The full-text search functionality allows selecting social media posts (for the time being only for the Twitter collection) based on some post attributes: The attributes can be either equal or not equal to a value entered by the user, forming a `’rule’. Rules can be composed in groups joined by logical operators (‘and’, ‘or’, ‘not’). The expression can be changed manually by editing it before pressing the ‘Execute query’ button. A ‘Reset’ button allows for clearing the expression. The results […]

Two outputs of the Topic Modelling pipeline (which is run daily, on a days worth of social media data) are the frequency of terms, as well as the development of word embeddings for the day. The API grants the ability to analyze the term frequency data, as well as interact with the word embedding models. The AIReD Term analysis tab interacts with the API to visualize the outputs, and examine the semantic relationships between words in a certain time period. […]

The AIReD topic graph tab provides users with an interactive way to examine the outputs of the Topic Modelling pipeline. The pipeline is run every day, on a days worth of social media data, using BERTopic, to generate a list of topic clusters. These clusters simulate the trending topics of the given day, and each cluster is associated with 30 representative terms – from which the topic of the cluster can be extrapolated. Topic graph construction In order to track the […]

The aggregation tab in AIReD provides ability to aggregate over the social media collections in four different ways, which called different endpoints at the API. Location Time (by day) Seasonality Language Each aggregation is controlled by a start and an end date, to specify the period over which the aggregation is desired. Upon loading the aggregation tab, one can select the start and end date by using the widget in the top left corner. Once these value are selected, all […]

The Australian Internet Observatory Research Dashboard (AIOReD) is an interactive web based application designed to provide user friendly access to the API, to perform social media analysis. The dashboard allows users to interact with data present in the databases, along with some of the topic modelling results and word embeddings that are generated by the ‘live’ machine learning pipelines. The dashboard works best a stable internet connection, and on a laptop or desktop computer, as it can be network intensive […]