structlm-labtalk
-
Upload
prachi-jain -
Category
Documents
-
view
220 -
download
0
Transcript of structlm-labtalk
-
8/2/2019 structlm-labtalk
1/24
Incorporating Language Modelinginto the Inference Network
Retrieval Framework
Don Metzler
-
8/2/2019 structlm-labtalk
2/24
Motivation
Great deal of information lost when forming queries Example: stemming information retrieval
InQuery informal (tf.idfobservation estimates) structured queries via inference network framework
Language Modeling formal (probabilistic model of documents)
unstructured InQuery + Language modeling
formal
structured
-
8/2/2019 structlm-labtalk
3/24
Motivation
Simple idea:
Replace tf.idfestimates in inference network
framework with language modeling estimates
Result is a system based on ideas from language
modeling that allows powerful structured queries
Overall goal:
Do as well as, or better than, InQuery within thismore formal framework
-
8/2/2019 structlm-labtalk
4/24
Outline
Review
Inference Network Framework
Language Modeling Combined Approach
Results
-
8/2/2019 structlm-labtalk
5/24
Review of Inference Networks
Directed acyclic graph
Compactly represents joint probabilitydistribution over a set of continuous and/ordiscrete random variables
Each node has a conditional probability tableassociated with it
Network topology defines conditionalindependence assumptions among nodes
In general, inference is NP-hard
-
8/2/2019 structlm-labtalk
6/24
Inference Network Framework
Node types
document (di)
concept (ri)
query (qi)
information need (I)
Set evidence at
document nodes
Run belief propagation
Documents are scored
by P(I= true | di= true)
-
8/2/2019 structlm-labtalk
7/24
Network Semantics
All events in network are binary
Events associated with each node:
di document iis observedri representation concept iis observed
qi query representation iis observed
I information need is satisfied
-
8/2/2019 structlm-labtalk
8/24
Query Language
-
8/2/2019 structlm-labtalk
9/24
Example Query
Unstructured:stemming information retrieval
Structured:#wand(1.5 #syn(#phrase(information retrieval) IR)
2.0 stemming)
-
8/2/2019 structlm-labtalk
10/24
Want to compute bel(n) for each node n in
the network (bel(n) = P(n = true | di= true))
Term/proximity node beliefs (InQuery)
Belief Propagation
1||log
5.0||log
||
||
5.15.0
)1()(
,
,
,
,
,
!
!
!
C
tf
C
idf
D
d
tf
tftf
idftfdbdbrbel
i
i
i
i
i
dr
r
avg
i
dr
dr
dr
rdrdb = default belief
tfr,di= number of times
representation r is matched in
document di
|di| = length of document i
|D|avg= average doc. length
|C|=
collection length
-
8/2/2019 structlm-labtalk
11/24
Belief Nodes In general, marginalization
is very costly
Assuming a nice functional
form, via link matrices,
marginalization becomeseasy
p1, ,pn are the beliefs at
the parent nodes ofq
W= w1 + + wn
!
!
!
!
!
!
!
i
Ww
iwand
i
iand
i
ii
wsum
ii
sum
n
i
ior
not
ipqbel
pqbelW
pw
qbel
n
pqbel
ppqbel
pqbel
pqbel
)/(
1max
1
)(
)(
)(
)(
),,max()(
)1(1)(
1)(
-
-
8/2/2019 structlm-labtalk
12/24
Link MatrixExample
A B
Q
-
8/2/2019 structlm-labtalk
13/24
Language Modeling
Models document generation as a stochastic
process
Assume words are drawn i.i.d. from anunderlying multinomial distribution
Use smoothed maximum likelihood estimate:
Query likelihood model:
!!
Qq
ddn qPqqQP )|()|( 1 UU-
||
)1(
||
)|(,
C
cf
d
tfwP w
dw
d PPU !
-
8/2/2019 structlm-labtalk
14/24
Rather than use tf.idfestimates for bel(r), use
smoothed language modeling estimates:
Use Jelinek-Mercer smoothing throughout forsimplicity
Inference Network + LM
||)1(
||)|(
)|()(
,
C
cf
d
tfdrP
drPrbel
r
i
dr
i
i
iPP !
!
-
8/2/2019 structlm-labtalk
15/24
Combining Evidence
InQuery combines query evidence via #wsum
operator i.e. all queries are of the form #wsum( )
#wsum does not work for combined model
resulting scoring function lacks idfcomponent
Must use #wand instead
Can be interpreted as normalized weighted
averages
arithmetic (InQuery)
geometric (combined model)
-
8/2/2019 structlm-labtalk
16/24
#wsumvs. #wand
-
8/2/2019 structlm-labtalk
17/24
Relation to Query Likelihood
Model subsumes query likelihood model
Given a query Q =q1,q2, ,qn (qi is a single
term) convert it to the following structuredquery:
#and(q1 q2 qn)
Result is query likelihood model
-
8/2/2019 structlm-labtalk
18/24
Smoothing
InQuery crude smoothing via default belief
Proximity node smoothing
Single term smoothing Other proximity node smoothing
Each type of proximity node can be
smoothed differently
-
8/2/2019 structlm-labtalk
19/24
Experiments
Data sets TREC 4 ad hoc(manual & automatic queries)
TREC 6, 7, and 8 ad hoc
Comparison
Query likelihood (QL)
InQuery
Combined approach (S
tructLM) Single term node smoothing = 0.6
Other proximity node smoothing = 0.1
-
8/2/2019 structlm-labtalk
20/24
Example Query
Topic: Is there data available tosuggest that capital punishment is a
deterrent to crime?
Manual structured query:#wsum(1.0 #wsum(1.0 capital 1.0 punishment
1.0 deterrent 1.0 crime
2.0 #uw20(capital punishment deterrent)
1.0 #phrase(capital punishment)
1.0 #passage200 (1.0 capital 1.0 punishment
1.0 deterrent 1.0 crime
1.0 #phrase(capital punishment)))
-
8/2/2019 structlm-labtalk
21/24
-
8/2/2019 structlm-labtalk
22/24
Proximity Node Smoothing
-
8/2/2019 structlm-labtalk
23/24
Conclusions
Good structured queries help
Combines inference networks structured
query language with formal languagemodeling probability estimates
Performs competitively against InQuery
Subsumes query likelihood model
-
8/2/2019 structlm-labtalk
24/24
Future Work
Smoothing
Try other smoothing techniques
Find optimal parameters for each node type Combine LM and tf.idfdocument
representations
Other estimates for bel(r)
Theoretical considerations