Gerhard Weikum Max Planck Institute for Informatics & Saarland University weikum

Gerhard Weikum Max Planck Institute for Informatics & Saarland Universityhttp://www.mpi-inf.mpg.de/~weikum/

Semantic Search:from Names and Phrasesto Entities and Relations

Acknowledgements

Big Picture: Opportunities Now !

KB Population

Info Extraction Semantic Authoring

Entity Linkage

Web of D

ataW

eb o

f Use

rs &

Con

tent

s

Very Large Knowledge Bases

Semantic Docs

Disambiguation

Big Picture: Opportunities Now !

KB Population

Info Extraction Semantic Authoring

Entity Linkage

Web of D

ataW

eb o

f Use

rs &

Con

tent

s

Very Large Knowledge Bases

Semantic Docs

Disambiguation

This talk:How Do We Search this World ofKnowledge, Data, and Text(and cope with ambiguity)

for Knowledge Harvestingsee talks at College de Franceand at VLDB School in Kunming

http://richard.cyganiak.de/2007/10/lod/lod-datasets_2011-09-19_colored.png

Web of Data: RDF, Tables, Microdata

YAGO

Cyc

TextRunner/ReVerb

WikiTaxonomy/WikiNet

SUMO

ConceptNet 5

BabelNet

ReadTheWeb

30 Bio. SPO triples (RDF) and growing



http://www.trueknowledge.com/













http://sig.ma/

http://sig.ma/

http://sig.ma/

http://sig.ma/


Web of Data: RDF, Tables, Microdata

YAGO

30 Bio. SPO triples (RDF) and growing

• 10M entities in 350K classes• 120M facts for 100 relations• 100 languages• 95% accuracy

• 4M entities in 250 classes• 500M facts for 6000 properties• live updates

• 25M entities in 2000 topics• 100M facts for 4000 properties• powers Google knowledge graph

Ennio_Morricone type composerEnnio_Morricone type GrammyAwardWinnercomposer subclassOf musicianEnnio_Morricone bornIn RomeRome locatedIn ItalyEnnio_Morricone created Ecstasy_of_GoldEnnio_Morricone wroteMusicFor The_Good,_the_Bad_,and_the_UglySergio_Leone directed The_Good,_the_Bad_,and_the_Ugly



owl:s

ameAs

rdf.freebase.com/ns/en.rome

owl:sameAs

owl:sameAs

data.nytimes.com/51688803696189142301

Coord

geonames.org/3169070/roma

N 41° 54' 10'' E 12° 29' 2''

dbpprop:citizenOf

dbpedia.org/resource/Rome

rdf:ty

pe

rdfs:subclassOf

yago/wordnet:Actor109765278

rdf:ty

pe

rdfs:subclassOfyago/wikicategory:ItalianComposer

yago/wordnet: Artist109812338

prop:actedInimdb.com/name/nm0910607/

Linked RDF Triples on the Web

prop: composedMusicFor

imdb.com/title/tt0361748/

dbpedia.org/resource/Ennio_Morricone

500 Mio. links

Embedding (RDF) Microdata in HTML Pages

May 2, 2011

Maestro Morricone will perform on the stage of the Smetana Hall to conduct the Czech National Symphony Orchestra and Choir. The concert will feature both Classical compositions and soundtracks such asthe Ecstasy of Gold.In programme two concerts for July 14th and 15th.

<html … May 2, 2011

<div typeof=event:music>

<span id="Maestro_Morricone">Maestro Morricone<a rel="sameAs"resource="dbpedia/Ennio_Morricone "/></span>…<span property = "event:location" >Smetana Hall </span>…<span property="rdf:type"resource="yago:performance">The concert </span> will feature …<span property="event:date" content="14-07-2011"></span>July 1

</div>

Supported by RDFaand microformats like schema.org

Outline

Opportunities Now

Entity Name Disambiguation

Question Answering

Disambiguation Reloaded

Wrap-Up

Semantic Search Today

Semantic Search Today (1)


Select ?x Where { ?x type composer [western movie] .?x wasBornIn ?y . ?y locatedIn Europe . }


Select ?x Where { ?x type composer .?x participatedIn ?y . ?y type western_film . }

Semantic Search Today (4)Key problem in semantic search: diversity and ambiguity of names and phrases !

Outline

Opportunities Now


Question Answering


Wrap-Up


Three Different NLP Problems

Harry fought with you know who. He defeats the dark lord.

1) named-entity detection: segment & label by HMM or CRF (e.g. Stanford NER tagger)

2) co-reference resolution: link to preceding NP (trained classifier over linguistic features)3) named-entity disambiguation: map each mention (name) to canonical entity (entry in KB)

Three NLP tasks:

HarryPotter

DirtyHarry

LordVoldemort

The Who(band)

Prince Harryof England

3-23

Sergio talked to Ennio aboutEli‘s role in theEcstasy scene. This sequence onthe graveyardwas a highlight inSergio‘s trilogyof western films.

Named Entity Disambiguation

D5 Overview May 30, 2011

Sergio means Sergio_LeoneSergio means Serge_GainsbourgEnnio means Ennio_AntonelliEnnio means Ennio_MorriconeEli means Eli_(bible)Eli means ExtremeLightInfrastructureEli means Eli_WallachEcstasy means Ecstasy_(drug)Ecstasy means Ecstasy_of_Goldtrilogy means Star_Wars_Trilogytrilogy means Lord_of_the_Ringstrilogy means Dollars_Trilogy … … …

KB

Eli (bible)

Eli Wallach

Mentions(surface names)

Entities(meanings)

Dollars Trilogy

Lord of the Rings

Star Wars Trilogy

Benny Andersson

Benny Goodman

Ecstasy of Gold

Ecstasy (drug)

?

3-24


Mention-Entity Graph

Dollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy (drug)

Eli (bible)

Eli Wallach

KB+Stats

weighted undirected graph with two types of nodes

Popularity(m,e):• freq(e|m)• length(e)• #links(e)

Similarity (m,e):• cos/Dice/KL (context(m), context(e))

bag-of-words orlanguage model:words, bigrams, phrases

3-25



Dollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy (drug)

Eli (bible)

Eli Wallach

KB+Stats


Popularity(m,e):• freq(e|m)• length(e)• #links(e)


jointmapping

3-26


27 / 20

Dollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy(drug)

Eli (bible)

Eli Wallach

KB+Stats


Popularity(m,e):• freq(m,e|m)• length(e)• #links(e)


Coherence (e,e‘):• dist(types)• overlap(links)• overlap (anchor words)


3-27


28 / 20

KB+Stats





American Jewsfilm actorsartistsAcademy Award winners

Metallica songsEnnio Morricone songsartifactssoundtrack music

spaghetti westernsfilm trilogiesmoviesartifactsDollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy (drug)

Eli (bible)

Eli Wallach


3-28


29 / 20

KB+Stats





http://.../wiki/Dollars_Trilogyhttp://.../wiki/The_Good,_the_Bad, _the_Uglyhttp://.../wiki/Clint_Eastwoodhttp://.../wiki/Honorary_Academy_Award

http://.../wiki/The_Good,_the_Bad,_the_Uglyhttp://.../wiki/Metallicahttp://.../wiki/Bellagio_(casino)http://.../wiki/Ennio_Morricone

http://.../wiki/Sergio_Leonehttp://.../wiki/The_Good,_the_Bad,_the_Uglyhttp://.../wiki/For_a_Few_Dollars_Morehttp://.../wiki/Ennio_MorriconeDollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy (drug)

Eli (bible)

Eli Wallach


3-29


30 / 20

KB+StatsPopularity(m,e):• freq(m,e|m)• length(e)• #links(e)



Metallica on Morricone tributeBellagio water fountain showYo-Yo MaEnnio Morricone composition

The Magnificent SevenThe Good, the Bad, and the UglyClint EastwoodUniversity of Texas at Austin

For a Few Dollars MoreThe Good, the Bad, and the UglyMan with No Name trilogysoundtrack by Ennio Morricone


Dollars Trilogy

Lord of the Rings

Star Wars

Ecstasy of Gold

Ecstasy (drug)

Eli (bible)

Eli Wallach


3-30

Joint Mapping

• Build mention-entity graph or joint-inference factor graph from knowledge and statistics in KB• Compute high-likelihood mapping (ML or MAP) or dense subgraph such that: each m is connected to exactly one e (or at most one e)

9030

5100

100

50 20

50

90

80 90

30

10 10

20

30

30

3-31

Coherence Graph Algorithm

• Compute dense subgraph to maximize min weighted degree among entity nodes such that: each m is connected to exactly one e (or at most one e)• Greedy approximation: iteratively remove weakest entity and its edges• Keep alternative solutions, then use local/randomized search

9030

5100

100

50 50

90

80 90

30

10 20

10

20

30

30

[J. Hoffart et al.: EMNLP‘11]140

180

50

470

145

230

3-32

Mention-Entity Popularity Weights

• Collect hyperlink anchor-text / link-target pairs from• Wikipedia redirects• Wikipedia links between articles• Interwiki links between Wikipedia editions• Web links pointing to Wikipedia articles

…• Build statistics to estimate P[entity | name]

• Need dictionary with entities‘ names:• full names: Arnold Alois Schwarzenegger, Los Angeles, Microsoft Corp.• short names: Arnold, Arnie, Mr. Schwarzenegger, New York, Microsoft, …• nicknames & aliases: Terminator, City of Angels, Evil Empire, …• acronyms: LA, UCLA, MS, MSFT• role names: the Austrian action hero, Californian governor, CEO of MS, …

… plus gender info (useful for resolving pronouns in context): Bill and Melinda met at MS. They fell in love and he kissed her.

[Milne/Witten 2008, Spitkovsky/Chang 2012]

3-33

Mention-Entity Similarity Edges

Extent of partial matches Weight of matched words

Precompute characteristic keyphrases q for each entity e:anchor texts or noun phrases in e page with high PMI:

)()(

),()(~)|(

mcontextinekeyphrasesq

mcover(q)distqscoremescore

1

)|(#~)|(qw

cover(q)w

e)|weight(w

ewweight

cover(q)oflengthwordsmatchingeqscore

)()(),(log),(efreqqfreq

eqfreqeqweight

Match keyphrase q of candidate e in context of mention m

Compute overall similarity of context(m) and candidate e

„Metallica tribute to Ennio Morricone“

The Ecstasy piece was covered by Metallica on the Morricone tribute album.

3-34

Entity-Entity Coherence EdgesPrecompute overlap of incoming links for entities e1 and e2

))2(),1(min(log||log))2()1(log())2,1(max(log1

eineinEeineineein~e2)coh(e1,-mw

Alternatively compute overlap of anchor texts for e1 and e2

or overlap of keyphrases, or similarity of bag-of-words, or …

)2()1()2()1(

engramsengramsengramsengrams

~e2)coh(e1,-ngram

Optionally combine with type distance of e1 and e2(e.g., Jaccard index for type instances)

For special types of e1 and e2 (locations, people, etc.)use spatial or temporal distance

3-35

AIDA: Accurate Online Disambiguation

http://www.mpi-inf.mpg.de/yago-naga/aida/3-36

http://www.mpi-inf.mpg.de/yago-naga/aida/


AIDA: Very Difficult Example

3-38



AIDA: Very Difficult Example

3-39


Some NED Online Tools forJ. Hoffart et al.: EMNLP 2011, VLDB 2011https://d5gate.ag5.mpi-sb.mpg.de/webaida/P. Ferragina, U. Scaella: CIKM 2010http://tagme.di.unipi.it/R. Isele, C. Bizer: VLDB 2012http://spotlight.dbpedia.org/demo/index.htmlReuters Open Calaishttp://viewer.opencalais.com/ S. Kulkarni, A. Singh, G. Ramakrishnan, S. Chakrabarti: KDD 2009http://www.cse.iitb.ac.in/soumen/doc/CSAW/D. Milne, I. Witten: CIKM 2008http://wikipedia-miner.cms.waikato.ac.nz/demos/annotate/

perhaps more

some use Stanford NER tagger for detecting mentionshttp://nlp.stanford.edu/software/CRF-NER.shtml

3-42

https://d5gate.ag5.mpi-sb.mpg.de/webaida/

http://tagme.di.unipi.it/

http://spotlight.dbpedia.org/demo/index.html

http://viewer.opencalais.com/

http://www.cse.iitb.ac.in/soumen/doc/CSAW/

http://wikipedia-miner.cms.waikato.ac.nz/demos/annotate/

http://wikipedia-miner.cms.waikato.ac.nz/demos/annotate/

http://nlp.stanford.edu/software/CRF-NER.shtml

NED: Experimental EvaluationBenchmark:• Extended CoNLL 2003 dataset: 1400 newswire articles• originally annotated with mention markup (NER), now with NED mappings to Yago and Freebase• difficult texts: … Australia beats India … Australian_Cricket_Team … White House talks to Kreml … President_of_the_USA … EDS made a contract with … HP_Enterprise_Services

Results:Best: AIDA method with prior+sim+coh + robustness test82% precision @100% recall, 87% mean average precisionComparison to other methods, see paper

J. Hoffart et al.: Robust Disambiguation of Named Entities in Text, EMNLP 2011http://www.mpi-inf.mpg.de/yago-naga/aida/

3-43

Ongoing Research & Remaining Challenges• More efficient graph algorithms (multicore, etc.)

• Short and difficult texts: • tweets, headlines, etc.• fictional texts: novels, song lyrics, etc.• incoherent texts

• Disambiguation beyond entity names:• coreferences: pronouns, paraphrases, etc.• common nouns, verbal phrases (general WSD)

• Leverage deep-parsing structures, leverage semantic types Example: Page played Kashmir on his Gibson

subj obj

mod

• Allow mentions of unknown entities, mapped to null

• Structured Web data: tables and lists

3-44

Variants of NED at Web Scale

• How to run this on big batch of 1 Mio. input texts? partition inputs across distributed machines, organize dictionary appropriately, … exploit cross-document contexts

• How to handle Web-scale inputs (100 Mio. pages) restricted to a set of interesting entities? (e.g. tracking politicians and companies)

Tools can map short text onto entities in a few seconds

3-45

Outline

Opportunities Now


Question Answering


Wrap-Up


Deep Question Answering

99 cents got me a 4-pack of Ytterlig coasters from this Swedish chain

This town is known as "Sin City" & its downtown is "Glitter Gulch"

William Wilkinson's "An Account of the Principalities of Wallachia and Moldavia" inspired this author's most famous novel

As of 2010, this is the only former Yugoslav republic in the EU

YAGO

knowledgeback-ends

questionclassification &decomposition

D. Ferrucci et al.: Building Watson. AI Magazine, Fall 2010.IBM Journal of R&D 56(3/4), 2012: This is Watson.

Semantic Keyword Search Need to map (groups of) keywords onto entities & relationshipsbased on name-entity similarities/probabilities

q: composer Rome scores westerns

[Ilyas et al. Sigmod‘10]

Media Composervideo editor

Western Digital

Rome(Italy)

goal infootball

film music

composer(creatorof music)

Rome(NY)

LazioRoma

western movies

western world

Western (airline)ASRoma

Western (NY)

… born in … … plays for … … used in … … recorded at …

Natural Language Questions are Natural

Who composed scores for westerns and is from Rome?

translate question into Sparql query:• dependency parsing to decompose question• mapping of question units onto entities, classes, relations

Who composedscores for westernsand is from Rome?

map resultsinto tabular or visual presentationor speech

From Questions to Queries

NL question:

Who composed scores for westerns and is from Rome?

scores for westerns

is from Rome Who composed scores

Dependency parsing exposes structure of question „triploids“ (sub-cues)

2-50

From Triploids to TriplesWho composed scores for westerns and is from Rome?

Who is from Rome

Who composed scores

scores for westerns

?x composed scores

?x bornIn Rome

scores contributesTo ?y?y type westernMovie

?x type composer?x composed ?s

?s contributesTo ?y

?s type music

2-51

Pattern Dictionary for Relations[N. Nakashole et al.: EMNLP 2012]

WordNet-style dictionary/taxonomy for relational phrases based on SOL patterns (syntactic-lexical-ontological)

• Relational phrases can be synonymous

• One relational phrase can subsume another

• Relational phrases are typed

Problem: cope with language diversity & ambiguityExample: composed …, wrote …, created …, …

“graduated from” “obtained degree in * from”“and $PRP ADJ advisor” “under the supervision of”

“wife of” “ spouse of”

<person> graduated from <university><singer> released <album><singer> covered <song> <book> covered <event>

PATTY: Pattern Taxonomy for Relations[N. Nakashole et al.: EMNLP 2012, demo at VLDB 2012]

350 000 SOL patterns with 4 Mio. instancesDerived from large data (Wikipedia, NYT, ClueWeb)by scalable sequence miningaccessible at: www.mpi-inf.mpg.de/yago-naga/patty

http://www.mpi-inf.mpg.de/yago-naga/patty

Disambiguation Mapping for TriploidsWho composed scores for westerns and is from Rome?

composed

composedscores

scores for

westerns

is from

Rome

Who

q1

q2

q3

q4

Combinatorial Optimization by ILP (with type constraints etc.)

e: Rome (Italy)e: Lazio Roma

c: personc: musiciane: WHO

r: createdr: wroteCompositionr: wroteSoftware

c:soundtrackr: soundtrackForr: shootsGoalFor

r: bornInr: actedIn

c: western moviee: Western Digital

wei

ghte

d ed

ges

(coh

eren

ce, s

imila

rity,

etc

.)

Relaxing Overconstrained QueriesSelect ?p Where {

?p composed ?s . ?s type music . ?s for ?m . ?m type movie .?p bornIn Rome . }

Select ?p Where {

?p composed ?s . ?s type music . ?s for ?m . ?m type movie [western] .?p bornIn Rome . }

Select ?p Where {

?p ?rel1 ?s [composed] . ?s type music . ?s ?rel2 ?m . ?m type movie [western] .?p bornIn Rome . }

with extended SPARQL-FullText: SPOX quad patterns

(S. Elbassuoni et al.: CIKM‘10, ESWC’11, SIGIR‘12)

Select ?p Where {?p composed ?s . ?s type music . ?s for ?m . ?m type movie [western] .?p bornIn Rome . }

Preliminary Results (M. Yahya et al.: WWW‘12, EMNLP‘12)

http://www.mpi-inf.mpg.de/yago-naga/deanna/

http://www.mpi-inf.mpg.de/yago-naga/deanna/

Outline

Opportunities Now


Question Answering


Wrap-Up


Disambiguation MappingWho composed scores for westerns and is from Rome?

composed

composedscores

scores for

westerns

is from

Rome

Who

q1

q2

q3

q4

e:Rome (Italy)e:Lazio Roma

c:personc:musiciane:WHO

r:createdr:wroteCompositionr:wroteSoftware

c:soundtrackr:soundtrackForr:shootsGoalFor

r:bornInr:actedIn

c:western moviee:Western Digital

wei

ghte

d ed

ges

(coh

eren

ce, s

imila

rity,

etc

.)

Selection: Xi Assignment: YijJointMapping: Zkl

[M.Yahya et al.: EMNLP‘12]

Disambig. Mapping: Objective FunctionWho composed scores for westerns and is from Rome?

composed

composedscores

scores for

westerns

is from

Rome

Who

q1

q2

q3

q4





r:bornInr:actedIn


wei

ghte

d ed

ges

(coh

eren

ce, s

imila

rity,

etc

.)


maximize i,j wij Yij + k,l vkl Zkl +… subject to:1) Yij Xi for all i,j2) j Yij 1 for all i3) Zkl i,j Yik and Zkl j Yil for all k,l4) Xi,Yij,Zkl {0,1}

wijvkl

Disambig. Mapping: ConstraintsWho composed scores for westerns and is from Rome?

composed

composedscores

scores for

westerns

is from

Rome

Who

q1

q2

q3

q4





r:bornInr:actedIn


wei

ghte

d ed

ges

(coh

eren

ce, s

imila

rity,

etc

.)


maximize i,j wij Yij + k,l vkl Zkl +… subject to:5) Qhi = 1 g Qhg = 3 for all h,i6) Xi + Xg 1 for all mutually exclusive i,g7) Qhi = 1 g,j Qhg Ygj = 1 for relation nodes j

wijvkl

Selection: Qhi

Disambig. Mapping: Type ConstraintsWho composed scores for westerns and is from Rome?

composed

composedscores

scores for

westerns

is from

Rome

Who

q1

q2

q3

q4


c:personc:musiciane: WHO



r:bornInr:actedIn


wei

ghte

d ed

ges

(coh

eren

ce, s

imila

rity,

etc

.)


maximize i,j wij Yij + k,l vkl Zkl +… subject to:8) Yij = 1 and j is relation node and Zkj=1 and Zjl=1 domain(j) types(k) and range(j) types(l)

wijvkl

Selection: Qhi

ILP optimizers like Gurobisolve this in 1 or 2 seconds

Outline

Opportunities Now


Question Answering


Wrap-Up


Summary

• Web of Data & Knowledge & Text (RDF + Phrases) Calls for Semantic Search by Entities, Classes & Relations

• Diversity & Ambiguity of Names and Phrases Calls for Disambiguation Mapping

• Strong Story for Entity Name Disambiguation

• Ongoing Work on Relation Phrase Disambiguation

• Cornerstone of Question Answering with Natural Language or Advanced Keywords

Great opportunity towards next-generation searchChallenging problems: robustness, scale, dynamics & transfer

Take-Home Message

Solve „Who composed the Ecstasy and other pieces for westerns?“

can solve semantic search with natural-language disambiguation

Gerhard Weikum Max Planck Institute for Informatics & Saarland University weikum

Documents

Transcript of Gerhard Weikum Max Planck Institute for Informatics & Saarland University weikum