Using an Optimized Chinese Address Matching Method to Develop a Geocoding Service: A Case Study of Shenzhen, China

Measuring Accessibility to Healthcare Using Taxi Trajectories Data: A Case Study of Acute Myocardial Infarction Cases in Beijing

Several methods have been applied to measure healthcare accessibility, ie, the Euclidean distance, the network distance, and the transport time based on speed limits. However, these methods generally produce less accurate estimates than actual measurements. This research proposed a method to estimate historical healthcare accessibility more accurately by using taxi Global Positioning System (GPS) traces. The proposed method's advantages were evaluated vis a case study using acute myocardial infarction (AMI) cases in Beijing in 2008. Comparative analyses of the new measure and three conventionally used measures suggested that the median estimated transport time to the closest hospital with percutaneous coronary intervention (PCI) capability for AMI patients was 5.72 minutes by the taxi GPS trace-based measure, 2.42 minutes by the network distance-based measure, 2.28 minutes by the speed limit-based measure, 1.73 minutes by the Euclidean distance-based measure; and the estimated proportion of patients who lived within 5 minutes of a PCI-capable hospital was 38.17%, 89.20%, 92.52%, 95.05%, respectively. The three conventionally used measures underestimated the travel time cost and overestimated the percentage of patients with timely access to healthcare facilities. In addition, the new measure more accurately identifies the areas with low or high access to healthcare facilities. The taxi GPS trace-based accessibility measure provides a promising start for more accurately estimating accessibility to healthcare facilities, increasing the use of medical records in studying the effects of historical healthcare accessibility on health outcomes, and evaluating how accessibility to healthcare changes over time.

Automatic Identification of Addresses: A Systematic Literature Review

Address matching continues to play a central role at various levels, through geocoding and data integration from different sources, with a view to promote activities such as urban planning, location-based services, and the construction of databases like those used in census operations. However, the task of address matching continues to face several challenges, such as non-standard or incomplete address records or addresses written in more complex languages. In order to better understand how current limitations can be overcome, this paper conducted a systematic literature review focused on automated approaches to address matching and their evolution across time. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed, resulting in a final set of 41 papers published between 2002 and 2021, the great majority of which are after 2017, with Chinese authors leading the way. The main findings revealed a consistent move from more traditional approaches to deep learning methods based on semantics, encoder-decoder architectures, and attention mechanisms, as well as the very recent adoption of hybrid approaches making an increased use of spatial constraints and entities. The adoption of evolutionary-based approaches and privacy preserving methods stand as some of the research gaps to address in future studies.

Recognition Method of New Address Elements in Chinese Address Matching Based on Deep Learning

Location services based on address matching play an important role in people’s daily lives. However, with the rapid development of cities, new addresses are constantly emerging. Due to the untimely updating of word segmentation dictionaries and address databases, the accuracy of address segmentation and the certainty of address matching face severe challenges. Therefore, a new address element recognition method for address matching is proposed. The method first uses the bidirectional encoder representations from transformers (BERT) model to learn the contextual information and address model features. Second, the conditional random field (CRF) is used to model the constraint relationships among the tags. Finally, a new address element is recognized according to the tag, and the new address element is put into the word segmentation dictionary. The spatial information is assigned to it, and it is put into the address database. Different sequence tagging models and different vector representations of addresses are used for comparative evaluation. The experimental results show that the method introduced in this paper achieves the maximum generalization ability, its F1 score is 0.78, and the F1 score on the testing dataset also achieves a high value (0.95).

Bidirectional Gated Recurrent Unit Neural Network for Chinese Address Element Segmentation

Chinese address element segmentation is a basic and key step in geocoding technology, and the segmentation results directly affect the accuracy and certainty of geocoding. However, due to the lack of obvious word boundaries in Chinese text, the grammatical and semantic features of Chinese text are complicated. Coupled with the diversity and complexity in Chinese address expressions, the segmentation of Chinese address elements is a substantial challenge. Therefore, this paper proposes a method of Chinese address element segmentation based on a bidirectional gated recurrent unit (Bi-GRU) neural network. This method uses the Bi-GRU neural network to generate tag features based on Chinese word segmentation and then uses the Viterbi algorithm to perform tag inference to achieve the segmentation of Chinese address elements. The neural network model is trained and verified based on the point of interest (POI) address data and partial directory data from the Baidu map of Beijing. The results show that the method is superior to previous neural network models in terms of segmentation performance and efficiency.

Improving a Street-Based Geocoding Algorithm Using Machine Learning Techniques

Address matching is a crucial step in geocoding; however, this step forms a bottleneck for geocoding accuracy, as precise input is the biggest challenge for establishing perfect matches. Matches still have to be established despite the inevitability of incorrect address inputs such as misspellings, abbreviations, informal and non-standard names, slangs, or coded terms. Thus, this study suggests an address geocoding system using machine learning to enhance the address matching implemented on street-based addresses. Three different kinds of machine learning methods are tested to find the best method showing the highest accuracy. The performance of address matching using machine learning models is compared to multiple text similarity metrics, which are generally used for the word matching. It was proved that extreme gradient boosting with the optimal hyper-parameters was the best machine learning method with the highest accuracy in the address matching process, and the accuracy of extreme gradient boosting outperformed similarity metrics when using training data or input data. The address matching process using machine learning achieved high accuracy and can be applied to any geocoding systems to precisely convert addresses into geographic coordinates for various research and applications, including car navigation.

Spatial Context-Based Local Toponym Extraction and Chinese Textual Address Segmentation from Urban POI Data

Georeferencing by place names (known as toponyms) is the most common way of associating textual information with geographic locations. While computers use numeric coordinates (such as longitude-latitude pairs) to represent places, people generally refer to places via their toponyms. Query by toponym is an effective way to find information about a geographic area. However, segmenting and parsing textual addresses to extract local toponyms is a difficult task in the geocoding field, especially in China. In this paper, a local spatial context-based framework is proposed to extract local toponyms and segment Chinese textual addresses. We collect urban points of interest (POIs) as an input data source; in this dataset, the textual address and geospatial position coordinates correspond at a one-to-one basis and can be easily used to explore the spatial distribution of local toponyms. The proposed framework involves two steps: address element identification and local toponym extraction. The first step identifies as many address element candidates as possible from a continuous string of textual addresses for each urban POI. The second step focuses on merging neighboring candidate pairs into local toponyms. A series of experiments are conducted to determine the thresholds for local toponym extraction based on precision-recall curves. Finally, we evaluate our framework by comparing its performance with three well-known Chinese word segmentation models. The comparative experimental results demonstrate that our framework achieves a better performance than do other models.

Measuring Spatial Patterns of Health Care Facilities and Their Relationships with Hypertension Inpatients in a Network-Constrained Urban System

There is evidence of a strong correlation between inequality in health care access and disparities in chronic health conditions. Equal access to health care is an important indicator for overall population health, and the urban road network has a significant influence on the spatial distribution of urban service facilities. In this study, the network kernel density estimation was applied to detect the hot spots of health care service along the road network of Shenzhen, and we further explored the influences of population and road density on the aggregate intensity distributions at the community level, using spatial stratified heterogeneity analyses. Then, we measured the spatial clustering patterns of health care facilities in each of the ten districts of Shenzhen using the network K-function, and the interrelationships between health care facilities and hypertension patients. The results can be used to examine the reasonability of the existing health care system, which would be valuable for developing more effective prevention, control, and treatment of chronic health conditions. Further research should consider the influence of nonspatial factors on health care service access.

Prediction of Lime Tolerance in Rhododendron Based on Herbarium Specimen and Geochemical Data

Rhododendrons are typically known to be calcifuges that cannot grow well in lime soils. Data on lime tolerance of different taxa in Rhododendron are scarce. Habitats of naturally distributed specimens of genus Rhododendron were compiled as Chinese text-based locations from the Chinese Virtual Herbarium. The locations were then geocoded into latitude/longitude pairs and subsequently connected to soil characteristics including pH and CaCO₃ from the Harmonized World Soil Database (HWSD). Using the upper quartile values of pH > 7.2 and CaCO₃ > 2% weight in topsoil as threshold, we predicted the lime tolerant taxa. A dataset of 31,146 Rhododendron specimens including the information on taxonomy, GPS locations and soil parameters for both top- and subsoil was built. The majority of the specimens were distributed in soils with moderately acidic pH and without presence of CaCO₃. 76 taxa with potential lime tolerance were predicted out of 525 taxa. The large scale data analysis based on combined data of geocoded herbarium specimens and HWSD allows identification of valuable Rhododendron species, subspecies or botanical varieties with potential tolerance to lime soils with higher pH. The predicted tolerant taxa are valuable resources for an in-depth evaluation of lime tolerance or for further use in horticulture and breeding.