Archive for April, 2012

Transcription System: QC & Delivery

Friday, April 20th, 2012

This is a series of posts on our human-powered audio transcription system. The following are links to the previous parts: OverviewWorkflowCertificationTranscription & Reviews, Proofreading.

Delivery is the final step of our workflow. Here we perform quality checks on the final transcript that was prepared in the previous step to determine if the transcript is deliverable. When we are satisfied with it we convert the file into MS Word, Adobe PDF, OpenOffiice.org Text file using a template file. Then a notification is sent out to the customer that the transcripts are available for download.

We do two types of quality checks; random sampling and keyword analysis. The keyword analysis helps us spot the least relevant terms in the transcript. The audio around those terms are checked again to ensure that they’re correct. We have developed our own tools which help us do these checks quickly. Typically we don’t have to spend more than the duration of the audio file to complete these checks and deliver the file.

If the quality checks fail — which does happen once in a while — the transcript is proofread once more. We also do a root cause analysis after that to prevent these cases from happening again. That helps us improve our system. Mostly it happens because of poor audio, or difficult accent and/or diction of the speaker.

This post completes this series about our Transcription System. We hope we have given you valuable insight into how we work. We believe we have superior transcription process which performs better and is more scalable than other systems out there. But don’t take our word for it, try out our transcription service today and see for yourself.

Transcript Progress

Thursday, April 19th, 2012

One of the unique features of our Transcription System is that you can monitor the progress and check the Work-In-Progress transcript anytime from your account. The progress is the percentage completion based on the stage of the workflow the file is in at any given point of time. It maps in the following way.

Range Stage
0-59% Transcription
60-89% Review
90-97% Proofreading
98% Quality Checks

The mapping is approximate. Sometimes one part file has not been reviewed but some other part files have. In that case the completion will be 60% even though the transcription is not yet complete.

The WIP transcript may have higher number of mistakes and inconsistencies till the Proofreading stage. The ‘transcript’ link just collates the part files together and displays the result. The Proofreading stage is where a lot of inconsistencies and mistakes are corrected.

The WIP transcript and the progress tracking feature is one of the differentiating features of Scribie.com. We put our customer first and strive to provide an easy-to-use and transparent system which allows you outsource your audio transcription work without any hassles.

New Transcript Order Options

Thursday, April 19th, 2012

We have added new transcript order options in the Invoices page. The Option button just below the table brings it up. A screenshot is below.

 

A description of each of the option follows:

Flexible Delivery:

The Flexible Delivery files are delivered anytime with a week instead of the normal 1 day. If within the week we are able to fit in the file in our delivery schedule, we will deliver it. Otherwise you are guaranteed delivery on the seventh day. We offer a 25% discount for these orders.

This option has to be specified before the order is placed. It cannot be enabled/disabled afterwards.

Additional Proofreading:

Proofreading is part of our workflow where we correct all inconsistencies in the combined transcript. It drastically improves the quality of the transcript. For difficult files (eg. files with poor audio quality, background noise, speakers with unclear accents, non-native speakers, etc.) you can request an additional proofreading. We pay special attention to these files and ensure that you get an accurate transcript regardless of the difficulty level of these files. We charge an additional $0.50 per minute of the audio for additional proofreading.

This option has to be specified before the order is placed. It cannot be enabled/disabled afterwards.

Strict Verbatim:

Set this option if you want us to include all ah’s, uhm’s, umm’s as spoken in the audio file. By default our transcripts are non-strict verbatim transcripts which means that these filler words are not transcribed unless necessary (ie. for a affirmative or negatory answer to a question). However these filler terms can indicate hesitation or confusion on the interviewee’s part and including them may prove useful. If this option is enabled then all the filler terms will be included in the transcript.

This option has to be specified before the order is placed. It cannot be enabled/disabled afterwards. No additional charges apply for this option.

Time-coding:

By default we time-code the transcripts. The time-code is the time stamp of the audio when the particular paragraph started. It is useful when you want to cross check the transcript agains the audio file. Time-codes are formatted as H:MM:SS.

This option can be enabled/disabled anytime from the Invoice Details page. No additional charges apply for this option.

Speaker Tracking:

Before each paragraph we prepend the speaker name or initial which identifies the person who was speaking at that time in the audio file. If speaker names are provided or spoken in the audio, then they are used. If not we use the generic format of Speaker 1, Speaker 2 and so on.

This option has two sub-options relating to formatting of the names. If speaker initials is chosen then the first occurrence will be the full name of the speaker and rest will be initials. Eg. if the name is Rajiv Poddar then the first time Rajiv speaks the speaker tracking will contain Rajiv Poddar and all subsequent one’s will be RP. This is illustrated in the sample transcript. If full names option is specified then the full name will be used throughout.

The Speaker Names link allows you to specify each speaker’s first and last name.

This option can be enabled/disabled anytime from the Invoice Details page. No additional charges apply for this option.

Transcript Template:

The transcripts are prepared as a text file and before the delivery we convert the file to MS Word, OpenOffice Text and Adobe PDF format. We use a template for this conversion. There are four different templates you can specify; Scribie default template with single/double line spacing and blank template with single/double line spacing. The Scribie default template has a title page, a header, footer and the last page contains some stats. Pleas check the sample transcript for an example. Blank template  only contains the transcript. Double and single line spacing adjusts the spacing in between the lines in the delivered file. If you want your own template to be used instead then please contact support and they will help you out with it.

This option can be enabled/disabled anytime from the Invoice Details page. No additional charges apply for this option.

Spelling Style:

The spelling style of the transcript can also be specified. Currently we support American, British, Australian and Canadian styles of spelling. Please contact us for more options.

This option can be enabled/disabled anytime from the Invoice Details page. No additional charges apply for this option.

Transcription System: Proofreading

Thursday, April 19th, 2012

This is a series of posts on our human-powered audio transcription system. The following are links to the previous parts: OverviewWorkflowCertification, Transcription & Reviews.

In our workflow, we first break the audio file into smaller parts and each part is then transcribed and reviewed by different people. Due to this methodology the transcript may contain inconsistencies. Proofreading is the step where these inconsistencies are corrected. Our proofreaders are the best of the best amongst our certified transcribers. We employ them as contractors and train them. The proofreader goes through all the parts of a file, does all the corrections necessary and prepares the final transcript for delivery.

We have specialist proofreaders for different accents and subject matter. Tough accents such as Indian, African, etc. and mixed accent files are handled by them. Subject matter experts take care of Medical, Legal, Academic and technology (eg, Web Development, Telecom etc.) files. Additionally our proofreaders are trained to research specific terms and acronyms. We also do additional proofreading if requested. Additional proofreading option can be specified while ordering and is recommended for difficult files.

Our delivery capacity is limited by number of hours we can proofread in a day. That is the reason behind our delivery limit of two hours per user per day. We stagger the deliveries over several days when more than two hours is ordered by one customer in a day. It is a safeguard to prevent us being overwhelmed by a single large order. However we do have the ability to ramp up very quickly and recruit new proofreaders whenever there is a need. We just have to be told about it in advance.

Proofreading is what sets us apart from other transcription services. It is designed to guarantee high transcript quality. If you are looking to get a highly accurate transcripts at a reasonable cost then our transcription service is the right choice for you.

The next part of the series is QC & Delivery.

Transcription System: Transcription & Reviews

Wednesday, April 18th, 2012

This is a series of posts on our human-powered audio transcription system. The following are links to the previous parts: Overview, Workflow, Certification.

The Transcription & Review subsystem is where the bulk of the work gets done. In transcription the file is played back and typed into something called a raw transcript. This is the first pass transcripts where the incomprehensible parts are marked with blanks. In review this raw transcript is checked against the audio mistakes. Timestamps and speaker tracking is also added during review. The output of both these steps produces a fairly accurate textual representation of the audio file.

In our workflow, we first break up the files into smaller parts. Our certified transcribers — the one’s who have successfully cleared the Transcription Test – can then login to their account and select these part files. Another innovation of our system is that we don’t actually assign files to them. Instead they are asked to choose from the files available. They can preview the file and check the quality before choosing. This creates a competition which in turn ensures that files get done quickly.

For performance monitoring we use a five point grading system; A+/Excellent to D/Poor. The files are graded after the review. Another small innovation of our system is the Diff Preview which shows the changes made during the review. It helps the reviewer to assess the quality of the raw transcript and grade accordingly. Based on the grades a Transcriber can be promoted to a Reviewer. There is a disputes and arbitration system in place too to investigate unfair grading.

Another innovative aspect of our system is it ensures a file is worked on by multiple transcribers and reviewers. The average for a 1 hour file is 15-20. More eyes and more ears on the file does wonders for the transcript quality. During Proofreading all the inconsistencies caused by this methodology are corrected. We will talk about more about Proofreading in the next part fo the series.

Till then if you want a high quality transcript of your audio file which has been checked multiple times by different people, then check out our transcription service today.

The next part of the series is available here.

Transcription System: Certification

Monday, April 16th, 2012

This is a series of posts on the Scribie.com transcription system. The previous parts are: OverviewWorkflow.

Our transcription system is 100% human powered. That makes the certification subsystem the first important component. This subsystem handles the process of certifying new transcribers and inducting them. To become a certified transcriber one has to apply and take Transcription Test. We publish guidelines on how the transcript should be prepared and provide recommendation for tools. The candidates are first added to a waiting list and invited for the test at their turn.

The test itself is a 3-6 minutes audio file which which they have to complete within 2 hours. We evaluate the submission and check the quality of transcript. We also look at the adherence to the guidelines and formatting. If everything is okay, they are certified as a Scribie.com transcriptionist and paid for the work done. We are closing in on the 2000 certified transcribers mark right now.

The goal of the transcription test is to ascertain whether a candidate is fit for this type of work. We get a lot of applications, but around 50% of them drop off at the test stage. The ones who pass through understand what to do and how to do it. They are given access to the next component which is the Transcription Subsystem. They can log in to their Scribie.com account anytime and choose from the available files. They get paid when their submissions are reviewed.

The number of active transcribers/reviewers, i.e. people who work regularly on Scribie.com, is around 1% of the total. The number might seem to be low, but 1% is considered a good active users ratio for internet services. To maintain this active users base, we certify new transcribers on a ongoing basis. New applicants are added to a wait list and are certified in turn. The Certification Subsystem manages the waiting list and sends out test invites as required.

Around 10% of our transcribers go on to become regulars. In fact few of our transcribers have been working working with us since the early days and are still active. That in itself is a testimonial to our system’s effectiveness.

If you are interested in working for us then please check our Freelance Transcription Program. If you want to outsource your audio transcription work to our certified transcribers then upload your files now.

The next part of the series talks about the Transcription and Reviews subsystem.

Transcription System: Workflow

Sunday, April 15th, 2012

This is a series on Scribie.com’s audio transcription system. The first part which provides an overview is here

Our workflow consists of five steps.

File Splitting -> Transcription -> Review -> Proofreading -> Delivery

We start by splitting the file into smaller parts. The file is split at the 6 minute boundary which produces one or more files of duration 6 minutes or shorter. This is the first little innovation of our transcription process. File splitting breaks down the work into smaller manageable chunks. It helps in many ways. The file can be worked on parallelly by number of transcribers. A huge amount of effort is not wasted if one part has to be re-done. Additionally, we can track the progress precisely.

Transcription is the typing part. On an average it takes around 15-20 minutes to transcribe a 6 minute file. For a lot of our transcribers–who are mostly home-based freelancers–this is not a huge investment of time. Therefore splitting increases the likely hood that the file will be transcribed quickly. In fact on an average it takes around 1 to 1.5 hours to complete the transcription part of a one hour file!

The accuracy of the transcript is very low at this stage; typically around 50 to 80%. Therefore we do a review. The transcript is checked against the audio and all mistakes are corrected. Time-coding and speaker tracking is also added at this stage. Review usually takes 5 to 8 minutes of effort. But it takes longer for all the parts to get reviewed because we have fewer reviewers than transcribers. This is by design since we promote only our best transcribers to reviewers. The review drastically improves the accuracy.

Once all parts are transcribed and reviewed, we can combine them together and prepare the final transcript. However one more round of review is required here. That’s because, since different parts are worked on by different people, there are bound to be inconsistencies. Proofreading is done by a one person who goes through all the parts together and corrects them. The proofreader is an employee of CGBiz LLC (our company). They are the best of the best we have. We train them and pay them a monthly salary rather than an hourly rate.

The transcript is almost done now. However things might not be perfect even now. The proofreader can make mistakes, some more research may be required for certain terms, etc. So before the delivery we do some random checks. We try to gauge whether the quality is indeed at the level we want it to be. We also use keyword analysis (tf-idf to be precise) to identify out-of-context terms and inconsistencies. We review it again if we are not happy with it. Over time we have found that a small percentage of files require re-review; around 2%. Those are generally the most difficult of files.

Once we are satisfied that the transcript is perfect, as best as it can be, we deliver the file. The file is converted into MS Word, Adobe PDF, OpenOffice Text and plain text formats and we notify the customer that the transcript is available for download.

All of the above happens in 1 day and is managed by our transcription system. We charge only $0.99 per minute of the audio for it. So if you want to get a high quality transcript quickly, please do try out our transcription service today.

The next part of the series talks about the Certification Subsystem.

Re-inventing Audio Transcription

Friday, April 13th, 2012

Last month we completed four years of our company. We launched CallGraph Skype Recorder in April of 2008 with intention of offering services around it. The transcription service was one of those and it quickly became the most popular. The past four years we have invested all our time and effort in developing a human-powered transcription system with a single goal in mind: deliver the best quality transcript with the lowest amount of effort. In this series of posts we are going to write about this system in-depth.

So why build another human-powered transcription system? Why not just use a Automatic Speech Recognition system. Speech recognition has been in the limelight recently most notably because of Siri which uses Nuance’s Technology. In fact in Google Glass it’s a central component. Even Evernote recently added the support for it. However all of these systems employ keyword recognition; eg. commands that are spoken aloud. Our requirement was conversational speech recognition. The technology for that is still very immature. In fact we tried out CMU Shpinx and results were so poor that we ruled it out.

The big issue with human-powered systems is that it produces inconsistent results. Transcription is very labor intensive. And just like any labor intensive workflow, if you do not have processes in place, you will not be able to control the quality. The typical transcription process involves one person doing the typing work and maybe, another person proofreading it. On an average it  takes around four hours to type one hour of audio and around the same amount to edit it. This increases the cost of transcription. And even after that the transcript is bound to have mistakes, thereby affecting its quality.

So that was the starting point for us. Our system manages the transcription process end-to-end. It’s like a machine where you input the audio file and it outputs a high quality transcript in one day. This system is powered by our certified transcriptionists who do all the work. We have a well defined workflow and a robust process in place. We use some Machine Learning and Information Retrieval tools as well, but for the most part, it is all done by hand.

With this system we have completed more than 3000 hours of audio transcription till date and managed to survive four years in a highly competitive market. The best part is that we a high return rate of customers. For a startup, it might not be a stellar achievement like Instagram’s, but we believe that we have built something substantial; a scalable and reliable transcription service. The next post will cover the first part of our system, the transcriber certification process. Till then, if you are in need of a transcription service then you should try out our transcription service today. You will not be disappointed.

The next part of the series can be found here.

Starting Weekend Deliveries

Monday, April 2nd, 2012

Our transcription service just got a bit better. Starting this month, we are moving from a 5-day work week to a 7-day one. This means we will deliver transcript orders on all days of the week. If an order is placed on Friday or Saturday, it will be delivered within 1 day; Saturday and Sunday respectively. Previously, all orders placed over the weekends were delivered on Monday. Not anymore. We will be delivering all orders within our usual 1-day turnaround time.

There’s no additional charge for weekend deliveries. The same rates ($0.99 per audio minute) and rules apply. Orders have to placed before 2:30 PM EST (US) to be delivered the next day. Orders that are placed after 2:30 PM EST will be delivered on the day following.

Get started today. Receive affordable, high quality transcripts of your audio files in one day. Start now.

https://scribie.com/files/upload/