Coupled pattern learner: Difference between revisions

Revision as of 22:36, 26 August 2015

Latest revision as of 04:29, 26 June 2025

(20 intermediate revisions by 14 users not shown)

Line 1:

{{Short description|Machine learning algorithm}}

{{Orphan|date=March 2012}}

'''Coupled Pattern Learner''' (CPL) is a [[machine learning]] algorithm which couples the [[semi-supervised learning]] of categories and relations to forestall the problem of semantic drift associated with boot-strap learning methods.

Coupled Pattern Learner (CPL) is a [[machine learning]] algorithm which couples the [[semi-supervised learning]] of categories and relations to forestall the problem of semantic drift associated with boot-strap learning methods.

== Coupled Pattern Learner ==

[[Semi-supervised learning]] approaches using a small number of labeled examples with many unlabeled examples are usually unreliable as they produce an internally consistent, but incorrect set of extractions. CPL solves this problem by simultaneously learning classifiersclassifiers for many different categories and relations in the presence of an [[Ontology (information science)|ontology]] definingdefining constraints that couple the training of these classifiersclassifiers. It was introduced by Andrew Carlson, Justin Betteridge, Estevam R. Hruschka Jr. and Tom M. Mitchell in 2009.<ref name=cbl2009>{{cite journal|last=Carlson|first=Andrew|author2=Justin Betteridge |author3=Estevam R. Hruschka Jr. |author4= Tom M. Mitchell |year=2009|title=Coupling semi-supervised learning of categories and relations|journal=Proceedings of the NAACL HLT 2009 Workshop on Semi-Supervised Learning for Natural Language Processing |publisher=Association for Computational Linguistics|___location=Colorado, USA|pages=1–9|isbn=9781932432381 |url=http://dl.acm.org/citation.cfm?id=1621829.1621830}}</ref><ref name=cpl2010>{{cite journalbook|last=Carlson|first=Andrew|author2=Justin Betteridge |author3=Richard C. Wang |author4=Estevam R. Hruschka Jr. |author5= Tom M. Mitchell |year=2010|title=Coupled semi-supervised learning for information extraction|journal=Proceedings of the third ACM international conference on Web search and data mining |chapter=Coupled semi-supervised learning for information extraction |year=2010|publisher=ACM|___location=NY, USA|pages=101–110|urldoi=http://dl10.acm.org1145/citation.cfm?doid=1718487.1718501|isbn=9781605588896|doi-access=free}}</ref>

== CPL Overviewoverview==

CPL is an approach to [[semi-supervised learning]] that yields more accurate results by coupling the training of many information extractors. Basic idea behind CPL is that semi-supervised training of a single type of extractor such as ‘coach’ is much more difficultdifficult than simultaneously training many extractors that cover a variety of inter-related entity and relation types. Using prior knowledge about the relationships between these different entities and relations CPL makes unlabeled data as a useful constraint during training. For e.g., ‘coach(x)’ implies ‘person(x)’ and ‘not sport(x)’.

== CPL Descriptiondescription ==

=== Coupling of Predicatespredicates ===

CPL primarily relies on the notion of coupling the [[learning]] of multiple functions so as to constrain the semi-supervised learning problem. CPL constrains the learned function in two ways.

# Sharing among same-arity predicates according to logical relations

Line 19 ⟶ 18:

=== Relation argument type-checking ===

This is a type checking information used to couple the learning of relations and categories. For example, the arguments of the ‘ceoOf’ relation are declared to be of the categories ‘person’ and ‘company’. CPL does not promote a pair of noun phrases as an instance of a relation unless the two noun phrases are classifiedclassified as belonging to the correct argument types.

=== Algorithm Descriptiondescription ===

Following is a quick summary of the CPL algorithm.<ref name=cpl2010 />

Line 27 ⟶ 26:

Output: Trusted instances/patterns for each predicate

'''for''' i=1,2,...,∞ '''do'''

'''foreach''' predicate p in O '''do'''

EXTRACT candidate instances/contextual patterns using recently promoted patterns/instances;

FILTER candidates that violate coupling;

RANK candidate instances/patterns;

PROMOTE top candidates;

'''end'''

'''end'''

==== Inputs ====

A large [[Text corpus|corpus]] of Part-Of-Speech tagged sentences and an initial ontology with predefinedpredefined categories, relations, mutually exclusive relationships between same-arity predicates, subset relationships between some categories, seed instances for all predicates, and seed patterns for the categories.

==== Candidate extraction ====

CPL findsfinds new candidate instances by using newly promoted patterns to extract the noun phrases that co-occur with those patterns in the text corpus. CPL extracts,

* Category Instances

* Category Patterns

Line 45 ⟶ 44:

* Relation Patterns

==== Candidate Filteringfiltering ====

Candidate instances and patterns are filteredfiltered to maintain high precision, and to avoid extremely specificspecific patterns. An instance is only considered for assessment if it co-occurs with at least two promoted patterns in the text corpus, and if its co-occurrence count with all promoted patterns is at least three times greater than its co-occurrence count with negative patterns.

==== Candidate Rankingranking ====

CPL ranks candidate instances using the number of promoted patterns that they co-occur with so that candidates that occur with more patterns are ranked higher. Patterns are ranked using an estimate of the precision of each pattern.

==== Candidate Promotionpromotion ====

CPL ranks the candidates according to their assessment scores and promotes at most 100 instances and 5 patterns for each predicate. Instances and patterns are only promoted if they co-occur with at least two promoted patterns or instances, respectively.

== Meta-Bootstrap Learner ==

Meta-Bootstrap Learner (MBL) was also proposed by the authors of CPL in.<ref name=cpl2010 /> Meta-Bootstrap learner couples the training of multiple extraction techniques with a multi-view constraint, which requires the extractors to agree. It makes addition of coupling constraints on top of existing extraction algorithms, while treating them as black boxes, feasible. MBL assumes that the errors made by different extraction techniques are independent. Following is a quick summary of MBL.

'''Input''': An ontology O, a set of extractors ε

'''Output''': Trusted instances for each predicate

'''for''' i=1,2,...,∞ '''do'''

'''foreach''' predicate p in O '''do'''

'''foreach''' extractor e in ε '''do'''

Extract new candidates for p using e with recently promoted instances;

'''end'''

FILTER candidates that violate mutual-exclusion or type-checking constraints;

PROMOTE candidates that were extracted by all extractors;

'''end'''

'''end'''

Line 72 ⟶ 71:

== Applications ==

In their paper <ref name=cbl2009 /> authors have presented results showing the potential of CPL to contribute new facts to existing repository of semantic knowledge, Freebase <ref>{{cite journal|year=2009 |title=Freebase data dumps |publisher=Metaweb Technologies |url=http://download.freebase.com/datadumps/ |url-status=dead |archiveurl=https://web.archive.org/web/20111206102101/http://download.freebase.com/datadumps/ |archivedate=December 6, 2011 }}</ref>

== See also ==

Line 83 ⟶ 82:

==References==

* {{cite journal|last=Liu|first=Qiuhua |author2=Xuejun Liao |author3=Lawrence Carin |year=2008|title=Semi-supervised multitask learning|journal=NIPS}}

* {{cite journal|last=Shinyama|first=Yusuke|author2=Satoshi Sekine|year=2006|title=Preemptive information extraction using unrestricted relation discovery|journal=HLT-NAACLNaacl}}

* {{cite journal|last=Chang|first=Ming-Wei|author2=Lev-Arie Ratinov |author3=Dan Roth |year=2007|title=Guiding semi-supervision with constraint driven learning|journal=ACL}}

* {{cite journal|last=Banko|first=Michele|author2=Michael J. Cafarella |author3=Stephen Soderland |author4=Matt Broadhead |author5=[[Oren Etzioni]] |author5-link=Oren Etzioni |year=2007|title=Open information extraction from the web|journal=IJCAI}}

* {{cite journalbook|last=Blum|first=Avrim|author2=Tom Mitchell|yeartitle=1998Proceedings of the eleventh annual conference on Computational learning theory |titlechapter=Combining labeled and unlabeled data with co-training |journalyear=COLT1998|pages=92–100|doi=10.1145/279943.279962|isbn=1581130570|s2cid=207228399 }}

* {{cite journal|last=Riloff|first=Ellen|author2=Rosie Jones|year=1999|title=Learning dictionaries for information extraction by multi-level bootstrapping|journal=AAAI}}

* {{cite journal|last=Rosenfeld|first=Benjamin|author2=Ronen Feldman|year=2007|title=Using corpus statistics on entities to improve semi-supervised relation extraction from the web|journal=ACL}}