By now, everyone is familiar with Amazon Alexa and other voice command-based technologies. These technologies typically require a user to address the artificial listener, in Amazon’s case “Alexa,” prior to giving a command. For example, “Alexa add milk to my grocery list.”
A recently published patent application filed by Amazon Technologies, Inc. shows that initially addressing the device may soon be unnecessary as Amazon’s device would, instead, be listening continuously. The application states this is being done for “targeted advertisement and product recommendations.” The device would use a “sniffer algorithm” that would identify trigger words, typically verbs, that indicate “some level of desire or interest in a noun that follows the trigger word,” for example, “I love hamburgers.”
The trigger words are then analyzed and compared with adjacent audio to determine keywords associated with the trigger word. The keywords are stored and advertisers will use those keywords to customize content relevant to the user. This would also allow Amazon to extend offers on certain products or encourage a user to take action on purchasing a product.
The algorithm would also distinguish between positive and negative keywords. Positive keywords, for example, “prefer” “love” “enjoy”, are words associated with the user in a positive manner. An emphasis is then put on the noun that follows. If the user spoke about a specific pair of shoes they loved, the word “love” would hold a higher value over another verb such as “like” and the user may begin to receive advertisements for the brand of shoe they stated they loved. If a certain keyword is used multiple times in a conversation it may also be assigned a higher priority than other tagged keywords.
Negative keywords such as “hate” “dislike” “returned” are still associated with the user except now they are an indicator of an item that is disliked or unfavorable and it would trigger the algorithm to not send the user advertisements associated with the negative word(s).
These keywords are temporarily stored in a database and may be timestamped. Identifiers are also stored to associate the keywords with the correct user.
Some may feel this technology is too invasive or too “Big Brother” for them. According to the patent application, the user has the “option of activating or deactivating the sniffing or voice capture processes, for purposes such as privacy and data security.” There must also be a “willingness [from the user] to have voice content analyzed” in order for the trigger-word algorithms to function properly.
A spokesman for Amazon stated “[w]e take privacy seriously and have built multiple layers of privacy into our devices. We do not use customers’ voice recordings for targeted advertising. Like many companies, we file a number of forward-looking patent applications that explore the full possibilities of new technology. Patents take multiple years to receive and do not necessarily reflect current developments to products and services.”
Suiter Swantz IP is a full-service intellectual property law firm, based in Omaha, NE, serving all of Nebraska, Iowa, and South Dakota. If you have any intellectual property questions or need assistance with your patent, trademark, or copyright matters and would like to speak with one of our patent attorneys please contact us.