Hybrid Machine Learning Models of Classifying Residential Requests for Smart Dispatching

Abstract

This paper presents a hybrid machine learning method of classifying residential requests in natural language to responsible departments that provide timely responses back to residents under the vision of digital government services in smart cities. Residential requests in natural language descriptions cover almost every aspect of a city's daily operation. Hence the responsible departments are fine-grained to even the level of local communities. There are no specific general categories or labels for each request sample. This causes two issues for supervised classification solutions, namely 1) the request sample data is unbalanced and 2) lack of specific labels for training. To solve these issues, we investigate a hybrid machine learning method that generates meta-class labels by means of unsupervised clustering algorithms; applies two-word embedding methods with three classifiers (including two hierarchical classifiers and one residual convolutional neural network); and selects the best performing classifier as the classification result. We demonstrate our approach performing better classification tasks compared two benchmarking machine learning models, Naive Bayes classifier and a Multiple Layer Perceptron (MLP). In addition, the hierarchical classification method provides insights into the source of classification errors.

Data-Centric Content Classification of Smart City Residential Services

“…Many department descriptions share similar semantics that refers to a single department name. In our previous work [1], a manually created dictionary mapped the same departments described by different location phrases to the same label, resulting in seventy-three classes.…”

“…We choose to use the ResNet model for the purpose of demonstrating the improvement of learning performance using our data-centric approach. ResNet was applied and demonstrated the best performance in the context of our previous work [1]. The feature vectors are embedded using the Word2Vec model.…”

A Novel Hybrid Machine Learning Classification for the Detection of Bruxism Patients Using Physiological Signals

“…Miskovic [56] suggested a hybrid model for the classification of decision support. Chen et al [57] used the hybrid model for categorizing residential requests in natural language to provide timely replies back to citizens under the vision of digital administration services in smart cities. In these approaches, the designed HML models outperformed in all classifications.…”

Data-Centric Content Classification of Smart City Residential Services

“…Our previous work [1] addressed the first issue by adopting means of unsupervised clustering algorithms to generate meta-class labels. For the second issue, our previous work developed a Word2Vec [2] model as word embedding and a Residual Neural Network [1,3] as the classifier model. The limitation is that Word2Vec embedding tends to measure the similarity between words.…”