Language-Independent Named Entity Analysis Using Parallel Projection and Rule-Based Disambiguation

Abstract

The 2017 shared task at the BaltoSlavic NLP workshop requires identifying coarse-grained named entities in seven languages, identifying each entity's base form, and clustering name mentions across the multilingual set of documents. The fact that no training data is provided to systems for building supervised classifiers further adds to the complexity. To complete the task we first use publicly available parallel texts to project named entity recognition capability from English to each evaluation language. We ignore entirely the subtask of identifying non-inflected forms of names. Finally, we create cross-document entity identifiers by clustering named mentions using a procedure-based approach.

The First Cross-Lingual Challenge on Recognition, Normalization, and Matching of Named Entities in Slavic Languages

“…The team (code "jhu") attempted all languages available in the Challenge. More details can be found in (Mayfield et al, 2017).…”

Finding next of kin: Cross-lingual embedding spaces for related languages

“…The baseline system (Piskorski et al . 2017) was based on large gazetteers developed by the JRC, while the only other submission covering all Slavonic languages from JHU (Mayfield, McNamee, and Costello 2017) was based on projection of NER labels via word-aligned parallel corpora, see Table 9, as well as a brief explanation of the projection approach in Section 4.…”

“…One set of approaches uses parallel corpora for projecting automatic annotations in one language to others, for example, for POS tagging (Das and Petrov 2011), parsing (Täckström, McDonald, and Nivre 2013; Tiedemann 2014) and NER (Mayfield et al . 2017). In the projection approaches, the donor part of a parallel corpus is annotated with an existing tool.…”

BSNLP2019 Shared Task Submission: Multisource Neural NER Transfer

“…The task was somewhat different from the 2019 task in that training data was not provided to participants. Approaches submitted to this task included a model based on parallel projection (Mayfield et al, 2017) and a model with language-specific features trained on found data (Marcińczuk et al, 2017). There has also been follow-up work on this dataset using cross-lingual embeddings (Sharoff, 2018).…”