05.03.2013 Views

PhD thesis - School of Informatics - University of Edinburgh

PhD thesis - School of Informatics - University of Edinburgh

PhD thesis - School of Informatics - University of Edinburgh

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Chapter 2. Background and Theory 31<br />

Excuse Me... Kyaa Re Excuse me...What’s it?<br />

Meraa Dil Tere Pe Fidaa Re... I am smitten by you<br />

Bus Stop Pe Dekhaa Tujhe Pehli Baar I saw you first at the bus stop<br />

JhaTke MeN Ho Gayaa Tere Se Pyaar I fell in love as the bus lurched<br />

Excuse Me... aaN Bolnaa Excuse me...yes?<br />

MaiN Pehle Se Shaadi Shudaa Re... I am already married.<br />

Abhi To HooN Saalaa Roadpati I am the master <strong>of</strong> street<br />

Ladki ChaahuuN KaroRpati... I want a billionaire girl<br />

Race Course MeN Dekha Tujhe Pehli Baar I saw you at the Race Course<br />

Counter Pe Ho Gayaa Tere Se Pyaar I fell in love at the counter<br />

Excuse Me...Kyaa Re Excuse me...What’s it?<br />

Police MeN Hai Meraa MiyaaN Re My husband is in the police<br />

...Excuse Me...Yes Please Excuse me...Yes please<br />

Ban Jaa Mera Bhaiyaa Re Better become my brother<br />

Figure 2.5: Hinglish song lyrics, example taken from Kachru (2006)<br />

<strong>of</strong> a bouncer outside a club when deciding who can enter or not (Dunn, forthcoming).<br />

It can be concluded that English is influencing many languages. The anglicisa-<br />

tion <strong>of</strong> those languages is a complex process with various reasons and motivations.<br />

Foreign words or proper names pose substantial difficulties to NLP applications, not<br />

only because they are hard to process but also because, theoretically, they are infinite<br />

in number. Moreover, it is impossible to predict which foreign words will enter a lan-<br />

guage, let alone to create an exhaustive gazetteer <strong>of</strong> them. Therefore, the increasing use<br />

<strong>of</strong> English forms in different languages presents a challenge to NLP systems that are<br />

conventionally designed to handle monolingual text. The task <strong>of</strong> recognising English<br />

inclusions is the main focus <strong>of</strong> the work presented in this <strong>thesis</strong>. Before introducing<br />

an automatic classifier able to detect English language portions in otherwise mono-<br />

lingual text (Chapters 3 and 4), this chapter will review previous work on automatic,<br />

specifically mixed-lingual, language identification (Section 2.2).

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!