Identrics. Customisable AI & NLP solutions that help you understand data.
Create content people care about with our topic engagement technology and stay relevant.
Focus on making the right content for the right occasion. Know what’s relevant now – from crypto and soft news to climate protection and the European football championship.
How it works
We can help you find out what your customers care about with respect to their personal data.
1
We look at LDA and user interactions
We employ two statistical methods. First, we provide Latent Dirichlet Allocation (LDA). Then, we look at the frequency of user reactions per document on social networks.
2
Daily monitoring of key pieces of data.
To ensure unbiased data, we keep a close eye on the date, location, language, different named entities, authors, sources, and offline data sets.
3
Quantifying and measuring different datasets.
Collecting data from Facebook comments, shares, and reactions to an article, Twitter article shares and comments; and direct article comments give us a true understanding of what garners the most attention online.
Services
Make content that will get the attention, buzz, and shares it deserves.
Our Topic Engagement services help you create relevant, unique, and unforgettable content.
You get to understand who clicks on what, how long does something keep the attention on an average basis, does it trigger any reactions, what kind of reactions, do people share it, and last but not least, how do people feel about it.
Case Studies
Media Intelligence
Client: Economedia
Problem: In the given media, journalists and editors had to closely analyze the content in other media every day, in order to be aware of what is happening in the world around them. Doing this manually required a huge amount of time and human effort, which could have been spent in a more useful way.
Solution: We applied a topic engagement model to the selection of media (national and regional) which is monitored daily. The average number per day was between 7K and 9K documents.
Thanks to the technological models we described above, we grouped all these documents into thematic cubes, which contain not only the quantitative measures, but also additional metadata which gives even greater insight for the individual topics. Thus, the time of journalists and editors, which otherwise goes into content analysis, can be invested in the development of additional materials or other activities. This way, we provide a much bigger objective view of all the information that needs to be analyzed.
