey Cortana" does nothing. All those third-party skills are no longer operative, and Cortana's notebook has officially been tossed into the digital dumpster.
Cortana used a Microsoft translator back-end in order to perform searches and answer questions. Microsoft translator speech API (Application Program Interface) is a cloud-based automatic translation service. Microsoft translators’ main and important mission is to break the language barrier by providing translation. It uses Bing as a web search engine which has been developed using ASP.NET. It was Tellme Networks, bought by Microsoft in 2007, from where natural processing capabilities of Cortana are derived & are coupled with a semantic search database called Satori. Its ultimate motive is to get more meaning from data and integrate rational concepts with more powerful abstraction concepts that we know from the field of artificial intelligence. The aim is to provide high level modeling primitives as an integral part of a data model in order to facilitate the representation of real world situations.
Speech recognition in Cortana involves the following:
- Signal processing: It is the technology present in all stages of sound processing, from the conversion from analog to digital, to the application of filters or the gain control. Regarding Cortana it enables extracting the information within signals to translate recognizable words.
- Speech recognition: This is the most important part of this process; here the actual recognition is done. The feature vectors sequence is then decoded into a sequence of words. This is done by using algorithms such as Dynamic Time Warping.Dynamic Time Warping is the algorithm that measures similarity between two temporal sequences which may vary in speed during time series analysis, in Cortana. It is used to cope with different speaking speeds. The program includes a big dictionary of popular words that exist in the language. Each feature vector is matched against the sound.
- Semantic interpretation: Here it checks if the language allows a particular syllable to appear after another. After that there will be a grammar check. It tries to find out whether or not the combination of words makes any sense. It uses “command mode’’ semantic property to decide how to respond to the user.
- Dialogue management: The errors encountered are tried to be corrected. Then the meaning of the combined words is extracted and the required task is performed. It is a framework that provides authentication in Cortana skills. The output of the dialogue management is a list of instructions to other parts of the dialogue system, usually in a semantic interpretation. This semantic interpretation is usually converted to human language by the natural language generation component.
- Response generation: After the task is performed, the response or the result of that task is generated. The response is either in the form of a speech or text. What words to use so as to maximize the user understanding, are decided here. If the response is to be given in the form of speech then text to speech conversion process is used. In many cases Cortana’s voice is not digitized at all. It is really Taylor’s Voice or whoever else owns Cortana’s voice in a given country. In case, there is little no time for an answer to be recorded Microsoft can have her own response be readout by Cortana using her digitally-generated voice font, based on the actor’s or actress’s real voice.
Features of Cortana
- Improved search: the internet may be an ocean of knowledge, but it isn’t the only resource available to Cortana. She can also search the contents of your OneDrive storage or the contents of your computer to find whatever you’re looking for. Cortana also understands natural-language requests, so to find something you’ve worked on recently, you can simply ask Cortana to find files from yesterday, last week, last month, last year or anything in between.
- Cortana Notebook: Like a new friend, Cortana will slowly learn your preferences and habits based on the stuff you tell her, but if there are important things you think she should know right away, you can add them to her notebook. You can include info such as your favorite sports team (Go, Yankees!), stocks you own and recommendations for your favorite cuisines (barbecue or sushi, please), even taking into account how much you're willing to spend. To get to the notebook, just click the Windows 10 search box and hit the book icon on the side of the menu.
- Cortana Home: With a simple click on the Windows 10 search box, you can get Cortana to provide you with a rundown of the important events of the day in one place. She aggregates things like reminders, flight times, reservations, calendar events and more into a single time line, so you know when and where you need to be.
- Voice activation: Another trait of a good friend is being dependable and having your back when you need it. Cortana has that in spades, and to get her attention, all you have to do is say, "Hey, Cortana." It's something so simple that we've started to take it for granted on Android and iOS, and it's about time Windows PCs caught up. Note that this feature may not be activated by default, in order to save battery life, but you can easily turn it on by enabling it under the Settings menu in Cortana's Notebook
- Reminders: One of Cortana's most powerful features is the ability to set reminders that are time-based, people-based and even location-based. That means Cortana can remind you to take out the trash just after you pull into the driveway, walk the dog after dinner at 8 p.m., or ask your friend Dan about his trip the next time he sends you a message or you send him an email.
- Multitasking: This isn't advertised on the box, but Cortana helps you do more in less time by listening to requests without interrupting what you're doing. That's possible because she does not take control of your window or mouse cursor when you ask her a question. This allows you to read and respond to emails while simultaneously asking the assistant to create reminders and add calendars events as they come up.
- Microsoft Edge with Cortana: Not content with just integrating Cortana into Windows 10, Microsoft also gave her a handle on things in its new browser, Edge. If you highlight unfamiliar text with your cursor, Cortana can pull up an explanation on the right side of the window, something I wish I had when I was in school. This also works for things like restaurants, so if you're looking for a place to go for dinner, you can highlight a restaurant's name, and Cortana will pull a list of reviews and links to help you make a reservation. The assistant will also give you helpful info like hours of operation and directions on how to get there.
- Compose an email: While high-powered CEOs still pay for assistants to transcribe dictated emails, in Windows 10, Cortana will do it for free. You have to remember to set up your email accounts first, but after that you can say a name from your address book, narrate your email and send it off without ever touching your keyboard.
Currently Cortana is the name for an ultimately different and new path of handling Windows Phone, and in no time maybe Windows too Everyday Cortana becomes more useful by learning more about us day by day. By paying attention to what you like and how you do things, it certainly provides us with an experience where a person’s individuality is celebrated and not ignored. Since we can decide what Cortana knows about us, we're in control of what information we share. The first attempt from Microsoft to expand its Cortana assistant away from computers and phones and into a more useful home setting is Harman Kardon’s invoke speaker.