Stata/Python code for fuzzy matching of latin script location names in Hindi.

This project is being maintained here: https://github.com/devdatalab/masala-merge

Python 3.2 (may work with other versions)

$tmp and $MASALA_PATH must be defined. $tmp is a folder for storage of temporary files. $MASALA_PATH is the path containing lev.py, included in this package.

fix_spelling will magically correct spelling errors in a list of words, given a master list of correct words. For example, suppose you have a dataset with district names, you have a master list of district names (with state identifiers), and you want to modify your current district names to match the master key. The following command will "fix" your misspelled district names:

fix_spelling district_name, src(master_district_key.dta) group(state_id) replace

state_id is a group variable -- districts in state 1 in the open file will only be fixed based on districts in state 1 in the key file. With this format, district_name and state_id both need to appear in both files. If the variables have different names in the different datasets, use targetfield() and targetgroup() to specify the field and group variables in the using data.

Additional options:

• gen(varname) can be used instead of replace
• targetfield() and targetgroup() can be used if the group or merge variable have different names in the master dataset.
• If keepall is specified, your dataset will add rows from the target list that didn't match anything in your data

Example:

. fix_spelling pc01_district_name, src($keys/pc01_district_key) replace group(pc01_state_name) 

[...]

+--------------------------------------------------------------------------------------
| Spelling fixes and levenshtein distances:
+--------------------------------------------------------------------------------------

      +-------------------------------------------+
      | pc01_distri~e         __000000   __0000~t |
      |-------------------------------------------|
  80. |   karimanagar       karimnagar         .2 |
 155. |  mahabubnagar      mahbubnagar         .2 |
 422. |         buxor            buxar        .45 |
 462. |     jahanabad        jehanabad        .45 |
 480. |     khagari a         khagaria        .01 |
      |-------------------------------------------|
 544. |        purnea           purnia        .45 |
 624. |     ahmedabad        ahmadabad        .45 |
 700. |   banaskantha     banas kantha        .01 |
 757. |         dahod            dohad         .8 |
 888. |    panchmahal     panch mahals       1.01 |
      |-------------------------------------------|
 932. |   sabarkantha     sabar kantha        .01 |
 991. |       vadodra         vadodara         .2 |
1490. |         angul           anugul          1 |
1546. |         boudh            baudh        .45 |
1569. |       deogarh         debagarh        1.2 |
      |-------------------------------------------|
1609. | jagatsinghpur   jagatsinghapur         .2 |
1617. |        jajpur          jajapur         .2 |
1674. |        khurda          khordha        .65 |
1722. |   nabarangpur     nabarangapur         .2 |
1922. |    puducherry      pondicherry       1.35 |
      +-------------------------------------------+

masala_merge is the underlying program used by fix_spelling and has more customizable behavior.

Given datasets that already align on state, district and subdistrict identifiers, the following command will run a fuzzy merge on village names:

masala_merge state_id district_id subdistrict_id using match_data.dta, S1(village_name) OUTfile($tmp/foo)

Only one field can be fuzzy matched at a time.

The outfile is a temporary file that will be generated with the full set of matches. masala_merge will create a dataset that automatically picks the matches that it thinks are best and rejects the matches that are too ambiguous.

For example, if a match with modified levenshtein distance 0.4 will be rejected if there is an another possible match with

Some Additional options:

fuzziness(real): Default is 1. Higher numbers generate merges with more
           tolerance for differences. But even 0 will allow some
	   fuzziness.

quietly: Suppress output

keepusing(string): Works like keepusing() in standard merge command

sortwords: Will sort words before running merge

Masala_merge runs a Levenshtein "edit distance" algorithm that counts the number of insertions, deletions and character changes that are required to get from one string to another. We have modified the Levenshtein algorithm to use smaller penalties for very common alternate spellings in Hindi. For instance "laxmi" and "lakshmi" have a Levenshtein distance of 3, but a Masala Levenshtein distance of only 0.4.

The low cost character changes are described in a list in lev.py. If you would like to modify this for other languages with other common spelling inconsistencies, you only need to modify these lists and the costs associated with each kind of change.