Lista usług¶

Narzędzia podstawowe dla języka polskiego:¶

W podpunktach podano podstawowy potok przetwarzania w notacji LPMN

• Morpho
  • any2txt|maca({"morfeusz2":true})
• Tager
  • WCRFT2: any2txt|wcrft2({"guesser":false, "morfeusz2":true})
  • Morphodita: any2txt|morphoDita({"guesser":false, "allforms":false, "model":"XXI"})
• Chunker
  • any2txt|wcrft2|iobber
• Ner
  • any2txt|wcrft2|liner2({"model":"names"})
• Serel
  • any2txt|wcrft2|liner2({"model":"5nam"})|serel
• Spatial
  • any2txt|wcrft2({"morfeusz2":false})|liner2({"model":"names","output":"tei:gz"})|spatial
• Spejd
  • any2txt|wcrft2({"morfeusz2":false})|iobber|liner2({"model":"n82","output":"tei:gz"})|spejd
• Parser
  • any2txt|wcrft2|dependpar|out("conll_")|conll2svg|out("svg_")
• WSD
  • any2txt|wcrft2({"morfeusz2":false})|wsd({"use_mwe":false})
• Ner DE
  • any2txt|spacy({"annotate_entities":true,"lang":"de"})
• Ner EN
  • any2txt|spacy({"annotate_entities":true})
• Ner EN NLTK
  • any2txt|nltk({"annotate_entities":true})
• Parser DE
  • any2txt|spacy({"method":"parser","lang":"de"})|out("conll_")|conll2svg|out("svg_")
• Parser EN
  • any2txt|spacy({"method":"parser"})|out("conll_")|conll2svg|out("svg_")
• Tager DE
  • any2txt|spacy({"lang":"de"})
• Tager EN
  • any2txt|spacy
• Tager EN NLTK
  • any2txt|spacy
• Tager ML
  • any2txt|tagger({"lang":"polish"})
• Inkluz
  • any2txt|inkluz
• Respa
  • any2txt|wcrft2|respa
• Sentyment
  • any2txt|wcrft2({"morfeusz2":true})|wsd|sentiment|out("senti")|sentimerge({"split_paragraphs":"False"})
• Summarize
  • any2txt|wcrft2({"morfeusz2":false})|liner2({"model":"names"})|summarize
• TermoPL
  • any2txt|wcrft2({"morfeusz2":false})|termopl
• TF-IDF
  • any2txt|wcrft2|tfidf
• WebSty
  • any2txt|wcrft2|fextor2({"features":"base interp_signs bigrams","base_modification":"startlist","orth_modification":"startlist","lang":"pl","filters":{"base":[{"type":"lemma_stoplist","args":{"stoplist":"@resources/fextor/nkjp360-meaningless-no-prep-freq-above-3500.txt"}}]}})|dir|out("output_fextor")|featfilt({"similarity":"cosine","weighting":"all:tf","filter":"min_tf-1 min_df-1"})|cluto({"no_clusters":2,"analysis_type":"plottree"})
• WebStyML
  • any2txt|div(20000)|tagger({"lang":"polish"})|fextor2({"features":"base interp_signs bigrams","base_modification":"startlist","orth_modification":"startlist","lang":"ud","filters":{"base":[{"type":"lemma_stoplist","args":{"stoplist":"@resources/fextor/ml/polish_base_startlist.txt"}}]}})|dir|out("output_fextor")|featfilt({"similarity":"cosine","weighting":"all:tf","filter":"min_tf-1 min_df-1"})|cluto({"no_clusters":2,"analysis_type":"plottree"})
• Topic
  • converter({"type":"topicmodel"})|dir|topicmodel({"min_df":2,"max_df":0.7,"no_topics":20,"method":"lda_gensim"})
• Lem
  • any2txt|wcrft2({"morfeusz2":false})|converter({"type":"ccl2base"})|dir|shimext({"ext":".txt"})|makezip
• MeWeX
  • any2txt|wcrft2({"morfeusz2":true})|mewex({"mewex_options":{"ranker_func":"vector_association_measure","num_kbest":"500","wccl_rels":["agr_noun_adj","all_burk_noun","all_ger_qub","sth_adjgen","gndnoun_adj","all_noun_noun","all_noun_self","ppron3gen_noun","all_num_noun","adj_noun_adj","noun_prep_noun"]}})

Techniczne:¶

• dir - grupowanie danych do jednego katalogu
• makezip - pakowanie (zip) danych

Konwersje:¶

• any2txt
• converter

Grupowanie:¶

• cluto
• ward

Wydobywanie informacji¶

• fextor
• featfilt
• feature
• summarize
• respa
• tfidf

Język angielski, niemiecki¶

• spacy

Kontakt
Pytania i uwagi proszę kierować na maila: webserwisy ( at ) clarin-pl.eu