For a news content distribution platform like Dailyhunt, Named Entity Recognition is a pivotal task for building better user recommendation and notification algorithms. Apart from identifying names, locations, organisations from the news for 13+ Indian languages and use them in algorithms, we also need to identify n-grams which do not necessarily fit in the definition of Named-Entity, yet they are important. For example, "me too movement", "beef ban", "alwar mob lynching". In this exercise, given an English language text, we are trying to detect case-less n-grams which convey important information and can be used as topics and/or hashtags for a news. Model is built using Wikipedia titles data, private English news corpus, BERT-Multilingual pre-trained model, Bi-GRU and CRF architecture. It shows promising results when compared with industry best Flair, Spacy and Stanford-caseless-NER in terms of F1 and especially Recall.
-
Sample Train Data File
=> FILE : sample_train_data.tsv.gz
-
Cased NERs output
=> FILE : sample_english_news.out.limited
-
Sample Model output
=> FILE : english_sample_news.tsv
-
All possible NERs for benchmark files
=> all_possible_ner_set.txt
-
Models Output Counts
=> FILE : benchmark_file1.tsv, benchmark_file11.tsv
=> TAB Separated
=> COLUMNS : text, actual_ner_count, curr_model_match, curr_model_count_total, curr_stanford_match, curr_stanford_count_total, curr_spacy_match, curr_spacy_count_total, curr_flair_match, curr_flair_count_total
-
Reference wikipedia titles for given validation set
=> all_possible_wiki_title_matches.txt
-
Benchmark outputs for all models
=> benchmark_file_222.tsv
=> TAB separated
=> COLUMNS : text, actual_ner_list, model_ner_list, stanford_ner_list, spacy_ner_list, flair_ner_list