Updated July 25th, 2020 at 12:12 IST

Alexa Conversations launched for a more 'natural' conversational experiences

Amazon launched ‘Alexa Conversations’ (beta) that enables one to create skills that customers can interact within a natural, less constrained way. 

Reported by: Bhavya Sukheja
| Image:self
Advertisement

Amazon, at a virtual event for Alexa vendors and developer partners, unveiled tools and resources designed to enable new Alexa voice app experiences. The e-commerce giant rolled out deep neural networks aimed at making Alexa ‘more natural’ and more accurate for custom apps. Amazon launched ‘Alexa Conversations’ (beta), which is a new Al-driven approach to dialogue management that enables one to create skills that customers can interact within a natural, less constrained way. 

According to a press release, the main issue that Alexa Conversations try to solve is the combinatorial explosion of conversation states and code required to manage all possible dialogue paths in a natural conversation. The new Alexa skills will allow users to provide multiple answers in one utterance, to correct or override part of their inputs seamlessly, to handle changes of contexts back and forth, and so on. The e-commerce giant informed that Alexa Conversations are able to extrapolate the full range of dialogue paths and phrasing variations from a few developers-provided samples. 

READ: From Workout To Linguistics: Amazon Reveals How Indians Used Alexa During COVID-19

From happy paths, phrasing variations to uncommon alternatives, Amazon informed that the new Alexa Conversations would generate thousands of dialogues. The large dataset will then be used to train a variety of deep-learning neural networks, including deep transformer-based encoders, recurrent neural networks and attention-based pointer networks. According to the press note, defining a skill with Alexa Conversations requires developers to write dialogues, then organising possible use utterances into sets, and finally define responses. 

Amazon informed, “The trained model can predict the next steps in the dialogue based on the entire conversation’s history, the current state, and the capabilities of the developer’s APIs. It can take action to drive the conversation forward, such as confirming inputs, eliciting missing information, retrieving information through an API call through your skill, or gracefully following the customer’s direction”. 

READ: Google Shopping Announces Commission Free Sales To Take On E-commerce Giant Amazon

Alexa for Apps 

Furthermore, Amazon also announced Alexa for Apps which the e-commerce giant will be bringing to its Alexa ecosystems to enable voice-based interaction with mobile phones. The new feature leverages deep-linking and can be thus used to control any existing mobile app supporting deep-links. The use of Alexa for Apps daily include using voice to launch a search, access additional information, or control an app. 

READ: Rioters Loot Amazon Go Store In Seattle; Trump Jr Calls Them 'undocumented Shoppers'

READ: Amazon Unveils Shopping Cart That Knows What You're Buying

Advertisement

Published July 25th, 2020 at 12:13 IST