What Happened to Theory Anyway? - PowerPoint PPT Presentation

About This Presentation
Title:

What Happened to Theory Anyway?

Description:

Title: CS206 --- Electronic Commerce Author: Jeff Ullman Last modified by: Jeffrey D. Ullman Created Date: 3/23/2002 8:14:09 PM Document presentation format – PowerPoint PPT presentation

Number of Views:92
Avg rating:3.0/5.0
Slides: 19
Provided by: Jeff554
Category:

less

Transcript and Presenter's Notes

Title: What Happened to Theory Anyway?


1
What Happened to Theory Anyway?
  • Role of Theory
  • What Makes a Good Theory?
  • Prospects

2
The 2 Most Annoying Things I Like to Say
  1. Theoreticians stop playing games and start
    worrying about what this stuff is good for.
  2. Implementers take a little time to appreciate
    the theory in the area where you are about to
    create a major hack.

3
Example Universal Relation
  • There is a theory about what happens when you
    imagine that all your relations are projections
    of a single relation.
  • Some things become simpler (natural joins, e.g.)
    other things become harder.
  • Ted Codd hated the idea.
  • Rule 217 Thou shalt not use the ideas of
    Maier-Ullman-Vardi.

4
UR --- (2)
  • Whatever you may think of the idea, it keeps
    getting rediscovered.
  • Moshe Vardi kept a file of over a dozen papers
    that thought they had discovered a remarkable,
    new interface to databases.
  • Good example of why it pays to learn the theory
    before you proceed.

5
But Theory is no Less Guilty
  • Confession Ive been guilty of at least the
    following
  • Working on stuff just because it is fun.
  • Or so that I could get a paper out and maybe
    tenure.
  • Or to impress my friends.

6
What Makes a Good Theory?
  • The first test is always whether it accomplishes
    something important.
  • But theories take a long time to develop.
  • Incremental work OK here.
  • Decide where to put effort by cleanliness
    theory should offer a high ratio of power to
    mechanism.

7
Example Relational Model
  • The relational model is the best example by far
    in our field.
  • One basic construct.
  • Good for anything.
  • Perfect for almost nothing.

8
Example Database logic
  • We thought Datalog was going to take over for
    SQL it didnt.
  • But it is natural and clean, and the study turned
    out to be worth the effort --- for reasons we
    didnt anticipate.
  • Prolog baroque execution model.
  • AI/logic way too much mechanism.

9
Application Information Integration
  • Logical rules seem better tuned to data
    transformation than querying.
  • name(XY) - first(X) last(Y)
  • Work of Halevy, collaborators, many others.

10
Application Data Exchange
  • Work of Fagin, Kolaitis, c.
  • Data sharing, peer-to-peer.
  • Uses not only logic, but related dependency
    theory and also some ideas from the
    universal-relation theory.

11
Application Software Security
  • Work of Monica Lam, c.
  • Datalog to express program analysis.
  • More than rules --- stratified negation, e.g.
  • Datalog compiled into BDDs to exploit regularity
    in data.
  • Atypical for normal databases.

12
Application Networks
  • Work of Hellerstein c.
  • Describes network properties and constraints in
    Datalog.
  • Clean, succinct descriptions of network
    protocols.
  • Rapid prototyping.

13
Future
  • Funny data models.
  • I mean XML.
  • Data with regularity.
  • Is there more to BDDs as a database
    implementation?
  • Data mining.
  • Lets get those terrorists.

14
Tree/Graph Data Models
  • It seems very hard to switch from relations to
    semistructured data models.
  • Which seem rather clean at first face.
  • Is there a compromise that will give us the
    composability of relational algebra clean
    queries in the Datalog (or even SQL) style?

15
Structure Within Data
  • BDDs work for apps when the data has regularity.
  • Aside BDDs are a notation for boolean functions
    that are terrible except for the functions one
    actually cares about.
  • BDDs have good algorithms for relational algebra
    operations.
  • Can the technique carry over to challenging DB
    apps, e.g., design DBs?

16
Data Mining
  • Database systems offers an under-appreciated view
    of data mining its queries, not building
    statistical models.
  • Example given a set of numbers, compute AVG, not
    the mean of the most likely Gaussian
    distribution.
  • Move beyond association rules.

17
Data Mining on Steroids
  • Biggest data-mining problem track terrorism
    while protecting privacy.
  • Petabytes of unstructured data.
  • x phoned y z bought ammonium nitrate with credit
    card c.
  • Algorithms for multijoin queries that cant
    possibly be evaluated fully.

18
Thanks
  • While its terrific to receive an award like
    this, it really acknowledges the work of students
    and colleagues who developed the ideas and those
    who, in the fullness of time, put them into
    practice.
Write a Comment
User Comments (0)
About PowerShow.com