An algorithm for event detection based on social media data

Cui, Wenjuan; Wang, Pengfei; Du, Yi; Chen, Xin; Guo, Danhuai; Zhou, Yuanchun

doi:10.1016/j.neucom.2016.09.127

Cited by 29 publications

(19 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figure 4 shows the ROC curves for the results of a single fold of Naive Bayes classification that uses the features extracted by each selection methods. The classification results of the proposed method outperformed the benchmarks and state of the art developed by [40], [8], [39], [10], [1], [25].…”

Section: Experiments and Resultsmentioning

confidence: 86%

Event Detection in Twitter: A Keyword Volume Approach

Hossny

Mitchell

2018

2018 IEEE International Conference on Data Mining Workshops (ICDMW)

View full text Add to dashboard Cite

Event detection using social media streams needs a set of informative features with strong signals that need minimal preprocessing and are highly associated with events of interest. Identifying these informative features as keywords from Twitter is challenging, as people use informal language to express their thoughts and feelings. This informality includes acronyms, misspelled words, synonyms, transliteration and ambiguous terms. In this paper, we propose an efficient method to select the keywords frequently used in Twitter that are mostly associated with events of interest such as protests. The volume of these keywords is tracked in real time to identify the events of interest in a binary classification scheme. We use keywords within wordpairs to capture the context. The proposed method is to binarize vectors of daily counts for each word-pair by applying a spike detection temporal filter, then use the Jaccard metric to measure the similarity of the binary vector for each word-pair with the binary vector describing event occurrence. The top n wordpairs are used as features to classify any day to be an event or non-event day. The selected features are tested using multiple classifiers such as Naive Bayes, SVM, Logistic Regression, KNN and decision trees. They all produced AUC ROC scores up to 0.91 and F1 scores up to 0.79. The experiment is performed using the English language in multiple cities such as Melbourne, Sydney and Brisbane as well as the Indonesian language in Jakarta. The two experiments, comprising different languages and locations, yielded similar results.

show abstract

Section: Experiments and Resultsmentioning

confidence: 86%

Event Detection in Twitter: A Keyword Volume Approach

Hossny

Mitchell

2018

2018 IEEE International Conference on Data Mining Workshops (ICDMW)

View full text Add to dashboard Cite

show abstract

“…Cui et al [19] proposes a foodborne disease event detection method. Faced with massive microblog data, a SVM (Support Vector Machine) classifier is trained to filter noise information by taking into account ten features of user's fan number, user's attention number, user's profile length, post number, the average forwarding number, the average like number, the average comment number, publishing time, the average link number and so on.…”

Section: Related Workmentioning

confidence: 99%

Sina Weibo Bursty Event Detection Method

Chen

Liang

2019

IEEE Access

View full text Add to dashboard Cite

In response to the recent frequent outbreaks of bursty events, if the supervision is not strengthened, the grievances caused by it will have an extremely negative impact on society. Therefore, how to effectively detect bursty events in social networks has become the focus of research. In order to eliminate the interference of local daily events generated by noisy data, this paper proposes a bursty event detection method for quantifying the influence of microblog text. Through the analysis of high-impact microblogs, the burst words are mined, the potential bursty event data sets are constructed, and the k-means cluster analysis method is applied to detect the event. In order to verify the method, this paper compares with the commonly used detection methods on the real dataset. The experimental results show that the method has certain improvement in accuracy and efficiency.

show abstract

“…These learningbased methods modeled history sensor data based on various classification models, e.g. Neural Network [18], Support Vector Machine [19], Markov Random Field [20] etc., and then classified real-time sensor data. Singh et al [19] proposed a distributed machine learning approach for event detection in two phases, base phase and meta phase.…”

Section: ) Learning-based Approachmentioning

confidence: 99%

A Declarative Service-Based Method for Adaptive Aggregation of Sensor Streams

Zhang

Liu

et al. 2019

IEEE Access

View full text Add to dashboard Cite

The dynamic interventions among sensor streams bring new challenges for the Internet of Things applications to derive meaningful information from a large amount of sensor streams. This paper proposes a declarative service-based method for capturing meaningful events from dynamic sensor streams based on our previous service abstraction, which can increase the value density of primitive sensor streams and raise the accessibility for higher level business applications. For aggregating related events from different sources, we set up declarative rules and compose cascaded events through service collaboration. In addition, for dealing with dynamic interventions among sensor streams, we propose a declarative rule evolution method to realize adaptive service collaboration at runtime. This paper also reports the tryout use of our method in the China Power Grid for detecting abnormal situations of power quality. Through a series of experiments based on real sensor data in the Power Grid, we verified that our method can capture events with high and steady precision.

show abstract

An algorithm for event detection based on social media data

Cited by 29 publications

References 16 publications

Event Detection in Twitter: A Keyword Volume Approach

Event Detection in Twitter: A Keyword Volume Approach

Sina Weibo Bursty Event Detection Method

A Declarative Service-Based Method for Adaptive Aggregation of Sensor Streams

Contact Info

Product

Resources

About