Machine learning for transliteration

1. A method comprising:

receiving a plurality of resources, the plurality of resources including a plurality of anchor text, wherein anchor text found in a specific resource of the plurality of resources links to a different resource identified by a uniform resource locator of the different resource;

determining one or more potential transliterations from the plurality of anchor text;

sorting the one or more potential transliterations, based on likelihoods of the one or more potential transliterations co-occurring with text that identifies a same resource or location;

identifying one or more potential transliteration pairs from the one or more potential transliterations, wherein each potential transliteration pair includes a first anchor text in a first writing system and a second anchor text in a second writing system, the second anchor text and the first anchor text linking to a same resource;

for each potential transliteration pair,

comparing the first anchor text and the second anchor text; and

first classifying the potential transliteration pair as being a transliteration pair or not being a transliteration pair, based on the comparison;

first sorting the first classified potential transliteration pairs, based on likelihoods of the first classified potential transliteration pairs being transliteration pairs, to produce first sorted potential transliteration pairs;

identifying a subset of the first sorted potential transliteration pairs, the subset including potential transliteration pairs classified as being transliteration pairs and potential transliteration pairs classified as not being transliteration pairs;

for each potential transliteration pair in the subset,

aligning the first anchor text and the second anchor text; and

extracting one or more edits from the alignment;

generating a classification model, based on the one or more edits and the subset; and

second classifying each of the first classified potential transliteration pairs as being a transliteration pair or not being a transliteration pair, using the classification model.

2. The method of claim 1 , where comparing the first anchor text and the second anchor text includes:

determining a first edit distance between the first anchor text and the second anchor text; and

comparing the first edit distance to a first threshold value;

wherein the aligning is based on minimizing the first edit distance.

3. The method of claim 2 , where the first threshold value is the lesser of a length of the first anchor text and a length of the second anchor text.

4. The method of claim 3 , where the potential transliteration pair is first classified as not being a transliteration pair when the first edit distance is greater than the first threshold value, and the potential transliteration pair is first classified as being a transliteration pair when the first edit distance is less than the first threshold value.

5. The method of claim 1 , where generating the classification model includes:

associating each of the one or more edits to a feature; and

generating a feature weight for each feature.

6. The method of claim 5 , where the second classifying includes:

for each of the first classified potential transliteration pairs,

comparing the first classified potential transliteration pair to the one or more features in the classification model;

determining one or more feature weights, based on the comparison; and

summing the one or more feature weights to produce a classification score.

7. The method of claim 6 , further comprising:

for the one or more edits associated with the first writing system,

second sorting the second classified potential transliteration pairs, based on likelihoods of the second classified potential transliteration pairs being transliteration pairs; and

for each second sorted potential transliteration pair,

reclassifying the second sorted potential transliteration pair as not being a transliteration pair when its corresponding classification score indicates that the second sorted potential transliteration pair is not a transliteration pair;

reclassifying the second sorted potential transliteration pair as a best potential transliteration pair when the second sorted potential transliteration pair has the highest likelihood of being a transliteration pair and its corresponding classification score indicates that the second sorted potential transliteration pair is a transliteration pair;

determining a second edit distance between the second sorted potential transliteration pair and the best potential transliteration pair;

reclassifying the second sorted potential transliteration pair as being a transliteration pair when its second edit distance is less than a second threshold value and its corresponding classification score indicates that the second sorted potential transliteration pair is a transliteration pair; and

reclassifying the second sorted potential transliteration pair as not being a transliteration pair when its second edit distance is greater than the second threshold value.

8. The method of claim 7 , further comprising:

for the one or more edits associated with the second writing system,

third sorting the reclassified potential transliteration pairs, based on likelihoods of the reclassified potential transliteration pairs being transliteration pairs; and

for each third sorted potential transliteration pair classified as being a transliteration pair,

reclassifying the third sorted potential transliteration pair as not being a transliteration pair when its corresponding classification score indicates that the third sorted potential transliteration pair is not a transliteration pair;

reclassifying the third sorted potential transliteration pair as a best potential transliteration pair when the third sorted potential transliteration pair has the highest likelihood of being a transliteration pair and its corresponding classification score indicates that the third sorted potential transliteration pair is a transliteration pair;

determining a third edit distance between the third sorted potential transliteration pair and the best potential transliteration pair;

reclassifying the third sorted potential transliteration pair as being a transliteration pair when its third edit distance is less than a third threshold value and its corresponding classification score indicates that the third sorted potential transliteration pair is a transliteration pair; and

reclassifying the third sorted potential transliteration pair as not being a transliteration pair when its third edit distance is greater than the third threshold value.

9. The method of claim 1 , where the classification model uses a support vector machine (SVM).

10. The method of claim 1 , where the likelihoods are calculated using log likelihood ratios.

11. A method comprising:

receiving a plurality of resources, the plurality of resources including a plurality of anchor text, wherein anchor text found in a specific resource of the plurality of resources links to a different resource identified by a uniform resource locator of the different resource;

determining one or more potential transliterations from the plurality of anchor text;

identifying one or more potential transliteration pairs from the one or more potential transliterations, wherein each potential transliteration pair includes a first anchor text in a first writing system and a second anchor text in a second writing system, the second anchor text and the first anchor text linking to a same resource;

for each potential transliteration pair,

classifying the potential transliteration pair as being a transliteration pair or not being a transliteration pair;

aligning the first anchor text with the second anchor text; and

extracting one or more edits from the alignment; and

generating a classification model, based on the one or more edits and a subset of the classified potential transliteration pairs.

12. The method of claim 11 , further comprising identifying transliteration pairs from the potential transliteration pairs, using the classification model.

13. The method of claim 11 , wherein the one or more potential transliteration pairs are identified from the one or more potential transliterations, based on likelihoods of the one or more potential transliterations co-occurring with text that identifies a same resource or location.

14. The method of claim 11 , wherein the classifying includes:

determining a first edit distance between the first anchor text and the second anchor text;

comparing the first edit distance to a first threshold value; and

classifying the potential transliteration pair as being a transliteration pair or not being a transliteration pair, based on the comparison;

wherein the aligning is based on minimizing the first edit distance.

15. The method of claim 14 , where the first threshold value is the lesser of a length of the first anchor text and a length of the second anchor text.

16. The method of claim 15 , further comprising:

reclassifying the potential transliteration pairs based on refinement rules and the one or more edits in the first writing system.

17. The method of claim 16 , further comprising:

reclassifying the potential transliteration pairs classified as being a transliteration pair based on the refinement rules and the one or more edits in the second writing system.

18. The method of claim 11 , further comprising:

sorting the classified potential transliteration pairs based on likelihoods of the classified potential transliteration pairs being transliteration pairs.

19. The method of claim 18 , where the subset includes potential transliteration pairs classified as being a transliteration pair and potential transliteration pairs classified as not being a transliteration pair.

20. The method of claim 11 , where the classification model uses a support vector machine (SVM).

21. A computer program product, encoded on a computer readable storage medium, operable to cause one or more processors to perform operations comprising:

receiving a plurality of resources, the plurality of resources including a plurality of anchor text, wherein anchor found in a specific resource of the plurality of resources links to a different resource identified by a uniform resource locator of the different resource;

determining one or more potential transliterations from the plurality of anchor text;

identifying one or more potential transliteration pairs from the one or more potential transliterations, wherein each potential transliteration pair includes a first anchor text in a first writing system and a second anchor text in a second writing system, the second anchor text and the first anchor text linking to a same resource;

for each potential transliteration pair,

classifying the potential transliteration pair as being a transliteration pair or not being a transliteration pair;

aligning the first anchor text with the second anchor text; and

extracting one or more edits from the alignment; and

generating a classification model, based on the one or more edits and a subset of the classified potential transliteration pairs.

22. A system comprising:

a machine readable storage device including a program product; and

one or more computers operable to execute the program product and perform operations comprising:

receiving a plurality of resources, the plurality of resources including a plurality of anchor text, wherein anchor text found in a specific resource of the plurality of resources links to a different resource identified by a uniform resource locator of the different resource;

determining one or more potential transliterations from the plurality of anchor text;

identifying one or more potential transliteration pairs from the one or more potential transliterations, wherein each potential transliteration pair includes a first anchor text in a first writing system and a second anchor text in a second writing system, the second anchor text and the first anchor text linking to a same resource;

for each potential transliteration pair,

classifying the potential transliteration pair as being a transliteration pair or not being a transliteration pair;

aligning the first anchor text with the second anchor text; and

extracting one or more edits from the alignment; and

generating a classification model, based on the one or more edits and a subset of the classified potential transliteration pairs.