Social Media Data Visualizations




Introduction


We have vast experience in producing data visualizations from scraped data of all major online social networks and aggregators - Twitter, Instagram, Reddit and others. As part of our projects and platforms, we have scraped over billion data points (tweets, posts, followers, etc.) from online social networks.


Example of Visualization - Social Media Mentions on Twitter


For several of our projects, we provided mentions analytics - the number of mentions of an entity through time. An example of hourly mentions for #GDPR on Twitter (from one of our projects):



We use machine learning models for sentiment analysis and opinion mining of tweets, posts, reviews and other texts. Hourly sentiment of tweets containing hashtag #digitaleconomy (from one of our projects):



Sentiment and mentions analytics can be combined to assess in real-time the sentiment and interest in entities, decisions, events and other topics. Example for @Eu_commision:



We use advanced natural language processing (NLP) methods such as Latent Dirichlet Allocation (LDA) and Latent Semantic Indexing (LSI) to determine topics discussed in social media. Example of topics determined from tweets mentioning #digitalhealth on 18th July:



Named entity recognition (NER) denotes the process of entity extraction from texts, whereby entities can be persons, organizations, locations, countries, etc. NER system in combination with other analytics helps in numerous questions and analytical tasks, such as:
An example of our NER visualization as applied on a document „A new era in EU-China relations: more wide ranging strategic cooperation?“ with automatic extraction of persons, organizations, geopolitical entities, events and laws.