Both in timing and in content this could be a good lead-in to the University of Washington's Intro to Data Science class that looks like it will have more of a focus on 'big data', NoSQL, Hadoop, data mining, etc.
Is there a list anywhere of Coursera courses that have a self-study option? Unless there's something I'm misunderstanding about what that option means, these are the only Coursera classes I'm ever likely to use, because of my schedule. It seems odd that they don't appear to differentiate them.
There was a good deal of relational algebra / theoretical background that I had not wrestled with extensively elsewhere. This aspect was fun from a "mind-expanding" perspective. I have never found a specific need to directly apply this information in day-to-day activities.
The practical aspect of the course was heavily focused on SQL (Queries). Lots of interactive exercises that relied upon SQLLite which uses ANSI Standard SQL. This helped with focusing on constructing queries that did not use vendor specific features. There was a decent amount of XML related processing and querying. NoSQL was only mentioned in passing - and was essentially an afterthought in the 2011 course.
The format of the course was closer to traditional college format (with relatively long lectures). I generally don't care for this in online classes. Due to time constraints, I skipped lectures in areas with which I was familiar and went straight to the exercises/quizzes. I thought the quizzes were really excellent, and the problems presented (and online validation/checking of problems) was really great.
As one with a fairly extensive background in databases (Oracle DBA Cert, years of experience with Oracle, Informix, DB2) I really found the course helpful and well worth the time. It does not provide much practice in Data Modeling or NoSQL. All in all one of the best technical educational experiences I have had in any format.
I took it at Stanford, but it was using the same online content.
The course is probably way more theoretical than you'd expect. It goes a lot into relational algebra, schema design, and UML.
The online SQL quizzes are pretty cool little puzzles but the only truly exciting part of the class was creating a website using a large database as the final project, which I can't imagine will be offered in the online version.
If you don't know anything about databases or are trying to learn SQL, it could definitely be worthwhile.
> The online SQL quizzes are pretty cool little puzzles but the only truly exciting part of the class was creating a website using a large database as the final project, which I can't imagine will be offered in the online version.
The year I took it, they couldn't get the Oracle licenses sorted out in time so they forewent the entire final project. They ended up just talking about XPath and Data Warehousing instead (both of which seemed way too enterprise-y to me at the time to commit either to long-term memory). The schema design stuff I found helpful, if only to drill into my mind the importance of data normalization.
It sounds like I'll hopefully get something out of it, given I'm purely self-taught post-high school. Some rigour to my existing practical capacity would be useful.
A: Yes, I think so. I'm self thought guy myself, at least, when it comes to programming, and I really enjoyed watching the relational algebra videos and then the SQL ones (never got to the modelling part though, got all my learning time sucked by a big work project)... go through them at 1.5x or 2x speed and you won't regret it, at least if you had no encounter with relational algebra before, as it feels good to know there's some nice and simple math behind SQL :) (though now I want even more to shoot the guy that invented SQL's ugly syntax and all the database implementors that piled tons of "semi-compatible" shit on top...)
I took this one a year ago. I had zero idea how to do anything with databases. I won't say I am DBA material, but I'm confident in my abilities to design and deploy a database.
As far as I'm concerned, this is one of the best OCW courses available. Very challenging at the beginning. Though I think there are areas that could be improved upon, it will take you from zero to proficient. Although I've never used views in real life, I have used triggers; I know how to normalize okay (I guess?), and can figure out the logic to run fairly complex queries over raw data.
What you won't learn about is creating optimized and fast queries, though this is something you'll get with experience, trial and error, and research. The NoSQL section is wanting, but the goal of the course is to teach you relational theory and general SQL concepts, which it does a very good job at.
Of course, if you want to be a PostgreSQL or Oracle master, you'll probably have to buy the requisite books or comb the online docs. If you want to learn how to integrate the DBs into your programming language or platform, that is a self-study item as well.
I took the course and I found it very useful even though I had done good amount of work with databases before.
The sections on relational algebra and relational models are really good. The course covers the basics of db - sql, constraints, views, triggers and authorization very well. I particularly liked the coverage on transactions, an area that most web programmers will benefit from.
With respect to xml, I found the level of detail to be quite good again but I don't really use much of xquery or xslt these days.
While the course does cover Olap and nosql, it is quite light compared to the relational and xml units of the course.
I see that they have added a new section on json in the new version of the course. Some of the questions in the homework and exams are slightly tricky but nothing too hard.
I didn't really get to see the office hours videos, so can't comment on that. Overall, I think it is an excellent introductory db course. Just make sure you don't fall back on the schedule as some of the material can be quite heavy.
I took it a year ago and really enjoyed it. I had been using SQL quite a bit before, but I had never taken a database course at university. It was nice to get exposed to the theory , and I learnt things about SQL that became immediately useful at work. Also, Professor Widom has a nice sense of humor.
Nice and light. If you learned SQL and DBs "in the field" and had no exposure to relational algebra watch it at 1.5x - 2x speed, with pen & paper in your hands, to fill in your theory gaps... but I never got to the modelling part so can't give you a complete answer...
The theoretical stuff wasn't immediately applicable but it changed my mental model of databases and that has helped me tremendously in my work since then.
In the video, Prof. Widom mentions that the platform they are using now, Class2Go, has been developed at Stanford and is open source. A quick look at http://class.stanford.edu/ shows that it depends on external components - YouTube for video, and Piazza for forums - but it's good to have an open platform, and one of the aims is to avoid lock-in of content.
https://www.coursera.org/course/datasci
http://news.ycombinator.com/item?id=4881581