For one of my clients, I'm currently implementing the Salesforce API. Salesforce is a industry-leading CRM (Customer Relationship Management) tool, basically a massively powerful addressbook and scheduler. While reading documentation to get their SOAP client working, I came across this tutorial page:
Developing on Salesforce Using PHP(requires an account on that site).
While the actual content was a little remedial for my purposes, I was struck that this is a great model to emulate. They are generating pretty high quality video documentation from live events, and apparently taking the time to break the video up into sections, and tie it together with a table of contents, all managed through their flash video widget.
Another really interesting example is what the NYTimes is doing with their coverage of the US Presidential debates:
This interlinking of both section topics and specific trascript points is amazing. I'm sure it was a lot of work to set up, but it's without a doubt a great way to present the information.
When I think of Drupal Dojo 2.0, the first thing I think of is better organization, branding, etc, but beyond that these are the kinds of technological improvement that I think are really exciting, and would be great to pioneer within the open source world.
Comments
Salesforce API, eh?
Interesting that you should mention this today. I'm going to be launching into a Salesforce project tomorrow with a client, and as a matter of fact, am about to send an email to VictorKane about his custom module and see when it might be ready for testing. I do love the NY Times' use of video + transcription + Navigation Sections. That would SO work in the dojo 2.1!
Senpai
Joel Farris | my 'certified to rock' score
Transparatech
http://transparatech.com
619.717.2805
Dude
Dojo 2.1 is so tomorrow!
Here's a recent roadmap for salesforce.module
http://drupal.org/node/183684
Senpai
Joel Farris | my 'certified to rock' score
Transparatech
http://transparatech.com
619.717.2805
Transcription Services Could Help / OS Speech-to-Text Software?
I've done some work with aligning text transcripts with video, and it is a lot of work, but there are paths towards better automation.
These days, there are a lot more places who can be paid a nominal fee to get a transcript of an audio file.
This could also be done by volunteers, who would understand and be able to correctly capture all of the jargon. For example, I know that the DemocracyNow.org transcriptions are down by a small army of volunteers who split up the segments and rotate by day or week to get the full one-hour daily show transcribed and posted online. Also, videoblogger Ze Frank's energetic fans took it upon themselves to transcribe his 3-4 minute daily episodes. Here's an example transcript from this episode that he did on "How to choose a web developer" (warning: some graphic language).
But for timeliness and consistency sake, then some profession transcriptionist services are out there who can at least get the initial draft, which could then be thrown into a wiki to be cleaned up by those who listen to it afterwards.
I don't have any specific recommendations, but I do remember reading this transcription service review from almost 2 years ago. There are probably more services that are cheaper today.
Finally, the technology is probably getting even closer to being able to have speech-to-text desktop applications that could do a lot of the heavy conversion. I am aware of web services like EveryZing.com that has the capability to take a video/audio RSS feed and do a machine speech-to-text transcription hyperlinked to the word level. Here's an example. But I don't think it's easy to copy and paste their output.
But I'd be interested in hearing if anyone else knows of some open source speech-to-text transcription software that's out there. The capability certainly exists.
UPDATE: Just noticed that EveryZing advertises that they have a few patents for their speech-to-text service listed with more pending. For those interested in the technical specs check 5241619 & 5621859
So I guess the good news is that whenever these patents expire, then it could provide a good road map for an open source solution. :)
transcription & translation via dotSUB.com
Anyone interested in transcription (and translation!) of video should check out dotSUB.com. There is an episode of Rocketboom explaining how the service works. In a nutshell, dotSUB allows users to quickly and easily enter captions which are overlaid onto a Flash video file. The captions also allow other users to translate the content into other languages. Once subtitles are entered, the new language is immediately available for viewing, since the original video is shown with the new language overlaid on top.
At the very least, I think it would be worth investing time in a project to transcribe some of the general marketing-focused features screencasts and especially some of the internationalization and translation screencasts.
The site is in beta and it's been a while since I've checked it out in much detail, but as far as I know the service is currently free.