Updated: 8/10/2006; 7:02:07 PM.


on personal productivity in knowledge-intensive environments, weblog research, knowledge management, PhD, serendipity and lack of work-life balance...
If you search for mathemagenic that has nothing to do with weblogs try this

Earlier | Home | Later

  Tuesday, July 25, 2006

  Automatic translator for Skype chat and familyskyping project

A few weeks ago we had a very special incident. My mom thought that I was around home computer with Skype logged in as Robert and started a chat. In Russian.

Of course, it wasn't me next to it, but Robert himself. He went to an online translator, figured out what my mom was saying, got his reply translated and posted back. First my mom thought I was making fun of her typing strange Russian, but then she had to believe that she was chatting with Robert. Both of them were very happy and very proud that they found a mode for direct communication despite of speaking different languages.

Thinking of opportunities that this discovery could bring (=not having to translate back and forth between family members :) I thought that it would be great to find an automatic translator for Skype chat. I found one -  ULRTMT - Universal Language Real-Time Message Translator, which is in difficult to install beta. Haven't tried it yet, but already happy since these times things develop fast.

On a side note - it took ages to get my mom on Skype and she is still learning ins and outs, but I'm so much happier now. Between other small victories is having my grandma talking to me (wearing "I look like those guys from Spaceship control center" headset) and giving her a video tour of our house in real time.

Now, there are a few more things before my familyskyping project brings real fruits: get other family members on Skype, make sure my mom finds their webcam and my brother installs it, configure properly for my new Bluetooth headset so I don't have to change the settings manually...

More on: family life Skype 

  OneNote for recording and coding qualitative data (+ more on tagging metaphor for flags)

Last year I experimented with using OneNote to record interviews: it allows to record audio syncronised with notes. The main reason is that I don't have resources (and to be fair don't see the need) to make full-text transcription of all interviews, so I use notes plus audio combination to work on the data. 

Specific advantages of OneNote in this respect:

  • My notes include key phrases, issues and emphasis, so I can easily recover the contents of the interview even without having full-text transcription (I actually transcribed one of the interviews, only to realise that it didn't add any new insights on the data for myself). 
  • I can go directly to an interesting note, click a button and then audio starts automatically - this is a very easy way to retrieve specific fragments in case I want more details.
  • I can edit notes (e.g. in cases when I want to have a full-text quote) and add more notes without destroying the audio link. It's not perfectly synchronised, but works well enough for connecting right paragraphs to corresponding audio pieces.

Specific things you have to keep in mind if you want to use it this way:

  • Make sure you have an external microphone. I didn't have it, so sometimes the audio quality doesn't make me happy. Fortunately, since using OneNote was an experiment I had a back-up audio recorder, so sometimes I have to use that audio and miss the advantages of syncronisation.
  • There is a way for automatic transcription of (close to perfect) audio into notes (another reason to have an external microphone). I haven't tried it out yet, but if it works it would provide interesting opportunities (e.g. good quality automatic transcriptions would allow concentrate on meta-level in note-taking).

Now to the coding. I didn't want to port data from OneNote to any external qualitative data analysis system, since those I checked wouldn't allow to use audio or would break the connection between audio and notes. OneNote has some basic "coding" functionality, but unfortunately this is not enough for what I want.

I thought of coding the data with OneNote flags: flags allow to mark fragments of text and then retrieve it as summaries (makes sense to read this post if you need a better understanding of how it works). I guess OneNote flags were not intended for a use in a way I envisioned for them, so not surprising I have a long list of issues that don't make me happy. Anyway, I'll write them down - who know what inspiration the developers might need for the next version :)

What I miss:

  • Flags
    • User defined icons (or keywords/tags) to remember easily what is behind certain flag (it's a bit difficult to remember what did I mark with blue/yellow/green geometrical figure :)
    • Ability to define keyboard shortcuts for all flags I want (currently only nine flags have predefined shortcuts, to add others I have to go via a couple of menu levels - quite annoying). Alternatively, I could think of a general flag shortcut that would allow me to select a flag from existing ones or add a new one (and I really would love to add text (tags :) as flags).
    • Ability to flag section of text (e.g. one sentence, one word) instead of a whole paragraph as it is now. With pieces of interviews heavily flagged my summaries become too long and unreadable.
  • Flag summaries
    • Ability to see context (e.g. page name) in front of a flag if I go for a summary across different pages
    • Ability to see summaries for specific flags only.
    • Exporting flag summaries (e.g. in a table format with page name : flag : note - I'll find a way to port it into more powerful tools)
    • All kind of advanced stuff (e.g. ability to look at co-occurancies and relationsips between flags, see below for more).

In general, I believe that the whole idea of flags and summaries could grow much further then what is there now. Current functionality supporting flags seems to come from an idea of someone using just a few to mark important fragments or action points - it definitely doesn't scale beyond that. I'd think about flags in OneNote as tags - in a way those are used in del.icio.us, Flickr or blogs - not necessarily social, but definitely a way to organise, retrieve and analyse your own digital bits.

In that respect it would require a flexible flagging (tagging :) system, probably with text-based flags, and a set of tools to retrieve and display slices of data per flag (or combination of flags). Don't know how much it is implementation-wise, but ideas and examples of how it might work are out there in the open (and I'm prepared to work out a couple of scenarios on how this thing might work).

More on: methodology OneNote PhD tools 

Earlier | Home | Later

© Copyright 2002-2006 Lilia Efimova.

This weblog is my learning diary. Sometimes I write about things related to my work, but the views expressed here are personal and do not necessarily reflect the views of my employer.

July 2006
Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          
Jun   Aug

Edublog award 2004 as Best Research Based Blog. Click for more details...

Click to see the XML version of this web page. Click here to send an email to the editor of this weblog. Please, make sure that I recognise your name or you have a nice autorisation message - I tend to decline calls from people I don't know ;)

Locations of visitors to this page Technorati Profile