1
00:00:07,470 --> 00:00:12,110
Despite the growth of open data sets that
are available to the public, it can still

2
00:00:12,110 --> 00:00:17,710
be difficult to discover data sets that are
both high quality and have clearly defined

3
00:00:17,710 --> 00:00:20,109
license and usage terms.

4
00:00:20,109 --> 00:00:25,829
To help solve this challenge, IBM created
the Data Asset eXchange, or "DAX,”, which

5
00:00:25,829 --> 00:00:28,180
we’ll introduce in this video.

6
00:00:28,180 --> 00:00:32,989
DAX provides a trusted source for finding
open data sets that are ready for to use in

7
00:00:32,989 --> 00:00:35,110
enterprise applications.

8
00:00:35,110 --> 00:00:40,750
These data sets and which cover a wide variety
of domains, including images, video, text,

9
00:00:40,750 --> 00:00:41,860
and audio.

10
00:00:41,860 --> 00:00:47,330
Because DAX provides a high level of curation
for data set quality, as well as licensing

11
00:00:47,330 --> 00:00:54,300
and usage terms, DAX data sets are typically
easier to adopt, whether in research or commercial

12
00:00:54,300 --> 00:00:55,990
projects.

13
00:00:55,990 --> 00:01:01,240
Wherever possible, DAX aims to make data sets
available under one of the variants of the

14
00:01:01,240 --> 00:01:07,650
CDLACommunity Data License Agreement, in order
to foster data sharing and collaboration.

15
00:01:07,650 --> 00:01:14,000
DAX also provides a single place to access
unique data sets, in particular from IBM Research

16
00:01:14,000 --> 00:01:15,470
projects.

17
00:01:15,470 --> 00:01:20,880
To make it easier for developers to get started
with using the data sets, DAX also provides

18
00:01:20,880 --> 00:01:27,310
tutorials in the form of notebooks that walk
through the basics of data cleaning, pre-processing,

19
00:01:27,310 --> 00:01:30,740
and exploratory analysis.

20
00:01:30,740 --> 00:01:36,910
For some data sets, there are also notebooks
illustrating how to perform more complex analysis,

21
00:01:36,910 --> 00:01:43,560
such as creating charts, statistical analysis,
time-series analysis, training machine learning

22
00:01:43,560 --> 00:01:48,990
models, and integrating deep learning via
using the Model Asset eXchange, (a project

23
00:01:48,990 --> 00:01:54,840
closely related to DAX and also available
on the IBM Developer website).

24
00:01:54,840 --> 00:02:00,990
In this way, DAX helps developers to create
end-to-end analytic and machine learning workflows

25
00:02:00,990 --> 00:02:07,250
and to consume open data and models with confidence
under clearly defined license terms.

26
00:02:07,250 --> 00:02:11,870
Let’s say you’ve found a data set that
might be of interest to you.

27
00:02:11,870 --> 00:02:17,709
On the data set page you can download the
compressed data set archive from cloud storage,

28
00:02:17,709 --> 00:02:23,980
explore the data set using Jupyter Notebooks,
review the data set metadata, such as format,

29
00:02:23,980 --> 00:02:29,420
licensing terms and size, and preview some
parts of the data set.

30
00:02:29,420 --> 00:02:34,160
Most data sets on DAX are complemented by
one or more Jupyter Notebooks that you can

31
00:02:34,160 --> 00:02:39,319
use to perform data cleaning, pre-processing,
and exploratory analysis.

32
00:02:39,319 --> 00:02:44,810
These notebooks run "as is"as is in Watson
Studio, IBM’s Data Sciencedata science platform.

33
00:02:44,810 --> 00:02:49,670
Jupyter Notebooks and Watson Studio are covered
later during in this course.

34
00:02:49,670 --> 00:02:55,620
In this video, you’ve learned about IBM’s
open data repository, the Data Asset eXchange.

35
00:02:55,620 --> 00:02:59,209
In the hands-on lab you’ll have a chance
to explore the repository.