Question Answering - Application and Challenges
-
Upload
jens-lehmann -
Category
Science
-
view
1.158 -
download
0
Transcript of Question Answering - Application and Challenges
![Page 1: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/1.jpg)
Question Answering:
Applications and ChallengesWDAqua R&D Week
Contributions from Mohnish Dubey,Konrad Hoffner and Christina Unger
2016-02-09 Jens Lehmann
![Page 2: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/2.jpg)
Contents
![Page 3: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/3.jpg)
QA Overview and Recent Developments
![Page 4: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/4.jpg)
Definition of QA [3]
1 Ask questions in natural language
Example (Natural Language)
Which books are written by Dan Brown?
![Page 5: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/5.jpg)
Definition of QA [3]
1 Ask questions in natural language
2 People use their own terminology
Example
Which books are written by Dan Brown?Which books have Dan Brown as their author?What are notable works of Dan Brown?
![Page 6: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/6.jpg)
Definition of QA [3]
1 Ask questions in natural language
2 People use their own terminology
3 Receive a concise answer
Example (Formal Language)
select ?book {?book dbo:author dbr:Dan Brown.}
![Page 7: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/7.jpg)
Why Semantic Question Answering?
Motivation
• Large amounts of knowledge modelled as RDF data
• Contrary to the Web of Documents, Semantic Web is made formachines
• Lay users cannot use SPARQL
Approach Expressiveness Ease of Use Ease of Setup
SPARQL Endpoint ++ −− ++Facetted Search − + −1/+2
Question Answering + ++ −−1/−2
1new development2existing approach, new domain
![Page 8: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/8.jpg)
Dimensions of QA [4]
![Page 9: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/9.jpg)
QA Timeline
![Page 10: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/10.jpg)
First QA Systems
BASEBALL [2] 1961:
• First QA system in 1969
• Answered questions about the US baseball league over a period ofone year
LUNAR [5] 1971:
• Developed for NASA to answer questions on lunar geology
• First evaluated QA system with 78 % correctly answered questionsfrom lunar science convention visitors in 1971
• Compiles English into a “meaning representation language”
![Page 11: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/11.jpg)
2009: Wolfram—Alpha
![Page 12: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/12.jpg)
2009: Wolfram—Alpha
![Page 13: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/13.jpg)
2011: IBM Watson [1] wins at Jeopardy
![Page 14: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/14.jpg)
IBM Watson [1]
• Developed by DeepQA research team at IBM
• Won Jeopardy against human experts
• Massively parallel, runs on a supercomputer
• Aggregates many different sources, including semantic
• Industry project: some algorithms are published, but not open source
• Brmson1 platform approximates IBM Watson using open sourcetechnology
• QA engine Yoda QA2 inspired by Watson runs on a normal server PC
1http://brmlab.cz/project/brmson2http://ailao.eu/yodaqa/
![Page 15: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/15.jpg)
QA Challenges
![Page 16: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/16.jpg)
No challenge
Who developed Minecraft?
SELECT DISTINCT ?uri
WHERE {
dbpedia:Minecraft dbpedia -owl:developer ?uri .
}
![Page 17: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/17.jpg)
Lexical gap
How tall is Michael Jordan?
SELECT DISTINCT ?num
WHERE {
dbpedia:Michael_Jordan dbpedia -owl:height ?num .
}
![Page 18: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/18.jpg)
Lexical knowledge
Give me all taikonauts.
SELECT DISTINCT ?uri
WHERE {
?uri rdf:type dbpedia -owl:Astronaut .
?uri dbpedia -owl:nationality dbpedia:China .
}
![Page 19: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/19.jpg)
Lexical knowledge
• Who was the last princess of Joseon?
→ !BOUND :successor
• Which of the Beatles is still alive?
→ !BOUND :deathDate
![Page 20: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/20.jpg)
Ambiguity
How many banks are there in London?
SELECT DISTINCT count (?bank)
WHERE {
?bank a lgdo:Bank.
}
SELECT DISTINCT count (?bank)
WHERE {
?bank a lgdo:Riverbank.
}
![Page 21: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/21.jpg)
Granularity
Give me a list of all lakes in Denmark.
SELECT DISTINCT ?uri
WHERE {
{ ?uri rdf:type yago:LakesOfDenmark . }
UNION
{ ?uri rdf:type dbpedia -owl:Lake .
?uri dbpedia -owl:country dbpedia:Denmark . }
}
![Page 22: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/22.jpg)
Granularity
Give me a list of all lakes in Denmark.
SELECT DISTINCT ?uri
WHERE {
{ ?uri rdf:type yago:LakesOfDenmark . }
UNION
{ ?uri rdf:type dbpedia -owl:Lake .
?uri dbpedia -owl:country dbpedia:Denmark . }
}
![Page 23: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/23.jpg)
Alternative Resources
Who killed John Lennon?
SELECT DISTINCT ?uri
WHERE {
?uri rdf:type dbo:Person.
?uri dbp:conviction res:Death of John Lennon.
![Page 24: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/24.jpg)
Adjectives
Give me all communist countries.
SELECT DISTINCT ?uri
WHERE {
?uri rdf:type dbpedia -owl:Country .
?uri dbpedia -owl:governmentType dbpedia:Communism .
}
![Page 25: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/25.jpg)
Adjectives
Give me all animals that are extinct.
SELECT DISTINCT ?uri
WHERE {
?uri rdf:type dbpedia -owl:Animal .
?uri dbpedia -owl:conservationStatus ’EX’ .
}
![Page 26: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/26.jpg)
Clausal Query
How many companies were founded in the same year as Google?
SELECT COUNT(DISTINCT ?c)
WHERE {
?c rdf:type dbo:Company .
?c dbo:foundingYear ?year .
res:Google dbo:foundingYear ?year .
}
![Page 27: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/27.jpg)
Crossover of QA Systems Required
What’s the unemployment rate of females in European capitals in 2011 ?
- What are European Capitals?- QA (e.g. AskNow) over ontology answers e.g. “Berlin”
- What amount of the female population in Berlin were unemployed in2011 ?
- Now, the Data cube model and CubeQA could be used.Ensemble learning and QA system combination.
![Page 28: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/28.jpg)
Unanswerable Query
What is the best fruit to eat?
Who will be the next president of USA?
→ Will QA systems be expected to detect unanswerable queries (andprovide entertaining answers) in the future?
![Page 29: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/29.jpg)
Large-scale linguistic resources
![Page 30: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/30.jpg)
Large-scale linguistic resources
high
→ POS: adjective
→ forms: higher, highest
→ related form: heightPOS: noun
→ antonym: low
![Page 31: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/31.jpg)
Large-scale linguistic resources
![Page 32: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/32.jpg)
Declarative lexical knowledge
Ontology lexica capture rich linguistic information
• word forms
• part of speech
• subcategorization
• meaning
about how elements of a particular ontology are verbalized in a particularlanguage.
![Page 33: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/33.jpg)
![Page 34: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/34.jpg)
lemon (Lexicon Model for Ontologies)
lemon provides a meta-model for describing ontol-ogy lexica with RDF.
Semantics by reference
The meaning of lexical entries is specified by pointing to elements in anontology.
![Page 35: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/35.jpg)
lemon (Lexicon Model for Ontologies)
lemon provides a meta-model for describing ontol-ogy lexica with RDF.
Semantics by reference
The meaning of lexical entries is specified by pointing to elements in anontology.
![Page 36: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/36.jpg)
Machine learning approaches
• Joint building of meaning representations and instatiation w.r.t.dataset
→ Berant, Chou, Frostig, Liang: Semantic parsing on Freebase from
question-answer pairs. EMNLP 2013.
• Joint instantiation w.r.t. dataset
→ Xu, Feng, Zhao: Xser.
![Page 37: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/37.jpg)
The Semantic Web as interlingua
![Page 38: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/38.jpg)
Multilinguality
¿Que alto es el Hallasan?Wie hoch ist der Hallasan?Quelle est l’altitude du Hallasan?
SELECT DISTINCT ?height
WHERE {
dbpedia:Hallasan dbpedia -owl:elevation ?height .
}
![Page 39: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/39.jpg)
Scalability
• Distributed data
→ aggregating information from different datasets(Question Answering over linked data)
• Conflicting data
→ fact validation
• Missing and incomplete data
![Page 40: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/40.jpg)
Non-factoid questions
• What is the difference between impressionism and expressionism?
• How do I make cheese cake?
• What do most Americans think of gun control?
• What is the role of TRH in hypertension?
• How do histone methyltransferases cause histone modification?
![Page 41: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/41.jpg)
QAMEL Project
![Page 42: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/42.jpg)
HOBBIT Project
EU Big Linked Data Benchmarking Project
![Page 43: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/43.jpg)
QA Applications
![Page 44: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/44.jpg)
Personal assistants
• Personal assistants on smartphones with voice interface
• Need small memory, CPU and battery footprint
• Calendar events
• Weather
• Route planning
![Page 45: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/45.jpg)
Access to Encyclopedic Knowledge
”The world’s knowledge in your pocket“
• World knowledge available in DBpedia, Wikidata etc.
• Used by many SQA systems, such as ISOFT, CASIA, Intui,TBSL, AskNow, SINA
• Basis of many QALD evaluation initiative tasks
![Page 46: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/46.jpg)
When other devices are inconvenient . . .
• Crisis situations
• In-car QA systems
• Childcare ;-)
![Page 47: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/47.jpg)
Smart Home and Entertainment
• Will get increasingly complex → people will ask devices more complexquestions
• Networked devices (Internet of Things)
![Page 48: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/48.jpg)
Search
• QA is becoming part of main stream search engines
• Google:• Voice Search in 2008• Knowledge Graph in 2012• Question Intent Understanding in 2015
• Google can understand superlatives, ordered items, time e.g. ”Whowas the U.S. President when the Angels won the World Series?”
![Page 49: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/49.jpg)
Semantic QA Community
What can we do?
![Page 51: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/51.jpg)
The End
Jens Lehmannhttp://jens-lehmann.org
University of BonnFraunhofer IAIS
Thanks for your attention!
![Page 52: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/52.jpg)
References
A. M. Gliozzo and A. Kalyanpur.Predicting lexical answer types in open domain QA.International Journal On Semantic Web and Information Systems,8(3):74–88, July 2012.
B. F. Green Jr, A. K. Wolf, C. Chomsky, and K. Laughery.Baseball: an automatic question-answerer.In Papers presented at the May 9-11, 1961, western jointIRE-AIEE-ACM computer conference, pages 219–224. ACM, 1961.
L. Hirschman and R. Gaizauskas.Natural language question answering: the view from here.natural language engineering, 7(4):275–300, 2001.
V. Lopez, V. Uren, M. Sabou, and E. Motta.Is Question Answering fit for the Semantic Web?: A survey.Semantic Web Journal, 2(2):125–155, April 2011.
W. A. Woods.Semantics for a question-answering system, volume 27.Garland Pub., 1979.
![Page 53: Question Answering - Application and Challenges](https://reader033.fdocuments.us/reader033/viewer/2022042907/587c49a81a28abc62c8b472f/html5/thumbnails/53.jpg)
Outlook