1. December 2007NLP: Conflation Algorithms1 Natural Language Processing Conflation Algorithms

2. December 2007NLP: Conflation Algorithms2 Acknowledgements John Repici (2002) http://www.creativyst.com/Doc/Articles/SoundEx1/Sound Ex1.htm Porter, M.F., 1980, An algorithm for suffix stripping, reprinted in Sparck Jones, Karen, and Peter Willet, 1997, Readings in Information Retrieval, San Francisco: Morgan Kaufmann, ISBN 1-55860-454-4. [Vince has a copy of this] Jurafsky & Martin appendix B pp 833-836.

3. December 2007NLP: Conflation Algorithms3 Conflation COMPUTES COMPUTE COMPUTATION COMPUTABILITY COMPUTING COMPUTER COMPUT

4. December 2007NLP: Conflation Algorithms4 Word Conflation Algorithms Morphological analysis versus conflation Notion of word class is application dependent –Genealogy: Phonetic similarity –Information Retrieval: Semantic similarity Soundex Porter

5. December 2007NLP: Conflation Algorithms5 Problems with Names Names can be misspelt: Rossner Same name can be spelt in different ways Kirkop; Chircop Same name appears differently in different cultures: Tchaikovsky; Chaicowski To solve this problem, we need phonetically oriented algorithms which can find similar sounding terms and names. Just such a family of algorithms exist and are called SoundExes, after the first patented version.

6. December 2007NLP: Conflation Algorithms6 The Soundex Algorithm A Soundex algorithm takes a word as input and produces a character string which identifies a set of words that are (roughly) phonetically alike. It is very handy for searching large databases Originally developed 1918 by Margaret K. Odell and Robert C. Russell of the US Bureau of Archives, to simplify census-taking.

7. December 2007NLP: Conflation Algorithms7 Soundex Algorithm 1 The Soundex Algorithm uses the following steps to encode a word: 1.The first character of the word is retained as the first character of the Soundex code. 2.The following letters are discarded: a,e,i,o,u,h,w, and y. 3.Remaining consonants are given a code number. 4.If consonants having the same code number appear consecutively, the number will only be coded once. (e.g. "B233" becomes "B23")

8. December 2007NLP: Conflation Algorithms8 Code Numbers b, p, f, and v1 c, s, k, g, j, q, x, z2 d, t3 l4 m,n5 r6

9. December 2007NLP: Conflation Algorithms9 Soundex Algorithm: Example The Soundex Algorithm uses the following steps to encode a word: [ROSNER] 1.The first character of the word is retained as the first character of the Soundex code [R] 2.The following letters are discarded: a,e,i,o,u,h,w, and y. [RSNR] 3.Remaining consonants are given a code number. [R256] 4.If consonants having the same code number appear consecutively, the number will only be coded once. (e.g. "B233" becomes "B23") [R256]

10. December 2007NLP: Conflation Algorithms10 Soundex Algorithm 2 –The resulting code is modified so that it becomes exactly four characters long: If it is less than 4 characters, zeroes are added to the end (e.g. "B2" becomes "B200") –If it is more than 4 characters, the code is truncated (e.g. "B2435" becomes "B243")

11. December 2007NLP: Conflation Algorithms11 Uses for the Soundex Code Airline reservations - The soundex code for a passenger's surname is often recorded to avoid confusion when trying to pronounce it. U.S. Census - As is noted above, the U.S. Census Department was a frequent user of the Soundex algorithm while trying to compile a listing of families around the turn of the century. Genealogy - In genealogy, the Soundex code is most often used to avoid obstacles when dealing with names that might have alternate spellings.

12. December 2007NLP: Conflation Algorithms12 Improvements Preprocessing before applying the basic algorithm, e.g. identification of –DG with G –GH with H –GN with N (not 'ng') –KN with N –PH with F Question: where to stop? Question: how to evaluate?

13. December 2007NLP: Conflation Algorithms13 IR Applications Information Retrieval: Query → → Relevant Documents “Bag of Terms” document model What is a single term?

14. December 2007NLP: Conflation Algorithms14 Why Stemming is Necessary Frequently we get collections of words of the following kind in the same document compute, computer, computing, computation, computability …. Performance of IR system will be improved if all of these terms are conflated. –Less terms to worry about –More accurate statistics

15. December 2007NLP: Conflation Algorithms15 Issues Is a dictionary available? –Stems –Affixes Motivation: linguistic credibility or engineering performance? When to remove a affix versus when to leave it alone Porter (1980): W 1 and W 2 should be conflated if there appears to be no difference between the statements "this document is about W 1 /W 2 " relate/relativity vs. radioactive/radioactivity

16. December 2007NLP: Conflation Algorithms16 Consonants and Vowels A consonant is a letter other than a,e,i,o,u and other than y preceded by a consonant: sky, toy If a letter is not a consonant it is a vowel. A sequence of consonants (cc..c) or vowels (vv..v) will be represented by C or V respectively. For example the word troubles maps to C V C V C Any word or part of a word, therefore has one of the following forms: (CV) n ….C (CV) n ….V (VC) n ….C (VC) n ….V

17. December 2007NLP: Conflation Algorithms17 Measure All the above patterns can be replaced by the following regular expression (C) (VC) m (V) m is called the measure of any word or word part. m=0: tr, ee, tree, y, by m=1: trouble, oats, trees, ivy m=2: troubles; private

18. December 2007NLP: Conflation Algorithms18 Rules Rules for removing a suffix are given in the form (condition) S1 → S2 i.e. if a word ends with suffix S1, and the stem before S1 satisfies the condition, then it is replaced with S2. Example (m > 1) EMENT → Example: enlargement → enlarg

19. December 2007NLP: Conflation Algorithms19 Conditions *S - stem ends with s *Z - stem ends with z *T – stem ends with t *v* - stem contains a vowel *d - stem ends with a double consonant *o - stem ends cvc, where second c is not w, x or y e.g. –wil, -hop In conditions, Boolean operators are possible e.g. (m>1 and (*S or *T)) Sets of rules applied in 7 steps. Within each step, rule matching longest suffix applies.

20. December 2007NLP: Conflation Algorithms20 Organisation Step 1 Plurals and Third Person Singular Verbs Step 2 Verbal Past Tense and Progressive Step 3: Y to I Noun Inflections Steps 4 and 5 Derivational Morphology Multiple Suffixes visualisation → visualise Steps 6 Derivational Morphology Single Suffixes Step 7 Cleanup -s -ed, -ingfly/flies

21. December 2007NLP: Conflation Algorithms21 Step 1:Plural Nouns and 3 rd Person Singular Verbs conditionrewriteexample SSES → SScaresses → caress IES → Iponies → poni SS → SScaress → caress S →cats → cat

22. December 2007NLP: Conflation Algorithms22 Step 2a Verbal Past Tense and Progressive Forms conditionrewriteexample (m>0)EED → EEfeed → feed agreed → agree (*v*)ED → εplastered → plaster bled → bled (*v*)ING → εkilling → kill sing → sing

23. December 2007NLP: Conflation Algorithms23 Step 2b: Cleanup If 2 nd or 3 rd of last step succeeds conditionrewriteexample AT → ATEgenerat → generate BL → BLEtroubl → trouble IZ → IZEcapsiz → capsize *d and not (*L or *S or *Z) → single letter hopp → hop hiss → hiss

24. December 2007NLP: Conflation Algorithms24 Step 3: Y to I (*v*)Y → Ihappy → happi cry → cry

25. December 2007NLP: Conflation Algorithms25 STEP 4: Derivational Morphology 1 – Multiple Suffixes (excerpt) ConditionRewriteExample (m > 0)ATIONAL → ATErelational → relate (m > 0)TIONAL → TIONconditional → condition (m > 0)ENCI → ENCEvalenci → valence (m > 0)ABLI → ABLEcomfortabli → comfortable (m > 0)OUSLI → OUSanalagously → analagous (m > 0)IZATION → IZEdigitizer → digitize (m > 0)ATION → ATEgeneration → generate (m > 0)ATOR → ATEoperator → operate (m > 0)ALISM → ALformalism → formal (m > 0)IVENESS → IVEpensiveness → pensive (m > 0)FULNESS → FULhopefulness → hopeful (m > 0)OUSNESS → OUScallousness → callous (m > 0)ALITI → ALformality → formal (m > 0)BILITI → BLEpossibility → possible

26. December 200726 Step 6: Derivational Morphology III: Single Suffixes ConditionRewriteExample (m > 1)AL → εrevival → reviv (m > 1)ANCE → εallowance → allow (m > 1)ENCE → εinference → infer (m > 1)ER → εairliner → airlin (m > 1)IC → εCoptic → Copt (m > 1)ABLE → εlaughable → laugh (m > 1)ANT → εirritant → irrit (m > 1)EMENT → εreplacement → replac (m > 1)MENT → εadjustment → adjust (m > 1)ENT → εdependent → depend (m > 0) (*S or *T)ION → εadoption → adopt (m > 1)OU → εcallousness → callous (m > 1)ISM → εformalism→ formal (m > 1)ATE → εactivate → activ ITI → ε

27. December 2007NLP: Conflation Algorithms27 Porter Example INPUT in the first focus area, integrated projects shall help develop, principally, common open platforms for software and services supporting a distributed information and decision systems for risk and crisis management

28. December 2007NLP: Conflation Algorithms28 Porter Output Original WordStemmed Word first focusfocu area integratedintegr projectsproject help develop principallyprincip common open platformsplatform Original WordStemmed Word platformsplatform softwaresoftwar servicesservic supportingsupport distributeddistribut informationinform decisiondecis systemssystem risk crisiscrisi managementmanag

29. December 2007NLP: Conflation Algorithms29 Summary Conflation serves different purposes Generally, motivation is to achieve an engineering goal rather than linguistic fidelity. This can cause errors in the bag of words model. Soundex and Porter very well established and easily available.