Search

Following the tremendous amounts of text generated in social networks and news channels, and gaining valuable and dependable insights from diverse sources of information is a tedious task. The challenge is increased during specific periods, for example, in a pandemic event like Covid-19. Existing text categorization methods, such as sentiment classification, aim to help people tackle this challenge by categorizing and summarizing the text content. However, the inherent uncertainty of user-generated text limits their efficiency. This paper proposes a novel architecture based on fuzzy inference and deep learning for sentiment classification that overcomes this limitation. We evaluate the proposed method by applying it to well-known health-related text datasets and comparing the accuracy with state-of-the-art methods. The results show that the proposed fuzzy fusion methods increase the accuracy compared to individual pretrained models. The model also provides an expressive architecture for health news classification.

A challenging problem for human societies that affects human life is the spread of diseases such as COVID-19. Travel, especially air transportation, plays a significant role in the spread of such diseases. Analysis of air traffic data and its interaction with the disease outbreak is one of the applied and research topics. Flights from infected areas to less infected regions increase the outbreak. On the other hand, by following warnings and policies, air travel declines in such periods. The effects are reflected in the collected data with a time gap that requires efficient methods for rich analysis. This aspect is overlooked in existing research. In this paper, a new approach is proposed for spatio-temporal analysis that is based on granular computing and fuses the air traffic data with COVID-19 statistics. Data in diverse sources may be at different levels of granularity. Therefore, analyzing the interaction of two phenomena extracted from different sources is a challenging issue. Moreover, there may not be an exact match in other data sources for some data items, but applying a fuzzy inference system and constructing enriched data granules by using geographical knowledge bases eliminates this shortcoming. The proposed method fills the gaps and adjusts the differences to discover new patterns that other methods cannot detect. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Graphs and graph databases are applicable over a wide range of domains, including text mining and web mining. Using graphs to represent relationships between entities provides enriched models for emerging tasks of web search and information retrieval. Natural language processing algorithms use graphs to model structural relationships of texts efficiently, resulting in improved performance. However, the need to increase the accuracy of graph construction and weight allocation remains a fundamental challenge. Existing methods for these tasks provide limited efficiency and lack scalability for large graphs. In this study, we propose a novel graph-based method for text modeling and running a query to evaluate the similarity of text segments. In this method, the graph corresponding to the text is first created by modeling words and named entities by the state-of-the-art pre-trained BERT model. Graph nodes are then weighted in two stages. In the first stage, the nodes with more generalization obtain higher weights. The second weighting stage is done by the graph obtained from the query text. In this weighting step, nodes are considered important if they are specifically related to the query text. After determining the important nodes in the graph, the semantic similarity between the query text and the texts in the database is measured. The whole process of this framework uses a natural language processing pipeline in Apache Spark scalable platform. The efficiency of the model was evaluated for both distributed and non-distributed configuration and its scalability on a Spark cluster. Evaluation of the accuracy using the Pearson correlation coefficient shows that the proposed method performs higher performance than its competitors.