VhA`ddlmZmZddlmZmZmZmZddlm Z m Z ddl m Z ddl mZej fdeedeedeegefd eeeeefeeeefffd Zdeedeed eeeeefeeeefeeeefffd Zd eeeefd eeeefd eeeeefeeeefeeeefffdZefd eeeefd eeeefde d eeeeefeeeefeeeefffdZefdeedeede d eeeeefeeeefeeeefffdZe fd eeeefd eeeefde d eeeeefeeeefeeeefffdZe fdeedeede d eeeeefeeeefeeeefffdZee fd eeeefd eeeefde de d eeeeefeeeefeeeefff dZee fdeedeede de d eeeeefeeeefeeeefff dZdeeeefd efdZej ee dddfdeedeedeegefde de dededed efdZej ee dddfd eeedeedeegefde de dededed efd!Zy")#)chainproduct)CallableIterableListTuple)WordNetCorpusReaderwordnet)StemmerI) PorterStemmer hypothesis reference preprocessreturnct|trtd|t|trtd|tt t ||}tt t ||}||fS)a7 Takes in pre-tokenized inputs for hypothesis and reference and returns enumerated word lists for each of them :param hypothesis: pre-tokenized hypothesis :param reference: pre-tokenized reference :preprocess: preprocessing method (default str.lower) :return: enumerated words list z?"hypothesis" expects pre-tokenized hypothesis (Iterable[str]): z="reference" expects pre-tokenized reference (Iterable[str]): ) isinstancestr TypeErrorlist enumeratemap)r rrenum_hypothesis_listenum_reference_lists K/home/dcms/DCMS/lib/python3.12/site-packages/nltk/translate/meteor_score.py_generate_enumsrs*c"Mj\ Z  )S!KI; W   #j**E FGyZ)CDE !4 44c8t||\}}t||S)a matches exact words in hypothesis and reference and returns a word mapping based on the enumerated word id between hypothesis and reference :param hypothesis: pre-tokenized hypothesis :param reference: pre-tokenized reference :return: enumerated matched tuples, enumerated unmatched hypothesis tuples, enumerated unmatched reference tuples )r _match_enums)r rrrs r exact_matchr/s'1@ I0V-- ,.A BBrrrc4g}tt|dddD]u}tt|dddD]V}||d||dk(s|j||d||df|j||j|uw|||fS)a matches exact words in hypothesis and reference and returns a word mapping between enum_hypothesis_list and enum_reference_list based on the enumerated word id. :param enum_hypothesis_list: enumerated hypothesis list :param enum_reference_list: enumerated reference list :return: enumerated matched tuples, enumerated unmatched hypothesis tuples, enumerated unmatched reference tuples Nr)rangelenappendpop)rr word_matchijs rrr@sJ 3+, -dd 3s./026 A#A&q)-@-CA-FF!!)!,Q/1DQ1G1JK%((+#''*  +-@ @@rstemmerc|Dcgc]}|d|j|df}}|Dcgc]}|d|j|df}}t||Scc}wcc}w)af Stems each word and matches them in hypothesis and reference and returns a word mapping between enum_hypothesis_list and enum_reference_list based on the enumerated word id. The function also returns a enumerated list of unmatched words for hypothesis and reference. :param enum_hypothesis_list: enumerated hypothesis list :param enum_reference_list: enumerated reference list :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :return: enumerated matched tuples, enumerated unmatched hypothesis tuples, enumerated unmatched reference tuples rr")stemr)rrr* word_pairstemmed_enum_hypothesis_liststemmed_enum_reference_lists r_enum_stem_matchr0[s$EY$7@1w||IaL12$ $ EX#7@1w||IaL12## 46Q RR$#s A Ac<t||\}}t|||S)a Stems each word and matches them in hypothesis and reference and returns a word mapping between hypothesis and reference :param hypothesis: pre-tokenized hypothesis :param reference: pre-tokenized reference :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :return: enumerated matched tuples, enumerated unmatched hypothesis tuples, enumerated unmatched reference tuples r*)rr0)r rr*rrs r stem_matchr3ws)1@ I0V-- 02Ew WWrr c g}tt|dddD]}ttjd|j ||dDj ||dh}tt|dddD]O}||d|vs|j||d||df|j||j||||fS)a^ Matches each word in reference to a word in hypothesis if any synonym of a hypothesis word is the exact match to the reference word. :param enum_hypothesis_list: enumerated hypothesis list :param enum_reference_list: enumerated reference list :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) Nr!c3JK|]}d|jDyw)c3K|]6}|jjddkr|j8yw)_rN)namefind).0lemmas r z3_enum_wordnetsyn_match...s7zz|((-1JJLs<>N)lemmas)r:synsets rr<z)_enum_wordnetsyn_match..s-   !' s!#r"r) r#r$setr from_iterablesynsetsunionr%r&)rrr r'r(hypothesis_synsr)s r_enum_wordnetsyn_matchrDsJ 3+, -dd 3     .goo.B1.Ea.HI      %%a(+, - s./026 A"1%a(O;!!)!,Q/1DQ1G1JK%((+#''* & +-@ @@rc<t||\}}t|||S)ae Matches each word in reference to a word in hypothesis if any synonym of a hypothesis word is the exact match to the reference word. :param hypothesis: pre-tokenized hypothesis :param reference: pre-tokenized reference :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) :return: list of mapped tuples r )rrD)r rr rrs rwordnetsyn_matchrGs-1@ I0V-- !17 rct||\}}}t|||\}}}t|||\}}}t||z|zd||fS)a Aligns/matches words in the hypothesis to reference by sequentially applying exact match, stemmed match and wordnet based synonym match. in case there are multiple matches the match which has the least number of crossing is chosen. Takes enumerated list as input instead of string input :param enum_hypothesis_list: enumerated hypothesis list :param enum_reference_list: enumerated reference list :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) :return: sorted list of matched tuples, unmatched hypothesis list, unmatched reference list r2rFc |dS)Nr)wordpairs rz#_enum_align_words..s XVW[r)key)rr0rDsorted)rrr*r exact_matches stem_matches wns_matchess r_enum_align_wordsrRs(@L1@T17>:K%':  L (; 6t||\}}t||||S)aY Aligns/matches words in the hypothesis to reference by sequentially applying exact match, stemmed match and wordnet based synonym match. In case there are multiple matches the match which has the least number of crossing is chosen. :param hypothesis: pre-tokenized hypothesis :param reference: pre-tokenized reference :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) :return: sorted list of matched tuples, unmatched hypothesis list, unmatched reference list r*r )rrR)r rr*r rrs r align_wordsrUs/$1@ I0V-- 17G rmatchescd}d}|t|dz krP||dzd||ddzk(r||dzd||ddzk(r|dz }E|dz }|dz }|t|dz krP|S)ab Counts the fewest possible number of chunks such that matched unigrams of each chunk are adjacent to each other. This is used to calculate the fragmentation part of the metric. :param matches: list containing a mapping of matched words (output of align_words) :return: Number of chunks a sentence is divided into post alignment rr")r$)rVr(chunkss r _count_chunksrYs A F c'lQ  AEN1 A!2 2 AEN1 A!2 2 FA  Q!  c'lQ  Mrg?g@g?alphabetagammac^t|||\}} t|} t| } t|| ||\} } } t| } t|| z }t|| z }||z||zd|z |zzz }tt | }||z }|||zz}d|z |zS#t $rYywxYw)a Calculates METEOR score for single hypothesis and reference as per "Meteor: An Automatic Metric for MT Evaluation with HighLevels of Correlation with Human Judgments" by Alon Lavie and Abhaya Agarwal, in Proceedings of ACL. https://www.cs.cmu.edu/~alavie/METEOR/pdf/Lavie-Agarwal-2007-METEOR.pdf >>> hypothesis1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which', 'ensures', 'that', 'the', 'military', 'always', 'obeys', 'the', 'commands', 'of', 'the', 'party'] >>> reference1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'that', 'ensures', 'that', 'the', 'military', 'will', 'forever', 'heed', 'Party', 'commands'] >>> round(single_meteor_score(reference1, hypothesis1),4) 0.6944 If there is no words match during the alignment the method returns the score as 0. We can safely return a zero instead of raising a division by zero error as no match usually implies a bad translation. >>> round(single_meteor_score(['this', 'is', 'a', 'cat'], ['non', 'matching', 'hypothesis']),4) 0.0 :param reference: pre-tokenized reference :param hypothesis: pre-tokenized hypothesis :param preprocess: preprocessing function (default str.lower) :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) :param alpha: parameter for controlling relative weights of precision and recall. :param beta: parameter for controlling shape of penalty as a function of as a function of fragmentation. :param gamma: relative weight assigned to fragmentation penalty. :return: The sentence-level METEOR score. )rrTr"g)rr$rRfloatrYZeroDivisionError)rr rr*r rZr[r\enum_hypothesisenum_referencetranslation_lengthreference_lengthrVr7 matches_count precisionrecallfmean chunk_count frag_fracpenaltys rsingle_meteor_scorerksX'6I*'#O^_->*%'MGQLM-(+== }%(88V# (9QY&>> hypothesis1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which', 'ensures', 'that', 'the', 'military', 'always', 'obeys', 'the', 'commands', 'of', 'the', 'party'] >>> hypothesis2 = ['It', 'is', 'to', 'insure', 'the', 'troops', 'forever', 'hearing', 'the', 'activity', 'guidebook', 'that', 'party', 'direct'] >>> reference1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'that', 'ensures', 'that', 'the', 'military', 'will', 'forever', 'heed', 'Party', 'commands'] >>> reference2 = ['It', 'is', 'the', 'guiding', 'principle', 'which', 'guarantees', 'the', 'military', 'forces', 'always', 'being', 'under', 'the', 'command', 'of', 'the', 'Party'] >>> reference3 = ['It', 'is', 'the', 'practical', 'guide', 'for', 'the', 'army', 'always', 'to', 'heed', 'the', 'directions', 'of', 'the', 'party'] >>> round(meteor_score([reference1, reference2, reference3], hypothesis1),4) 0.6944 If there is no words match during the alignment the method returns the score as 0. We can safely return a zero instead of raising a division by zero error as no match usually implies a bad translation. >>> round(meteor_score([['this', 'is', 'a', 'cat']], ['non', 'matching', 'hypothesis']),4) 0.0 :param references: pre-tokenized reference sentences :param hypothesis: a pre-tokenized hypothesis sentence :param preprocess: preprocessing function (default str.lower) :param stemmer: nltk.stem.api.StemmerI object (default PorterStemmer()) :param wordnet: a wordnet corpus reader object (default nltk.corpus.wordnet) :param alpha: parameter for controlling relative weights of precision and recall. :param beta: parameter for controlling shape of penalty as a function of as a function of fragmentation. :param gamma: relative weight assigned to fragmentation penalty. :return: The sentence-level METEOR score. c 3DK|]}t|yw))rr*r rZr[r\N)rk) r:rrZr[r\r rr*r s rr<zmeteor_score..s=     !  s )max)rlr rr*r rZr[r\s ```````r meteor_scorerp[s'd   $   rN) itertoolsrrtypingrrrr nltk.corpusr r nltk.stem.apir nltk.stem.porterr rlowerintrrrr0r3rDrGrRrUrYr^rkrprJrrrxsc%224"* (+yy5 5}5#$5 4c3h $uS#X"7 78 5:C C*23-C 4c3h $uS#X"7eCHo9N NOC"AuS#X/AeCHo.A 4c3h $uS#X"7eCHo9N NOA<&SuS#X/SeCHo.SS 4c3h $uS#X"7eCHo9N NO S>&X X}XX 4c3h $uS#X"7eCHo9N NO X,$+"AuS#X/"AeCHo."A!"A 4c3h $uS#X"7eCHo9N NO "AP$+ }! 4c3h $uS#X"7eCHo9N NO .&#* &uS#X/&eCHo.&&! &  4c3h $uS#X"7eCHo9N NO &X&#*  }!   4c3h $uS#X"7eCHo9N NO 04c3h0S2(+yy%#*>!}>! >!#$>! >! ! >!  >! >! >! >!H(+yy%#*>#'> >#$> > ! >  > > > >r