Merge pull request #2 from workforce-data-initiative/mb/tpot

Add draft TPOT documentation

docs/tpot/index.md

@@ -0,0 +1,84 @@
+# Training Provider Outcomes Toolkit
+
+The Training Provider Outcomes Toolkit (TPOT) is a collection of tools for
+securely collecting, connecting, analyzing, aggregating, and publishing data on
+wage and employment outcomes for education and training participants.
+
+State and local workforce development boards aim to improve employment and
+workforce investment in their regions.
+
+## Collecting data from training providers
+
+There are many hundreds or thousands of training providers within the purview
+of each workforce development board. Each one must securely upload their
+participant data to their workforce board in order to be eligible for federal
+funds. The resulting aggregate statistics on participant outcomes will be
+invaluable marketing materials for successful programs, regardless of federal
+funding. This means that the workforce development boards must be equipped to
+receive and validate the data.
+
+Training providers range from small trade apprenticeships to community colleges
+to multi-state organizations, with a wide range of data sophistication. The
+way(s) in which the workforce data board collects participant outcomes must be
+easy and accessible to all organizations. At the same time, it must be easy
+for the board itself to automatically process and validate the datasets.
+
+### The Data Package
+
+In order to support these use-cases, the toolkit uses the [Data Package]()
+specification in order to encapsulate and describe training provider outcome
+data in a systematic fashion.
+
+#### Create a Data Package specification
+
+Each workforce development board should create a data package specification
+that defines the specific fields and values that they require the training
+providers to use.
+
+The [**Data Packagist**](http://datapackagist.okfnlabs.org/) tool allows you to
+easily define these packages in the browser. You can also view existing package
+specifications with the [online data package
+viewer](http://data.okfn.org/tools/view).  Finally, you can validate your
+specification with an [online validator](http://data.okfn.org/tools/validate).
+
+### Uploading website
+
+Once a data package specification is created, it may be used to easily
+customize a self-hosted website that enables validating and uploading data
+packages from training providers. This site is designed to be simple to modify
+and setup a custom instance for each training provider. A demo uploading
+website is available at [send.dataatwork.org](http://send.dataatwork.org). The
+source code is available for customization [on
+GitHub](https://github.com/workforce-data-initiative/datapackagist).
+
+### The data warehouse
+
+The upload website is designed to securely upload the data packages to a
+data warehouse for each workforce board.  _Stay tuned for updates._
+
+## Data matching
+
+Once the workforce board has collected the training providers' outcomes, it
+must link the participant information with wage and employment information from
+other departments. In order to facilitate this matching, we have developed an
+easy-to-use and open-source deduplication package,
+[SuperDeduper](https://github.com/dssg/superdeduper). This allows for a
+combination of exact record linkage and probabalistic fuzzy matching in order
+to link as many participants as possible to the other datasets.
+
+## Data processing
+
+A [wide variety of tools](http://frictionlessdata.io/tools/) have been
+developed that support the Data Package format. It is possible to load your
+data package into [many analysis
+languages](http://frictionlessdata.io/tools/#libraries) including
+[Python](https://github.com/frictionlessdata/datapackage-py) and
+[R](https://github.com/frictionlessdata/datapackage-py). It's also possible to
+load the data packages into [many database
+systems](http://frictionlessdata.io/tools/#use-data-packages-with-) including
+[SQL Server](https://github.com/bimlscript/BETDPI) and
+[PostgreSQL](https://github.com/okfn/dptools/blob/master/bin/load-postgresql.py)
+ . Work is underway to support
+[Excel](https://github.com/frictionlessdata/ideas/issues/41) and [Google
+Sheets](https://github.com/okfn/data.okfn.org/issues/24).
+

mkdocs.yml

@@ -1 +1,5 @@
-site_name: My Docs
+site_name: Data at Work
+
+pages:
+- Home: index.md
+- TPOT: tpot/index.md