Wednesday, 4 June 2008

Points of Interest 04. June, 2008

1. The evolution of language. It's not nature versus nurture but nature and nurture in my opinion: The Workings of Co-Evolution

2. The evolution of music. Check out the article on Nature if you can: The Evolution of Music

3. All hail our chimpanzee leaders: Do chimpanzees have a theory of mind? 30 years later

4. Ficition versus reality - the difference in a MRI: Meeting George Bush versus Meeting Cinderella

5. Buehler deserves some respect, bitches: Karl Bühler/Buhler/Buehler on the Evolution of Language

6. I'd nominate this headline for the most freaky one I've ever seen and beware the article itself is even stranger: Zombie caterpillars controlled by voodoo wasps

7. Oh well, too good to be true: Good News In Our DNA: Defects You Can Fix With Vitamins And Minerals

8. Cannabis makes you even more stupid (you have to be stupid to do drugs): Long-term Cannabis Users May Have Structural Brain Abnormalities

9. Micro-robots Dance On Something Smaller Than A Pin's Head

Saturday, 31 May 2008

Semantic Web

Today an interesting article was posted on New Scientist about the Semantic Web - unfortunately, not available for me yet. Everybody's talking about Web 2.0 and nobody knows what it's all about. The real thing that's going on at the moment is the so called Semantic Web project.
Tim Berners-Lee, so to say the founder of the World Wide Web and W3C head, suggested a new dimension of the Internet; Please note that the Internet and the WWW are not the same thing but often used equally and so do I. He expressed it by following words:

I have a dream for the Web [in which computers] become capable of analyzing all the data on the Web – the content, links, and transactions between people and computers. A 'Semantic Web', which should make this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to machines. The 'intelligent agents' people have touted for ages will finally materialize. (Wikipedia)

This is, by the way, one topic of my, hopefully, new course of study - computational linguistics. The idea is to categorise and connect ideas in a way that a machine can understand and interact on a (semi-)natural with humans, e.g. via meta-information:

Non-semantic web (Web 1.0 and Web 2.0):

<.item>Cat<./item>

Semantic web (part of Web 3.0):

<.animal Kingdom="Animalia" Phylum="Chordata" Class="Mammalia" Order="Carnivora" Family="Felidae" Genus="Felis">Cat<./animal> (Wikipedia)

This should create a Serendipity effect as many users already experience when they do 'wikihopping'. A very advanced usage would be that a user can now ask the web:

Person: "Oh, I have a lovely cat at home and could you please tell me the class of my cat?"
Semantic WWW: "Of course sir/madam. You're from Europe, Republic of Ireland so the most likely class would be Mammalia. May I show you some images for verification or do you want to know something more?"

or:

Person: "When did the Berlin wall go down?"
Semantic WWW: "In 1989. Do you want some videos related to the event?"

Because search engines are dumb and can't understand meaning, you'll receive often inaccurate answers:

1. BBC ON THIS DAY | 9 | 1989: The night the Wall came down
2. When did the Berlin Wall go down? - Blurtit
3. Why did the berlin wall come down? - Yahoo! Answers (Google)


The idea of meta data is, infact, not new. Via HTML, especially Dublin Core, you can realise meta data (via <.meta>) but almost nobody uses it because it is much more work to do and search engines do not use meta data to index websites anymore and many browsers don't care at all. This will also be one of the main problems for a global Semantic Web. Many coders/users just don't care about meta information. So you'll have to make them care with prescriptions or to make it interesting to use meta tags.

Thursday, 29 May 2008

A Kanji - A Day (29.05.08)

I'm a bit exicted since the fourth edition of Dungeons and Dragons will be released very soon (06.06.08). Nevertheless, I have to continue learning Japanese.


26. 山 means "a mountain" and is read:

On-Yomi:

a)
サン (san), e.g. 火山 (かざん) / a volcano

Kun-Yomi
:
b) やま (yama), e.g. 山場 (やまば) / the climax, the peak, the crisis


27. 四 means "four" and is read:

On-Yomi:

a)
シ (shi), e.g. 四月 (しがつ) / April

Kun-Yomi
:
b) よつ (yotsu), e.g. 四つ角 (よつかど) / a crossroads
c) よっつ (yottsu), e.g. 四つ (よっつ) / four
d) よん (yon), e.g. 四輪駆動 (よんりんくどう) / 4WD


28. 子 means "child", "kid", "baby" or "an infant" and is read:

On-Yomi:

a)
シ (shi), e.g. 子午線 (しごせん) / the meridian
b) ス (su), e.g. 様子 (ようす) / appearances, a look, the situation, the state of things, circumstances

Kun-Yomi
:
c) こ (ko), e.g. 子分 (こぶん) / a follower, a henchman, a following


29. 糸 means "thread", "yarn" or "string" and is read:

On-Yomi:

a)
シ (shi), e.g. 一糸乱れず (いっしみだれず) / in precise order, in perfect coordination

Kun-Yomi
:
b) いと (ito), e.g. 絹糸 (きぬいと) / silk thread


30. 字 means "a character", "a letter", "writting" or "script" and is read:

On-Yomi:

a)
ジ (ji), e.g. 文字 (もじ) / a character, a letter

Kun-Yomi
:
b) あざ (aza), e.g. 字 (あざ) / a section, a township


Speech examples can be found at the Japanese Kanji Dictionary.

Stroke order can be found at the Kanji Land.


First Class (30/80 signs): 一右雨円王音下火花貝学気休玉金九空月犬見五口校左三山四子糸字耳七車手十出女小上森人水正生青石赤先千川早草足村大男竹中虫町天田土二日入年白八百文本名木目夕立力林六

Wednesday, 28 May 2008

EEG: Code Switching & The Language You Think In

In 2004 a team of scientists under the lead of Alice Mado Proverbio, professor of cognitive electrophysiology at the Milano-Bicocca University in Milan, analysed the brain activity of fifteen subjects via EEG while they were shown sentences in their native and secondary language (L2). The subjects were interpreters for the European Union and fluent in English. They were shown sentences which "could be entirely in Italian or in English (unmixed); alternatively, the body of the sentence could be in English and the final word in Italian or vice versa (mixed)." (Proverbio 2004) The objective was to determine "[...] neural mechanisms subserving code switching [...]." (ibid.) They monitored an alteration in the brain waves as the participants where code switching.

This sounds quite interesting. Unfortunately, from a linguistic point of view, you can't compare two different languages this way I think. The slightly higher amplitude can rise due to the different morphology, syntax and phonology. Italian, as a Romance language, focus much more on grammatical morphemes than English.
Take the phonology: The IPA notes for Standard English (RP) 24 consonants, 21 vowels (Sauer 1990) compared to Standard Italian with 46 consonants (short and long consonants) and 27 vowels (Rogers & d'Arcangeli 2004).
Take the morphology: "Nouns in Italian have gender (masculine or feminine, but no neuter), and number (singular or plural). The gender and number is always shown by the leading article (definite or indefinite), and usually by the final vowel. [...] Adjectives, like nouns, have two genders and two numbers. Italian verb infinitives have one of three endings, either -are, -ere, or -ire. (Wikipedia) In English we have the third person pronouns, the Saxon genitive and only 8 possible inflection forms for verbs.
In contrast to the rudimentary English verb conjugation system the Italian language has a distinctive system derived from Latin. Furthermore, the lexicon from L1 may be stored deeper in the left hemisphere than the lexicon from L2.

Sentences like:

(A) Occorre impedire l’uso fraudolento delle carte di CREDITO.

in comparison to

(B) The fraudulent use of CREDIT cards must be prevented.

may not be equal and produce different results.

Earlier this year another paper of Proverbio (et al.) was published in Biological Psychology which is about the difference between native language, secondary and even tertiary language (L2, L3) usage and their representation in the brain. The brain activity was recognisably higher when they were shown words from their native language.

"Proverbio attributed the differences to the fact the brain absorbs the mother tongue at a time when it is also storing early visual, acoustic, emotional and other nonlinguistic knowledge." (MSNBC)

Unfortunately, this article is not available for my university and therefore out of reach. I doubt whether an EEG could so easily find the difference between first and second language. You can't easily compare an Italic language with a Germanic language that way; they were separated from their Indo-European roots 2000 years ago. At least you have to consider the difference between morphology, phonology, semantics and syntax. Semantics I havn't mentioned yet. Italian words have sometimes 'more or less meanings' attached they are not equal to e.g. German words.

Lingformant
MSNBC
Proverbio, M. Alice (et al.). Language switching mechanisms in simultaneous interpreters: an ERP study. In: Neuropsychologia 42 (12). 2004: 1636-1656.
Proverbio, M. Alice (et al.). Inferring native language from early bio-electrical activity. In: Biological Psychology. 2008.
Rogers & d'Arcangeli. Vowels of Italian. 2004.
Sauer, Walter. A Drillbook of English Phonetics. 1990.


_

Note:
I changed this article due to Ms. Proverbio's correction in the comments. The article on MSNBC was misleading since they mixed the content of the 2004 paper with the paper from 2008.

Tuesday, 27 May 2008

Aussagenlogik (Teil 2)

In Teil 1 wurde die Terminologie und die Syntax der Aussagenlogik besprochen. Um das ganze abzurunden beschäftigen wir uns nun mit dem letzten Teil der Aussagenlogik, der Semantik und dem Tableaux als Beweisverfahren.

In den vorherigen Gesetzen ist eine Formel als φ definiert und die aussagenlogische Belegung als g. Nun gibt es Belegungen die allgemeingültig oder gar unerfüllbar sind. Ersteres nennt man Tautologie (gr. τὸ αὐτό "dasselbe" und λόγος "Wort", "Rede", "Sinn"). "Sinnlos", aber nicht "unsinnig" seien die Tautologien, welche "nichts" bedeuten, schreibt Wittgenstein in seinem Tractatus1. Als Formel wird eine Tautologie folgendermaßen ausgedrückt:

⊨φ

Eine Tautologie ist also unabhängig der Wahrheitswerte ihrer Elemente wahr. Das einfachste Beispiel ist (A ∨ ¬A); Es ist also wahr oder nicht wahr. Das einfachste Beispiel eines Widerspruchs wäre im Umkehrschluss (¬A ∧ A). Shakespeare lässt seinen Hamlet ganz treffend sagen: "Sein oder nicht sein"2 - wobei er wohl noch keine Bekanntschaft mit Schrödingers Katze gemacht hat, denn die kann bekanntlicherweise lebendig und tot zugleich sein3.

Damit wäre nun die Semantik abgeschlossen. Werfen wir noch ein paar Definitionen in den Raum bevor wir überhaupt zu einem anspruchsvollen Beweisverfahren kommen.
Wenn man eine Menge von Formeln hat, definiert als Φ, dann folgt aus dieser Menge eine Formel ψ, wenn die Formel ψ die gleichen Belegungen erfüllt wie die Menge Φ:
Man schreibt dafür Φ ⊨ ψ und nennt die Elemente von Φ die Prämissen und ψ die Konklusion.4 Zum Beispiel ist (A ∧ B) ⊨ (A ∨ B), denn jede Belegung die den ersten Teil der Formel erfüllt, macht auch den zweiten Teil wahr. Sind zwei Formeln gleichwertig, dann spricht man von Äquivalenz. Die einfachste Äquivalenz wäre:

¬¬φ ≡ φ

In Teil 1 haben wir Gebrauch von Tabellen gemacht um die Wahrheitswerte herauszufinden, das ist allerdings zu aufwendig, denn je mehr Elemente eine Formel hat, desto größer (2n um genau zu sein; n = Anzahl der Elemente) wird unsere Tabelle und platzt schließlich aus allen Nähten. Deshalb wurde die Tableaux-Methode entwickelt um in einem Widerlegungsverfahren die Erfüllbarkeit von Formeln der Aussagenlogik zu (be-) widerlegen. Ich verwende hierfür das Beispiel aus Klabunde (et al.).5
Die Frage ist, wie man beweisen kann, dass A ⇒ (B ⇒ A) eine Tautologie ist. Wir wissen, dass eine Formel genau dann allgemeingültig ist, wenn die Negation dieser Formel unerfüllbar ist. Wir benutzen also ein Hintertürchen und beweisen die Unerfüllbarkeit:

¬(A ⇒ (B ⇒ A))

Um diese Aussage also wahr zu machen, muss A ⇒ (B ⇒ A) falsch gemacht werden. Vergegenwärtigen wir uns noch einmal die allgemeine Regel:

I ((φ ⇒ ψ)) = 1, falls I (φ) = 0 oder I (ψ) = 1, sonst I ((φ ⇒ ψ)) = 0

Es muss (B ⇒ A) = 0 (d.h. ¬(B ⇒ A) = 1) und A = 1 sein. Im Tableaux wird es folgendermaßen dargestellt:

1. ¬(A ⇒ (B ⇒ A))
2. A
3. ¬(B ⇒ A)

Zeile 2 kann nicht weiter zerlegt werden, aber Zeile 3 können wir zerlegen in:

4. B
5. ¬A

Oder anders: Damit ¬(B ⇒ A) wahr wird muss B (siehe Zeile 4) wahr und A (siehe Zeile 5) falsch sein. Nun steht in Zeile 2 aber dass A = 1 sein muss und in Zeile 5 steht, dass A = 0 sein muss. Dieser Widerspruch schließt das Tableaux und wir haben bewiesen, dass ¬(A ⇒ (B ⇒ A)) nicht zu erfüllen und damit (A ⇒ (B ⇒ A)) eine Tautologie ist. Die Reihenfolge der Expansionen ist dabei fakultativ. Ein Tableaux ist dann geschlossen, wenn φ als auch ¬φ darin vorkommen. Beweisbar ist eine Formel, wenn ¬φ in ein geschlossenes Tableaux übergeht. Man spricht formal von einem Theorem:

⊦φ

Um eine Tautologie handelt es sich:

wenn ⊦φ dann ⊨ φ

Auch für das Tableaux gibt es allgemeingültige Regeln, so haben wir in Zeile 3 folgendes Schema angewandt:

F→: ¬(φ ⇒ ψ)
________
φ
¬ψ

Diese Regel wird Expansionsregel genannt. Daraus lassen sich natürlich auch andere Expansionsregeln wie:

W∧: (φ ∧ ψ)
______
φ
ψ

erstellen. Mit dem Problem, dass (φ ∨ ψ) wahr ist, wenn das erste oder das zweite Element wahr ist muss man umgehen indem man im Tableaux eine Verzweigung in der betreffenden Zeile erstellt.

Dass die Möglichkeiten der Aussagenlogik begrenzt ist, habe ich schon Anfangs erwähnt. So ist es zum Beispiel nicht möglich über die Eigenschaften der Objekte zu sprechen. Dies ist das Gebiet der Prädikatenlogik, mit der ich mich demnächst beschäftigen werden. Ein anderes Thema, welches für mich im Moment relevant ist, ist MathML und LaTeX bzw. TeX um mathematische Formeln dazustellen. HTML alleine ist dazu kaum in der Lage.


_

1 Wittgenstein, Ludwig. Tractatus Logico-Philosophicus. These 6.11.
2 Shakespeare, William. Hamlet. 3. Aufzug, 1. Szene.
3 Schrödinger, Erwin. Naturwissenschaften, 48, 807; 49, 823; 50, 844.
4 Klabunde, Ralf (et al.). Computerlinguistik und Sprachtechnologie. 36.
5 Klabunde, Ralf (et al.). Computerlinguistik und Sprachtechnologie. 38ff.

Sunday, 25 May 2008

Points of Interest 25. May, 2008

1. Interesting and humorous, unfortunately I'm light years from any degree:
Lord of the Rings: an allegory of the PhD?

2. About the mascot of the Nara 1300th-anniversary. It's about the capital move from Fujiwara to Nara. The second part with the poem is interesting. It shows the complexity of Japanese poems: Sento out!

3. Interesting facts about the Ainu language: Ainu, Utari, Ainu

4. The powerless will always be powerless: Lacking power diminishes cognitive function

5. Reason versus emotion: The MRI of morality?

6. Be careful with the interpretation of MRIs: MRI - the new phrenology

7. Parents' influence on kids' behavior: Not much

8. What Chutes and Ladders has to do with learning Math

9. Can we learn from errors? What if we're running a nuclear power plant?

10. Doing algebra - it's the little things that count

11. Space and Syntax

Monday, 19 May 2008

Points of Interest 19. May, 2008

1. The title says it all. However, I must add one thing: Magpies are the only known birds so far who are aware of themselves:

Joshua Klein on “the amazing intelligence of crows”


2. A fascinating ancient compendium of runic inscriptions on burial headstones of warriors and heroes:

Danicorum Monumentorum


3. A humorous propaganda world map from 1911:

Asia on the World


4. An antique children's book about the cheerful cricket with a twisted story:

The Cheerful Cricket


A really good urban animation sequence found on Digital World Tokyo:

Stunning urban animation crosses genre boundaries - and walls