Apple could smarten up Siri using crowdsourcing


Apple software users who’re frustrated when Sire fails to reply to a query could someday be ready to tap into the gang to get a greater response. The Apple patent software “Crowdsourcing data to satisfy consumer requests,” published Thursday through the United States Patent and Trademark Place of Job, envisions a few easy methods to assist Sire when it comes up empty.

Sire usually relies on websites such as Bing, Yelp, and Wolfram Alpha to collect the information to answer a question. However, from time to time, these sources are not geared up to offer an accurate response, particularly if the question itself is phrased in a method that appears to confuse Sire.

Instead, Apple’s invention would use other sources reminiscent of skilled data services and products, basic knowledge sources, and online Q&A boards to offer information. Questions and answers are also stored in a crowdsourced database that Sire would get entry to. Such a database can be searched using pure language queries, providing a greater shot at answering questions that may stump Sire in any other case.


Read More Article :

Or that you may select a specific source to test when asking the query. These other sources could come into play robotically if Sire cannot respond to a question. You also have to charge the response to your query, which might help Sire study from its hits and misses.

However, the more difficult the query, the longer you wait for a response. The patent software suggests a range of time frames, from a couple of minutes to several weeks, depending on the nature of the request and the helpful information required from the exterior source.

As Apple describes it in the average patentees:

A user request is received from a cellular client instrument. The user request comprises a speech enter and seeks an informational resolution or efficiency of a task. A failure to supply a satisfactory response to the consumer request is detected. In response to detecting the failure, data related to the user request is crowdsourced by querying several crowdsourcing information sources. Several answers are obtained from the gang sourcing data sources. The response to the consumer request is generated in keeping with at least probably one or more solutions acquired from one or more crowdsourcing information sources. A patent utility doesn’t imply that the expertise will reach the shopper market. However, crowdsourced knowledge seemis a logical step in Sari’s ongoing evolution.