Open Named Entity Modeling from Embedding Distribution
This addresses the problem of closed and labor-intensive named entity definitions in NLP, offering an open, multilingual approach that could benefit resource-poor languages.
The authors discovered that named entities cluster together in word embedding spaces regardless of type or language, enabling them to model all named entities using a geometric structure called the named entity hypersphere. This model provides an open definition for multilingual named entities, potentially enhancing state-of-the-art named entity recognition systems.
In this paper, we report our discovery on named entity distribution in a general word embedding space, which helps an open definition on multilingual named entity definition rather than previous closed and constraint definition on named entities through a named entity dictionary, which is usually derived from human labor and replies on schedule update. Our initial visualization of monolingual word embeddings indicates named entities tend to gather together despite of named entity types and language difference, which enable us to model all named entities using a specific geometric structure inside embedding space, namely, the named entity hypersphere. For monolingual cases, the proposed named entity model gives an open description of diverse named entity types and different languages. For cross-lingual cases, mapping the proposed named entity model provides a novel way to build a named entity dataset for resource-poor languages. At last, the proposed named entity model may be shown as a handy clue to enhance state-of-the-art named entity recognition systems generally.