Web services for predictive text analytics

Citizen participation has become an integral part of the democratic process. Tree company, a leading authority in this field, was asked by the Flemish Minister of Education to gauge public opinion on education reforms. Textgain aided in the collection and analysis of opinions on social media and helped categorize the content of the discussion forum using automatic text classification techniques. We are currently collaborating with Tree Company to develop a pipeline to streamline this process under the auspices of VLAIO.
Scanning social media for unwanted content is very much like trying to find a needle in a haystack. Text analytics can provide a solution to this problem by automating the search for specific types of content in large data streams. For Zumo Media, we developed text classifiers to detect jihadist propaganda on social media. This allowed us to gain insight into the rhetoric and prevalence of the phenomenon on various channels, such as Twitter, Telegram and topical discussion forums.
Loyal and happy customers are the foundation of every successful company and their satisfaction is the most effective way to guarantee continued success in the market. Online questionnaires are a great way to get feedback from the customer. However, questionnaires with free-text elements are often only evaluated manually. OTARIS (www.otaris.de) use the Textgain API to analyze and classify free text elements in surveys.
We did a few projects for our Alma Mater, the University of Antwerp. For CLiPS, we collected user generated content in order to evaluate the research group's demographic profiling techniques. We also developed a Concordance Interface for the literary archives of the Ruusbroec Institute. For the MIOS research center, we implemented techniques to efficiently collect live data from a custom social media platform in the context of user interaction research.
Many digital publishers are unaware of the value locked inside their treasure chest of articles. Using text analytics, we can discover historical trends and provide strategies for future directions. We worked with the - always innovative - Mo* Magazine, to convert their historical data into actionable visualizations and insights.
We developed Swahili part-of-speech tagging and sentiment analysis for Africa's Voices, the go-to organization for robust, credible and innovative research within the African context. They will use our tools to perform social media monitoring and expand the range of African voices they can listen to.
For the Department of Dutch linguistics, we harvested a massive amount of historical news paper data from the archives of the National Library of the Netherlands. We subsequently wrangled the data to allow for effective corpus-based research on diachronic language variation.
DEME commissioned a study from Textgain to stimulate bridge-building between the company's engineers and marketeers. Our report outlined several innovation strategies in computational creativity and text analytics and related them to specific engineering and marketing challenges at DEME.