Pages

Sunday, May 8, 2016

Can Voice Recognition Software Help You With Your Genealogy?

Voice recognition software has come a long way, but has it come along far enough to be of benefit to genealogists doing research and all the other things that genealogists do? Over the years, I have written about my experiences with VR (voice recognition) software. To some extent, it has progressed from being a cranky toy into being a useful tool, but it is still a tool with some significant limitations.

Many people today, are familiar with VR in the form of Siri on Apple iPhones and iPads. Other similar products are available on other mobile devices. I absolutely have no use for Siri or any similar program. I don't need to try and figure out how to ask a question Siri can answer correctly. I can almost always enter the searches I need to do faster than figuring out Siri's responses. If you don't know what I writing about, then you are probably a long way from utilizing VR.

Essentially the idea of VR software is that you speak into a microphone and your words are transformed into text on the computer screen. Nice theory. More difficult in practice than in theory. The reality is that if I speak at a very constant speed and say all of my words very clearly, the latest versions of the software work pretty well. Anything I speak-write however must be very carefully reviewed and edited and any particular time savings are usually lost in editing time. For some time now, I have been using a program on the iMac called Dragon Dictate. It is the Apple OS X version of Dragon Naturally Speaking both from Nuance.com. The programs are generally available online from Amazon.com and other online retailers or directly from Nuance.com.

It is not my goal in writing this to evaluate any one software product, I am primarily asking whether or not this type of software will help the average genealogist now someone like me who writes constantly. Let's looks at some of the writing activities you might use as a genealogist and see if any of them match the advantages of VR. The is, of course, one strongly motivating factor for using VR; if you are disabled or cannot type.

Entering information into genealogy programs

Some VR programs do a fairly good job of recognizing names. But the command set for entering information into fields in forms is complicated and probably will turn out to more frustrating than just typing. In this case, it may be possible to use a mixture of VR and typing by hand, but I haven't found that to be the case from my standpoint.

Entering long blocks of text into programs

The more text involved, the better chance there is that the VR software will help in data entry. I still find that the programs substitute words and misunderstand the context of what I am trying to say. Sometimes, when I am pressed for time or have a passage of text that cannot be easily copied into a word processing program, I will plug in my microphone and dictate the text. I'm not always convinced that dictating the text using VR software is the answer. Again, sometimes it takes me more time to correct my entries than it would have taken to simply type them in the first place.

[See if you can tell when I started using VR software to write this post. I switched in the middle of post.]

General writing such as letters, blog posts, email, etc.

The matter how sophisticated the program seem to get, they do not completely respond correctly to dictation. In many cases, I have to slow down considerably and speak very clearly. Even though I do this I am forced to watch what is translated very carefully to avoid having words substituted and the text garbled.

I would have to take the position that voice recognition software is a good solution and some limited cases where accuracy, consistency, and mixtures of numbers and dates and places is involved, it is it is just as fast to enter the data by hand.

If you're at all interested however, I would strongly suggest you try it out and see if it works for you. Come prepared to spend a fair bit of time learning the programs and teaching them to respond correctly. They also need to teach you and that takes some time.


1 comment:

  1. I've been using Dragon at work for about eight years. As my employer won't update the software, I have an older version that requires a lot of proofreading. That said, life without it would be miserably painful, so I love my Dragon. The quality of the microphone used is key. The one that came with the program was junk. I requested, and received, a better quality microphone headset (gaming quality)., which helped a lot. That said, oddly enough, it works better at understanding me if I'm chewing gum while I dictate, which seems counterintuitive to me.

    I use a newer version at home to write up my research notes, draft blog posts, and send email. It works drastically better than my work version. I love it. It gets virtually nothing wrong, and only requires such proofreading as I would do for any regularly typed document (but I already knew how to use it so there was no learning curve on my part -- only it learning my speech patterns). I do not speak slowly. I can't keep that up. It just had to learn me (that took only days with the newer version at home and weeks of constant use with the older version at work). But using it to fill out forms and such (genealogy programs) doesn't work very well as there isn't enough context for it to figure out which words I'm using, which has led to some hilariously unhelpful results.)

    ReplyDelete