A test of the three major voice-activated virtual personal assistants found that when it comes to multi-language use, Siri is more accurate than its rivals. 18 questions in English, Italian, French, German and Mandarin were asked to each of the three personal assistants. Answers that quickly came back correct received a 1.0 score. Those with correct answers that were not quite as easy to find (for example, they might be hidden among search results) were scored a 0.5. Answers that were irrelevant to the question, or which stated that the task could not be completed, were scored a 0.0.
When it came to English, all three scored well and within a few points of each other. Siri (14.5/18) topped Cortana (12.5/18) and Google Now (11.5/18). Siri managed to duplicate her success in Italian with a score of 14.5 while Google Now and Cortana trailed badly with scores of 8.5 and 6, respectively. While most of the other languages saw similar results, Mandarin saw Siri tally an impressive 13.5 while Cortana scored a 5 and Google Now performed miserably with a 1.5 score.
Interestingly, Cortana was the personal assistant who was more likely to respond to an inquiry with a web search. In terms of accuracy across all the languages, Siri scored the highest at 76% topping the 46% scored by Google Now and the 42% achieved by Cortana. Check out the infographic produced about the test, which we cut into easy to read morsels in the slideshow below.
source: VentureBeat via WMPoweruser