Massive Social Media Data Research Articles

Revealing the spatial co-occurrence patterns of multi-emotions from social media data

Emotions play a critical role in understanding human behaviors and are direct indicators of residents' well-being and quality of life. Assessing spatial-emotional interactions is crucial for human-centered urban planning and public mental health. However, prior research has focused on the spatial analysis of every single emotion, ignoring the intricate interactions between multi-emotions and space. To address this gap, we propose a novel framework to reveal the spatial co-occurrence patterns of multi-emotions using massive social media data in Wuhan, China. Specifically, the BERT (bidirectional encoder representations from transformers) pre-trained model is utilized to classify each post into one of five basic emotions. Given the implementation of the K-means algorithm on these emotional results, the emotion-based similarities among different grids are investigated. The qualitative and quantitative results reveal six spatial co-occurrence patterns of conflicting or consistent emotions in urban space, namely, happiness-fear, happiness-anger, balanced emotion, happiness dominated, happiness-surprise, and happiness-sadness. In particular, the balanced emotion pattern is the most prevalent and tends to be spatially concentrated in the city center, while patterns of happiness-anger and happiness-sadness are mainly observed in the suburbs. Plus, results of the Multinomial Logit Model (MNLM) indicate that the spatial multi-emotions co-occurrence patterns are significantly correlated with land use characteristics based on points-of-interest (POIs) data. These findings provide an innovative perspective for understanding the complex interactions between emotions and space, with theoretical and practical implications for designing and maintaining an emotionally healthy city.

Content and sentiment surveillance (CSI): A critical component for modeling modern epidemics.

Comprehensive surveillance systems are the key to provide accurate data for effective modeling. Traditional symptom-based case surveillance has been joined with recent genomic, serologic, and environment surveillance to provide more integrated disease surveillance systems. A major gap in comprehensive disease surveillance is to accurately monitor potential population behavioral changes in real-time. Population-wide behaviors such as compliance with various interventions and vaccination acceptance significantly influence and drive the overall epidemic dynamics in the society. Original infoveillance utilizes online query data (e.g., Google and Wikipedia search of a specific content topic such as an epidemic) and later focuses on large volumes of online discourse data about the from social media platforms and further augments epidemic modeling. It mainly uses number of posts to approximate public awareness of the disease, and further compares with observed epidemic dynamics for better projection. The current COVID-19 pandemic shows that there is an urgency to further harness the rich, detailed content and sentiment information, which can provide more accurate and granular information on public awareness and perceptions toward multiple aspects of the disease, especially various interventions. In this perspective paper, we describe a novel conceptual analytical framework of content and sentiment infoveillance (CSI) and integration with epidemic modeling. This CSI framework includes data retrieval and pre-processing; information extraction via natural language processing to identify and quantify detailed time, location, content, and sentiment information; and integrating infoveillance with common epidemic modeling techniques of both mechanistic and data-driven methods. CSI complements and significantly enhances current epidemic models for more informed decision by integrating behavioral aspects from detailed, instantaneous infoveillance from massive social media data.

Exploring Semantic Relations for Social Media Sentiment Analysis

With the massive social media data available online, the conventional single modality emotion classification has developed into more complex models of multimodal sentiment analysis. Most existing works simply extracted image features at a coarse level, resulting in the absence of partially detailed visual features. Besides, social media data usually contain multiple images, while existing works considered a single image case and used only one image for representing visual features. In fact, it is nontrivial to extend the single image case to the multiple images case, due to the complex relations among multiple images. To solve the above issues, in this paper, we propose a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</b> ated <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">F</b> usion <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S</b> emantic <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</b> elation (GFSR) network to explore semantic relations for social media sentiment analysis. In addition to inter-relations between visual and textual modalities, we also exploit intra-relations among multiple images, potentially improving the sentiment analysis performance. Specifically, we design a gated fusion network to fuse global image embeddings and the corresponding local Adjective Noun Pair (ANP) embeddings. Then, apart from textual relations and cross-modal relations, we employ the multi-head cross attention mechanism between images and ANPs to capture similar semantic contents. Eventually, the updated textual and visual representations are concatenated for the final sentiment prediction. Extensive experiments are conducted on real-world <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Yelp</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Flickr30k</i> datasets, showing that our GFSR can improve about 0.10% to 3.66% in terms of accuracy on the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Yelp</i> dataset with multiple images, and achieve the best accuracy for two classes and the best macro F1 for three classes on the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Flickr30k</i> dataset with a single image.

On streaming disaster damage assessment in social sensing: A crowd-driven dynamic neural architecture searching approach

Motivated by the recent advances in Internet and communication techniques and the proliferation of online social media, social sensing has emerged as a new sensing paradigm to obtain timely observations of the physical world from “human sensors”. In this study, we focus on an emerging application in social sensing – streaming disaster damage assessment (DDA), which aims to automatically assess the damage severity of affected areas in a disaster event on the fly by leveraging the streaming imagery data about the disaster on social media. In particular, we study a dynamic optimal neural architecture searching (NAS) problem in streaming DDA applications. Our goal is to dynamically determine the optimal neural network architecture that accurately estimates the damage severity for each newly arrived image in the stream by leveraging human intelligence from the crowdsourcing systems. The present study is motivated by the observation that the neural network architectures in current DDA solutions are mainly designed by artificial intelligence (AI) experts, which often leads to non-negligible costs and errors given the dynamic nature of the streaming DDA applications and the lack of real-time annotations of the massive social media data inputs. Two critical technical challenges exist in solving our problem: (i) it is non-trivial to dynamically identify the optimal neural network architecture for each image on the fly without knowing its ground-truth label a priori; (ii) it is challenging to effectively leverage the imperfect crowd intelligence to correctly identify the optimal neural network architecture for each image. To address the above challenges, we developed CD-NAS, a dynamic crowd-AI collaborative NAS framework that carefully explores the human intelligence from crowdsourcing systems to solve the dynamic optimal NAS problem and optimize the performance of streaming DDA applications. The evaluation results from a real-world streaming DDA application show that CD-NAS consistently outperforms the state-of-the-art AI and NAS baselines by achieving the highest disaster damage assessment accuracy while maintaining the lowest computational cost.

Using Social Media to Detect Socio-Economic Disaster Recovery

There has been growing interest in harnessing Artificial Intelligence (AI) to improve situational awareness for disaster management. However, to the authors’ best knowledge, few studies have focused on socio-economic recovery. Here, as a first step toward investigating the possibility of developing an AI-based method for detecting socio-economic recovery, this study provides fundamental insights about the correlations between public sentiment on social media and socio-economic recovery activities as reflected in market data. Our result shows multiple correlations between sentiment on social media and the socio-economic recovery activities involved in restarting daily routines. Conventional socio-economic recovery indicators, such as governmental statistical data, have a significant time lag before publishing. Therefore, by taking advantages of the real timeliness and the effectiveness of seizing communication trends of massive social media data, using public sentiment on social media can improve situational awareness in recovery operations.

Analyzing social media data: A mixed-methods framework combining computational and qualitative text analysis.

To qualitative researchers, social media offers a novel opportunity to harvest a massive and diverse range of content without the need for intrusive or intensive data collection procedures. However, performing a qualitative analysis across a massive social media data set is cumbersome and impractical. Instead, researchers often extract a subset of content to analyze, but a framework to facilitate this process is currently lacking. We present a four-phased framework for improving this extraction process, which blends the capacities of data science techniques to compress large data sets into smaller spaces, with the capabilities of qualitative analysis to address research questions. We demonstrate this framework by investigating the topics of Australian Twitter commentary on climate change, using quantitative (non-negative matrix inter-joint factorization; topic alignment) and qualitative (thematic analysis) techniques. Our approach is useful for researchers seeking to perform qualitative analyses of social media, or researchers wanting to supplement their quantitative work with a qualitative analysis of broader social context and meaning.

When social computing meets soft computing: opportunities and insights

The characteristics of the massive social media data, diverse mobile sensing devices as well as the highly complex and dynamic user’s social behavioral patterns have led to the generation of huge amounts of high dimension, uncertain, imprecision and noisy data from social networks. Thanks to the emerging soft computing techniques which unlike the conventional hard computing. It is widely used for coping with the tolerant of imprecision, uncertainty, partial truth, and approximation. One of the most important and promising applications is social network analysis (SNA) that is the process of investigating social structures and relevant properties through the use of network and graph theories. This paper aims to survey various SNA approaches using soft computing techniques such as fuzzy logic, formal concept analysis, rough sets theory and soft set theory. In addition, the relevant software packages about SNA are clearly summarized.

Real-Time Multimedia Social Event Detection in Microblog.

Detecting events from massive social media data in social networks can facilitate browsing, search, and monitoring of real-time events by corporations, governments, and users. The short, conversational, heterogeneous, and real-time characteristics of social media data bring great challenges for event detection. The existing event detection approaches rely mainly on textual information, while the visual content of microblogs and the intrinsic correlation among the heterogeneous data are scarcely explored. To deal with the above challenges, we propose a novel real-time event detection method by generating an intermediate semantic level from social multimedia data, named microblog clique (MC), which is able to explore the high correlations among different microblogs. Specifically, the proposed method comprises three stages. First, the heterogeneous data in microblogs is formulated in a hypergraph structure. Hypergraph cut is conducted to group the highly correlated microblogs with the same topics as the MCs, which can address the information inadequateness and data sparseness issues. Second, a bipartite graph is constructed based on the generated MCs and the transfer cut partition is performed to detect the events. Finally, for new incoming microblogs, incremental hypergraph is constructed based on the latest MCs to generate new MCs, which are classified by bipartite graph partition into existing events or new ones. Extensive experiments are conducted on the events in the Brand-Social-Net dataset and the results demonstrate the superiority of the proposed method, as compared to the state-of-the-art approaches.

SentiStory: multi-grained sentiment analysis and event summarization with crowdsourced social media data

The massive social media data bring timely, multi-dimensional and rich information. Recently, many researchers have worked on event summarization with crowdsourced social media data. While existing works mostly focus on text-based summary, they only summarize representative microblogs. Public sentiment for the event is also valuable; however, this is not explored in microblogging event summary. In this paper, we propose SentiStory, which is a multi-grained sentiment analysis and event summarization system that summarizes event from two levels: coarse-grained and fine-grained sentiment analysis. In coarse-grained analysis, it discovers microblogs which are important in sentiment, while in fine-grained analysis, it detects significant change of sentiment in the event and identifies which microblog causes the change. Specifically, the proposed system comprises two modules: (1) the microblog preprocessing module firstly reduces redundant information and extracts useful information from the microblog database, and then, it separates different aspects of the event and clusters the same aspect together in a clue. (2) The multi-grained sentiment analysis model analyzes microblogs from two levels: coarse-grained and fine-grained. We perform detailed experimental study on real dataset collected from Sina Weibo, and the results demonstrate the effectiveness of our approach.

Multi-Modal Event Topic Model for Social Event Analysis

With the massive growth of social events in Internet, it has become more and more difficult to exactly find and organize the interesting events from massive social media data, which is useful to browse, search, and monitor social events by users or governments. To deal with this problem, we propose a novel multi-modal social event tracking and evolution framework to not only effectively capture multi-modal topics of social events, but also obtain the evolutionary trends of social events and generate effective event summary details over time. To achieve this goal, we propose a novel multi-modal event topic model (mmETM), which can effectively model social media documents, including long text with related images, and learn the correlations between textual and visual modalities to separate the visual-representative topics and non-visual-representative topics. To apply the mmETM model to social event tracking, we adopt an incremental learning strategy denoted as incremental mmETM, which can obtain informative textual and visual topics of social events over time to help understand these events and their evolutionary trends. To evaluate the effectiveness of our proposed algorithm, we collect a real-world dataset to conduct various experiments. Both qualitative and quantitative evaluations demonstrate that the proposed mmETM algorithm performs favorably against several state-of-the-art methods.

The Role of Social Media in World as a Key Instrument

It’s no secret that “word of mouth” has always been considered one of the most powerful forms of marketing for a business. Social media today is the ultimate word-of-mouth marketing. Getting people to connect with you online in the social space helps spread the word about who your company is and your products and services. Social media has changed the way the world communicates, creating real time interconnectivity between people, businesses, and geographies. As social media continues to grow in popularity, brands are looking for ways to increase their fan base and leverage their social presence to increase sales and brand affinity. Social media is a powerful tool for engaging your customers and fans in a conversation. So there are really two functions of social media for business: 1) Are you participating in the conversation and sharing, and 2) Are you listening and monitoring what is being said about you. Social media takes work, there’s no doubt about it. Any business that has tried running their own Facebook and Twitter profiles knows that gaining new customers is more difficult than it seems. What role should social media play in your marketing? At the end of the day, with right tools and strategy in place, social media marketing can be scary and should/will play a large part in your overall digital strategy. Social media’s role in your overall digital strategy comes down to four main things: 1) Building a fan base 2) Reputation monitoring 3) Conducting customer service 4) Generating leads Social media mining is a process of representing, analyzing, and extracting actionable patterns from social media data. Social media mining introduces basic concepts and principal algorithms suitable for investigating massive social media data; it discusses theories and methodologies from different disciplines such as computer science, data mining, machine learning, social network analysis, network science, sociology, ethnography, statistics, optimization, and mathematics. It encompasses the tools to formally represent, measure, model, and mine meaningful patterns from large-scale social media data. Keywords : Internet, Social networking media, Social media,Communication etc.

Social Event Classification via Boosted Multimodal Supervised Latent Dirichlet Allocation

With the rapidly increasing popularity of social media sites (e.g., Flickr, YouTube, and Facebook), it is convenient for users to share their own comments on many social events, which successfully facilitates social event generation, sharing and propagation and results in a large amount of user-contributed media data (e.g., images, videos, and text) for a wide variety of real-world events of different types and scales. As a consequence, it has become more and more difficult to exactly find the interesting events from massive social media data, which is useful to browse, search and monitor social events by users or governments. To deal with these issues, we propose a novel boosted multimodal supervised Latent Dirichlet Allocation (BMM-SLDA) for social event classification by integrating a supervised topic model, denoted as multi-modal supervised Latent Dirichlet Allocation (mm-SLDA), in the boosting framework. Our proposed BMM-SLDA has a number of advantages. (1) Our mm-SLDA can effectively exploit the multimodality and the multiclass property of social events jointly, and make use of the supervised category label information to classify multiclass social event directly. (2) It is suitable for large-scale data analysis by utilizing boosting weighted sampling strategy to iteratively select a small subset of data to efficiently train the corresponding topic models. (3) It effectively exploits social event structure by the document weight distribution with classification error and can iteratively learn new topic model to correct the previously misclassified event documents. We evaluate our BMM-SLDA on a real world dataset and show extensive experimental results, which demonstrate that our model outperforms state-of-the-art methods.

SocialHelix: visual analysis of sentiment divergence in social media

Social media allow people to express and promote different opinions, on which people’s sentiments to a subject often diverge when their opinions conflict. An intuitive visualization that unfolds the process of sentiment divergence from the rich and massive social media data will have far-reaching impact on various domains including social science, politics and economics. In this paper, we propose a visual analysis system, SocialHelix, to achieve this goal. SocialHelix is a novel visual design which enables the users to detect and trace topics and events occurring in social media, and to understand when and why divergences occurred and how they evolved among different social groups. We demonstrate the effectiveness and usefulness of SocialHelix by conducting in-depth case studies on tweets related to the national political debates.

A data-driven framework for archiving and exploring social media data

Social media data are available and accumulated at the extra-byte level every day. As social media applications are widely deployed in various platforms from personal computers to mobile devices, they are becoming a natural extension to human sensory system. The synthesis of social media with human intelligence has the potential to be the intelligent sensor network of unprecedented scale and capacity. However, it also poses several grand challenges to archive and retrieve information from massive social media data. One of these challenges is how to archive, retrieve and mine such massive unstructured data set efficiently to support real-time emergency response. To explore potential solutions, this paper utilizes parallel computing methods to harvest social media data sets, using Twitter as an example, and to store, index, query and analyse them. Within this framework, a Not Only SQL database (DB), MongoDB, is used to store data as document entries rather than relational tables. To retrieve information from the massive data sets efficiently, several strategies are used: (1) data are archived in the MongoDB across multiple collections with each collection containing a subset of the accumulated data, (2) parallel computing is applied to query and process data from each collection and (3) data are duplicated across multiple servers to support massive concurrent access of the data sets. This study has also tested the performance of spatiotemporal query, concurrent user requests and sentiment analysis over multiple DB servers, and performance benchmark results showed that the proposed approach could provide a solution for processing massive social media data with more than 40% performance improvement. A proof-of-concept prototype implements the design to harvest, process and analyse tweets.

FluMapper: A cyberGIS application for interactive analysis of massive location‐based social media

SUMMARYSocial media have experienced a spectacular rise in popularity, attracting hundreds of millions of users and generating unprecedented amount of content that increasingly contain location and place information. Collectively, the massive location information in these data provides an excellent opportunity to better understand many geographic phenomena and geospatial dynamics in a timely fashion. Recent studies capitalizing on social networking and media data show significant societal impacts in many areas including prediction of stock market and infectious disease surveillance. However, because location‐based social media data are often massive, generated dynamically, and unstructured, significant computation, data, and visualization challenges need to be resolved. This research aims to demonstrate the use of massive social media data to interactively analyze spatiotemporal events across spatial and temporal scales, by establishing a data‐driven framework using cyberGIS—geographic information systems (GIS) based on advanced cyberinfrastructure—to resolve aforementioned challenges. Specifically, FluMapper—an application on the CyberGIS Gateway—is employed as a case study to demonstrate the data‐driven framework and seamless integration of massive location‐based social media data and spatial analytical services within the online problem solving environment of the Gateway. FluMapper presents integrated results from two complementary spatial analyses: (i) an interactive exploration of spatial distribution of flu risk and (ii) dynamic mapping of movement patterns, across multiple spatial, and temporal scales. The seamless integration of these two analyses through the framework illustrates the potential of cyberGIS to resolve the compute and data challenges of analyzing near real‐time social media data in an efficient and scalable manner and to support interactive visualization. Copyright © 2014 John Wiley &amp; Sons, Ltd.