A Deep Learning Architecture for Semantic Address Matching

Codes in this repository is for the paper Yue Lin, Mengjun Kang, Yuyang Wu, Qingyun Du & Tao Liu (accepted) A deep learning architecture for semantic address matching, International Journal of Geographical Information Science..

Codes are cited as Lin, Yue & Kang, Mengjun. (2019, October 8). yuelinnnnnnn/semantic_address_matching: Semantic address matching (Version v1.0). Zenodo. http://doi.org/10.5281/zenodo.3476673

Data are available at:

• Shenzhen address corpus (part): http://doi.org/10.5281/zenodo.3477632
• Labelled address dataset for semantic address matching: http://doi.org/10.5281/zenodo.3477006

Below is an overview of each file in this repository.

• geo_config.py Hyperparameter settings for the ESIM
• geo_data_prepare.py Tokenize the corpus and convert each address element into index
• geo_data_processor.py Process the labeled address dataset and divide it into training, development and test sets
• geo_ESIM.py Implementation of the enhanced sequential inference model (ESIM)
• geo_similarity.py Calculate statistical characteristics of the labeled address dataset
• geo_test.py Output predictive results of the ESIM on the test set
• geo_token.py Tokenize with the Jieba library
• geo_train.py Train the ESIM and evaluate its accuracy on the development set
• geo_word2vec.py Train word vectors of address elements
• other_CRF.py Tokenize using CRF [Comber and Arribas-Bel (2019)]
• other_crf_w2v.py Train word vectors of address elements (CRF tokenizer)
• other_string.py String similarity-based address matching methods: measure the string relevance
• other_w2v_cls.py Use word2vec embeddings directly for classification: calculat cosine similarity