Social media platforms have become part of many people's lives. Users are spending more and more time on these platforms, creating an active and passive digital footprint through their interaction. This footprint has high research potential in many research areas because understanding people's communication on social media is essential in understanding their values, attitudes, experiences and behaviors. Researchers found that the use of social networking sites impacts adolescents' eating behavior. If we define adolescents as individuals between ages 10 and 24 (WHO's definition), 76% of USA young people at age 18-⁠24 use Instagram, so the Instagram social network analysis is important for understanding young people's expressions in the context of healthy food. This study aims to identify the main topic associated with healthy food on the Instagram social network via hashtag and community analysis based on 2,045,653 messages created by 427,936 individual users. The results show that users most associate Healthy food with healthy lifestyle, fitness, weight loss and diet. In terms of food, these are foods that are Vegan, Homemade, Clean and Plant-based. Given that young people change their behavior in relation to people's behavior on social networks, it is possible to use this data to predict their future association with healthy food characteristics.

Healthy Food on Instagram Social Network: Vegan, Homemade and Clean Eating.

Online social networks have become an everyday aspect of many people's lives. Users spend more and more time on these platforms and, through their interactions on social media platforms, they create active and passive digital footprints. These data have a strong potential in many research areas; indeed, understanding people's communication on social media is essential for understanding their attitudes, experiences, behaviors and values. Researchers have found that the use of social networking sites impacts eating behavior; thus, analyzing social network data is important for understanding the meaning behind expressions used in the context of healthy food. This study performed a communication analysis of data from the social network Twitter, which included 666,178 messages posted by 168,134 individual users. These data comprised all tweets that used the #healthyfood hashtag between 2019 and 2020 on Twitter. The results revealed that users most commonly associate healthy food with a healthy lifestyle, diet, and fitness. Foods associated with this hashtag were vegan, homemade, and organic. Given that people change their behavior according to other people's behavior on social networks, these data could be used to identify current and future associations with current and future perceptions of healthy food characteristics.

Healthy Food on the Twitter Social Network: Vegan, Homemade, and Organic Food.

Social networks have become a common part of many people’s daily lives. Users spend more and more time on these platforms and create an active and passive digital footprint through their interaction with other subjects. These data have high research potential in many fields, because understanding people’s communication on social media is essential to understanding their attitudes, experiences and behaviours. Social media analysis is a relatively new subject. There is still a need to develop methods and tools for researchers to help solve typical problems associated with this area. A researcher will be able to focus on the subject of research entirely. This article describes the Social Media Analysis based on Hashtag Research (SMAHR) framework, which uses social network analysis methods to explore social media communication through a network of hashtags. The results show that social media analysis based on hashtags provides information applicable to theoretical research and practical strategic marketing and management applications.

Framework for Social Media Analysis Based on Hashtag Research

Multimodal sentiment analysis (MSA) is a very challenging task due to its complex and complementary interactions between multiple modalities, which can be widely applied into areas of product marketing, public opinion monitoring, and so on. However, previous works directly utilized the features extracted from multimodal data, in which the noise reduction within and among multiple modalities has been largely ignored before multimodal fusion. This paper proposes a multi-level attention map network (MAMN) to filter noise before multimodal fusion and capture the consistent and heterogeneous correlations among multi-granularity features for multimodal sentiment analysis. Architecturally, MAMN is comprised of three modules: multi-granularity feature extraction module, multi-level attention map generation module, and attention map fusion module. The first module is designed to sufficiently extract multi-granularity features from multimodal data. The second module is constructed to filter noise and enhance the representation ability for multi-granularity features before multimodal fusion. And the third module is built to extensibly mine the interactions among multi-level attention maps by the proposed extensible co-attention fusion method. Extensive experimental results on three public datasets show the proposed model is significantly superior to the state-of-the-art methods, and demonstrate its effectiveness on two tasks of document-based and aspect-based MSA tasks.

Multi-Level Attention Map Network for Multimodal Sentiment Analysis

The exponential growth of the internet and a multi-fold increase in social media users in the last decade have resulted in a massive growth of unstructured data. Aspect-Based Sentiment Analysis (ABSA) is challenging because it performs a fine-grain analysis; it is a text analysis technique where the opinions group is based on the aspect. The Aspect Extraction (AE) task is one of the core subtasks of ABSA; it helps to identify aspect terms in the text, comments, or reviews. The challenge of the Arabic AE task increases due to the complexity of the Arabic language. This work aims to develop the Arabic AE task by proposing transfer learning using state-of-art pre-trained contextual language models. We concatenate the Bidirectional Encoder Representation from Transformers (BERT) language model and contextualize string embeddings (Flair embedding) as a stacked embeddings layer for better word representation for Arabic language. Then, we extend it with different deep learning network architectures. For Arabic AE, the model is developed by concatenating the Arabic contextual language model, AraBERT, and Flair embedding as a contextual stacked embeddings layer with an extended layer, BiLSTM-CRF or BiGRU-CRF, for sequence labeling. Our proposed models are called BF-BiLSTM-CRF and BF-BiGRU-CRF. The proposed model is evaluated using the Arabic Hotel’s reviews dataset. For performance evaluation, we used the F1 score. The experimental results show that the proposed BF-BiLSTM-CRF configuration outperformed the baseline and other models by achieving an F1score of 79.7%.

https://ieeexplore.ieee.org/ielx7/6287639/9668973/09733905.pdf

Arabic Aspect Extraction Based on Stacked Contextualized Embedding With Deep Learning

In recent years, with the popularity of social media, users are increasingly keen to express their feelings and opinions in the form of pictures and text, which makes multimodal data with text and pictures the con tent type with the most growth. Most of the information posted by users on social media has obvious sentimental aspects, and multimodal sentiment analysis has become an important research field. Previous studies on multimodal sentiment analysis have primarily focused on extracting text and image features separately and then combining them for sentiment classification. These studies often ignore the interaction between text and images. Therefore, this paper proposes a new multimodal sentiment analysis model. The model first eliminates noise interference in textual data and extracts more important image features. Then, in the feature-fusion part based on the attention mechanism, the text and images learn the internal features from each other through symmetry. Then the fusion features are applied to sentiment classification tasks. The experimental results on two common multimodal sentiment datasets demonstrate the effectiveness of the proposed model.

https://www.mdpi.com/2073-8994/12/12/2010/pdf

Sentiment Analysis of Social Media via Multimodal Feature Fusion

With continuous developments in deep learning, image semantic segmentation technology has also undergone great advancements and been widely used in many fields with higher segmentation accuracy. This paper proposes an image semantic segmentation algorithm based on a deep neural network. Based on the Mask Scoring R-CNN, this algorithm uses a symmetrical feature pyramid network and adds a multiple-threshold architecture to improve the sample screening precision. We employ a probability model to optimize the mask branch of the model further to improve the algorithm accuracy for the segmentation of image edges. In addition, we adjust the loss function so that the experimental effect can be optimized. The experiments reveal that the algorithm improves the results.

https://www.mdpi.com/2073-8994/13/2/207/pdf

Image Semantic Segmentation Use Multiple-Threshold Probabilistic R-CNN with Feature Fusion

This paper explores the combination of a classic mathematical function named “hyperbolic tangent” with a metaheuristic algorithm, and proposes a novel hybrid genetic algorithm called NSGA-II-BnF for multi-objective decision making. Recently, many metaheuristic evolutionary algorithms have been proposed for tackling multi-objective optimization problems (MOPs). These algorithms demonstrate excellent capabilities and offer available solutions to decision makers. However, their convergence performance may be challenged by some MOPs with elaborate Pareto fronts such as CFs, WFGs, and UFs, primarily due to the neglect of diversity. We solve this problem by proposing an algorithm with elite exploitation strategy, which contains two parts: first, we design a biased elite allocation strategy, which allocates computation resources appropriately to elites of the population by crowding distance-based roulette. Second, we propose a self-guided fast individual exploitation approach, which guides elites to generate neighbors by a symmetry exploitation operator, which is based on mathematical hyperbolic tangent function. Furthermore, we designed a mechanism to emphasize the algorithm’s applicability, which allows decision makers to adjust the exploitation intensity with their preferences. We compare our proposed NSGA-II-BnF with four other improved versions of NSGA-II (NSGA-IIconflict, rNSGA-II, RPDNSGA-II, and NSGA-II-SDR) and four competitive and widely-used algorithms (MOEA/D-DE, dMOPSO, SPEA-II, and SMPSO) on 36 test problems (DTLZ1–DTLZ7, WGF1–WFG9, UF1–UF10, and CF1–CF10), and measured using two widely used indicators—inverted generational distance (IGD) and hypervolume (HV). Experiment results demonstrate that NSGA-II-BnF exhibits superior performance to most of the algorithms on all test problems.

Elite Exploitation: A Combination of Mathematical Concept and EMO Approach for Multi-Objective Decision Making

This study aimed to compare the behavior of rat bone marrow mesenchymal stem cells and rabbit auricle blastema cells implanted in non-cellular gingival tissue scaffold of human. In this regard, the tissues obtained from gingival surgeries in the dental clinic were de-cellulated using two detergents of sodium dodecyl sulfate and triton 100-X. After washing and sterilization, they were used as a scaffold for culture with bone marrow mesenchymal stem cells of rats. Using light and electron microscopy, these scaffolds were examined before and after 1, 2, and 4 weeks of cell culture. Also, the prepared three-dimensional scaffold was placed in the blastema ring obtained from the rabbit earlobe punch. The samples were evaluated 1, 2, and 4 weeks after culture based on histology techniques. The results showed that the study of scaffolds by electron microscopy showed preserving the epithelium matrix and the collagen fibers in the tissue. Structures similar to the epithelium were created in both samples. In addition, induction of cellular secretion was observed in scaffold cells migrating to the scaffold. In general, scaffolds made from human gums can be a good platform for studying cellular behaviors. Of course, further experiments to determine the nature of differentiated cells can help advance our knowledge of matrix cell interactions.

Preparation of three-dimensional matrices of human gingival tissue and comparison of induction of mesenchymal and blastema stem cell behaviors in prepared scaffolds.

Jianxin Liu

Papers

Sentiment Analysis of Social Media via Multimodal Feature Fusion

Image Semantic Segmentation Use Multiple-Threshold Probabilistic R-CNN with Feature Fusion

Elite Exploitation: A Combination of Mathematical Concept and EMO Approach for Multi-Objective Decision Making

Preparation of three-dimensional matrices of human gingival tissue and comparison of induction of mesenchymal and blastema stem cell behaviors in prepared scaffolds.