AWS Transcribe

QName: f:aws-transcribe

Empowers a node to automatically compute textual transcripts for audio content that it contains (as a binary attachment) or points to (as a property). These transcripts are computed every time the node or its binary attachment are modified.

The transcript text can then be applied back to the node, storing it onto a target property.

Once the transcript text is written back to the node, the node's index will be updated within Elastic Search, allowing for immediately full-text search of audio files and audio content.

This feature connects to the Amazon Transcribe service to perform transcription. To use this feature, you will need to have an AWS account along with IAM user credentials that provide the following access rights:

  • full rights to Amazon Transcribe
  • full rights to a single Amazon S3 bucket

For steps on how to configure the Amazon Transcribe connectivity from within Cloud CMS, see Create a Service Descriptor below.

Configuration

Property Type Required Default Read-Only Description
serviceDescriptorId text false The ID or Key of the service descriptor to use.
property text false transcript The path to the property that should store The ID or Key of the service descriptor to use.

Create a Service Descriptor

To use this feature, you will first need to declare a Service within Cloud CMS. Go to the Manage Platform > Services page and then click Create a New Service.

Given the service any Title and unique Key that you'd like.

The type should be aws_transcribe and the configuration should look like this:

{
    "accessKey": "",
    "secretKey": "",
    "region": "",
    "bucketName": ""
}

Save your service. That's all there is to it.

Apply the Feature to your Definition

As with all features, you can apply them at either the instance or the type level.

Suppose you have a content type called Episode which is meant to store audio files for episodes of a radio show or podcast. It might look like this:

{
    "title": "Episode",
    "_qname": "my:episode",
    "properties": {
        "title": {
            "title": "Title",
            "type": "string"
        }
    }
}

Suppose the MP3 audio is stored on the default attachment. We can add a mandatory feature to have transcription automatically occur and have the results placed onto our node (once they are available).

We might add a mandatory feature like this:

{
    "title": "Episode",
    "_qname": "my:episode",
    "properties": {
        "title": {
            "title": "Title",
            "type": "string"
        }
    },
    "mandatoryFeatures": {
        "aws:transcribe": {
            "serviceDescriptorId": "myAmazonTranscribeServiceKey",
            "property": "transcript"
        }
    }
}

Where myAmazonTranscribeServiceKey identifies the Service Descriptor you defined in the previous section and transcript is the name of the property on the JSON that will receive the transcript text.

With that in place, we can now create a new Episode (my:episode) and upload an MP3 attachment to it. Once the MP3 is uploaded, Cloud CMS will call out to the AWS Transcribe Service to transcribe the text of the audio file.

For example, if we uploaded freshair-terry-gross.mp3 and wait a few minutes, we will eventually see our content receive the following text:

This is fresh air, novelist helen schulman says in the author's note at the end of her new novel called come with me that her fears about an obsession with the internet have no bounds. Our book critic, maureen corrigan, says shulman finds an excellent way to deal with those fears by channelling them in tow art. Here's maureen's review the great internet novel like the great white whale, it's rumored to be out there somewhere beyond the horizon. So far, the novelist, who have been hailed as coming closest to writing it, have done so in dystopian door stoppers even longer than melville's moby dick. I'm thinking of the circle, by dave eggers and book of numbers by joshua cohen, both of which tells sweeping cautionary tales about the wired life within facebook type cult compounds. But helen shulman is taking a different tack in capturing the internet revolution. She zooming in tight and close on all those computers and smartphones scattered around the rooms where we live. Shulman has a gift for vividly tracing the fallout on the domestic realm from that hard to pinpoint historic moment. When the internet first drifted through the walls of the american home. In her best selling and harrowing twenty eleven novel, this beautiful life, shulman described how a sexting scandal broke apart and otherwise solid family and school community. Now, in her new and much more narratively ambitious novel called come with me, shulman spices together an old school family drama with high tech fantasy. It's a rich, closely observed story about regrets and stupid risk taking, set mostly among the coders, crunchers and ordinary citizens of palo alto, california the heroine of schulman story is amy reid. She's a middle aged mom who finds herself in the humbling position of working as a pr flunky for her old college roommates. Nineteen year old son dani is a brilliant stanford undergrad who's founded his own startup. He's just invented an algorithm that allows users to access all the alternative lives they might have lived. Or, as he calls them, multi verses. Here's how it works. Dani knows a lot about amy's early life from his gossipy mother and mohr, disturbingly from the cloud where apparently the most personal information floats ripe for the picking. Equipped with this data, donnie creates what he calls ah personalized crystal ball for amy. He instructs her to strap on a virtual reality headset and journey into her various what if lives. What if she'd married her handsome book feckless actor boyfriend ? What if she'd tripped that time ? She was running after her then toddler son and didn't catch up to him before he ran into the busy street. What if she carried that unplanned early pregnancy to term ? Even as amy is being traumatized at work by these deep dives into her alternative lives, the internet is changing the texture of her family's daily routine. At home, amy's teenage son, jack, maintains an intimate relationship with his out of town girlfriend, mostly via webcam, while her husband, dan, a former journalist who got phased out of his job at a dead tree publication, spends his days angrily tweeting, we're told. It's no wonder that dan, bitter and scared, clung to his iphone the way some cling to their religion and their guns. It was like an oxygen tank for his breath starved mind. Come with me is ingeniously structured around three non consecutive days in the lives both virtual and riel of amy and her family. There are a lot of story lines here as well as a lot of humor and heart. Amy is the fully realized moral center of the novel, intrigued by innovation but also graced with a witty feminist skepticism about the male dominated tech industry. For instance, reading an article that theorize is that tech may one day cheat death by making it possible to download the human brain into a robot. We're told. Amy reflects that when she was in college and studied literature, it was male novelists who'd believe they could use their craft to achieve immortality. Updike, mailer, bello. They behaved as if the written word could defy death, but clearly it had not. Her teenage son, jack, had never heard of any of them. Fair enough. But i'm still putting my money on literature to transcend time. In fact, if it this moment, i had to recommend one contemporary novel to the readers of the far future to give them a sense of how we newbies are struggling to grasp the unforeseen powers of the world wide web. I'd give them shulman's come with me. It poignantly captures the wonder as well as the cluelessness of how we live now. Maureen corrigan teaches literature, georgetown university she reviewed come with me by helen shulman if you're looking for a new book for yourself or to give us a gift for this holiday season, npr has compiled more than three hundred titles recommended by npr staff and critics, including maureen. You confined the twenty eighteen book concierge at npr dot org's slash best books and check out our podcast if you want to catch up on interviews you missed, like our interview with trauma surgeon joseph saccharine, who's been working for gun safety from the perspective of his experiences treating gunshot victims and as a gun shooting victim himself, he was shot in the neck when he was seventeen. He responded to an nra tweet with a tweet that helped launch a campaign mobilizing doctors and nurses around the issue of gun safety. Fresh air's executive producer is danny miller. Our interviews and reviews are produced and edited by amy salad, phyllis myers, sam bricker, lauren krenzel, heidi saman, theresa madden moods. Eighty there. Challenger and seth kelly. I'm terry gross. 

And this will be put onto the transcript property for your content item:

{
    "title": "Fresh Air with Terry Gross",
    "transcript": "This is fresh air, novelist helen schulman says in the author's note at the end of her new novel called come with me that..."
}

This will then be indexed for full-text search. A search for helen schulman, for example, will discover this Episode.