Trend predictions are possible at the touch of a button

Data-based trend and scenario analysis

© Sergey Nivens - AdobeStock.de

The strategic decisions that companies make are closely linked to developments in their environment – developments not only in their markets but also in the supporting technologies that can be used to optimize products, services and processes and make them sustainable.

This makes obtaining a valid forecast through specific monitoring of relevant major trends essential for sustainable and targeted corporate development; only in this way can opportunities and risks be identified in good time and strategic decisions be secured in a well-founded manner. But this calls for the right information.

It has never been easier than it is today to access this information: via the databases kept by industry associations or public and scientific institutions, and via news feeds or social media channels – worldwide.

Too much information overwhelms capacity for reading and analysis

However, it is precisely this flood of information from the most diverse sources and in different languages that overwhelms the capacity currently in place for reading and analyzing it all and poses major challenges for the economy: if even a simple search query generates several tens of thousands of results just in Google News in German, it is clear that neither companies themselves nor third-party market research partners are in a position to produce a comprehensive evaluation of all relevant published documents and news – even with a very substantial investment in personnel and time. And this challenge is constantly growing in our increasingly digital and volatile world, which demands ever-faster decisions even in the face of swelling volumes of data and information.

 

Fast and efficient results through new methods for semantic media analysis

 

This is why Fraunhofer SCS and the Nuremberg Institute of Technology are working together to develop automated text analysis systems for knowledge generation and trend analysis based on Semantic Web technologies. The aim is to be able to predict market trends and technological developments quickly without impairing their validity.

These new methods make automated trend predictions possible, where large volumes of unstructured texts in various languages sourced from selected websites and databases are scanned by machine and the information found is processed, fleshed out to address relevant questions and placed in the right (corporate) context.

This enables companies to efficiently assess the maturity of technologies and markets, to recognize changes in customer and competitor behavior at an early stage and to transfer industry-specific findings into strategic scenarios.

How we are predicting trends at the touch of a button

Our methods let us quickly identify valid trends from constantly growing volumes of unstructured information. To this end, we use a procedure we developed in-house to continuously process various data sources such as RSS feeds, homepages, newsletters, social media platform content, scientific publications or patent applications across languages.

By employing Semantic Web tools, we automatically annotate relationships, events and facts, link them using Linked Open Data principles and map them in a dynamic knowledge graph.

On this basis, we identify market- and technology-specific trends, derive industry-specific scenarios from these together with our customers and analyze the relevant indicators over time.

Our research focuses on the following methods:

  • Text mining for the automated exploitation of large, diversified datasets
  • Natural language processing and disambiguation with Semantic Web/Linked Open Data for data preparation and processing as well as for the development of dynamic, semantic data structures
  • Topic modeling for the detection of weak signals
  • Scenario techniques for the embedding of trends and scenarios in the specific business environment
Graphic Automated process for monitoring market and technology changes
© Fraunhofer IIS

Find out more about Fraunhofer SCS

 

Success and added value thanks to data

Sustained success in a changing world: this is the vision of Fraunhofer SCS.

 

Scientific expertise in the reference process

The methodological expertise developed in our fields of research is informed by our specially developed Reference Process for Digital Transformation. Read about what this initiative means to us, and how we can use our expertise to comprehensively support companies.