Event extraction from Twitter using Non-Parametric Bayesian Mixture Model with Word Embeddings

Abstract

To extract structured representations of newsworthy events from Twitter, unsupervised models typically assume that tweets involving the same named entities and expressed using similar words are likely to belong to the same event. Hence, they group tweets into clusters based on the cooccurrence patterns of named entities and topical keywords. However, there are two main limitations. First, they require the number of events to be known beforehand, which is not realistic in practical applications. Second, they don't recognise that the same named entity might be referred to by multiple mentions and tweets using different mentions would be wrongly assigned to different events. To overcome these limitations, we propose a nonparametric Bayesian mixture model with word embeddings for event extraction, in which the number of events can be inferred automatically and the issue of lexical variations for the same named entity can be dealt with properly. Our model has been evaluated on three datasets with sizes ranging between 2,499 and over 60 million tweets. Experimental results show that our model outperforms the baseline approach on all datasets by 5-8% in F-measure.

Open Event Extraction from Online Text using a Generative Adversarial Network

“…• DPEMM (Zhou et al, 2017) is a nonparametric mixture model for event extraction. It addresses the limitation of LEM that the number of events should be known beforehand.…”

“…Assuming that each document is assigned to a single event, which is modeled as a joint distribution over the named entities, the date and the location of the event, and the event-related keywords, Zhou et al (2014) proposed an unsupervised Latent Event Model (LEM) for open-domain event extraction. To address the limitation that LEM requires the number of events to be pre-set, Zhou et al (2017) further proposed the Dirichlet Process Event Mixture Model (DPEMM) in which the number of events can be learned automatically from data. However, both LEM and DPEMM have two limitations: (1) they assume that all words in a document are generated from a single event which can be represented by a quadruple <entity, location, keyword, date>.…”

Detecting Denial-of-Service Attacks from Social Media Text: Applying NLP to Computer Security

“…They showed you can cluster and extract events like concerts, movies, and performances into a calendar. General event detection from social media has continued in several threads (Benson et al, 2011;Popescu et al, 2011;Anantharam et al, 2015;Wei, 2016;Zhou et al, 2017). Guo et al (2013) link tweets to news stories using an annotated dataset.…”

Graph-based Event Extraction from Twitter

“…Very recently, Zhou et al (2017) use a nonparametric Bayesian Mixture Model leveraged with word embeddings to create event clusters from tweets. In this approach, events are modeled as a 4-tuple y, l, k, d modeling non-location NEs, location NEs, event keywords and date.…”

“…For each dataset, we compare our approach with state-of-the-art approaches. For the FSD dataset, we compare with LEM Bayesian model (Zhou et al, 2011) and DPEMM Bayesian model enriched with word embeddings (Zhou et al, 2017). For the EVENT2012 dataset, we compare our results with Named Entity-Based Event Detection approach (NEED) (McMinn and Jose, 2015) and Event Detection Onset (EDO) (Katragadda et al, 2016).…”