COLLECTION AND MERGING OF DATA

Data collection and preparation is crucial to be able to develop the data driven business model and maximize benefits in relation to your strategy. We collect your data, structured or unstructured, by applying the most advanced quantitative collection technics (such as NLP, machine learning, elastic search etc.) and qualitative technics (observation studies, field studies, sociology and ethnography).

To generate value, it is imperative to be able to collect and prepare data and merge it with other relevant data. We assist collect and assess your own data, but we also merge and enrich it with our own valuable data and relevant external data from public registries, open data sources, sensors, social media data etc.

DAMVAD Analytics typical services

  • Identify high value data assets and recommend applications to maximize your value
  • Assist in structuring data to be ready for analytics
  • Structure unstructured data by applying machine learning and natural language processing
  • Link and merge your data with DAMVAD Analytics unique proprietary data as well as public register data and open data sources
  • Extract data on companies, boards and managenets from public data sources
  • Access open data sources to extract data for analytics and intelligence (API, ETL etc.), e.g.
  • Collecting complex unstructured data through advanced qualitative techniques and tools
  • Structuring vast volumes of unstructured data

    Client: A European top-10 research foundation

    Goals: To digitalize, systemize and upgrade internal data in order to be able to monitor and measure impact

    For a large research foundation, we applied big data tools to digitalize, systemize and upgrade their data. The project enabled the foundation to measure the output of their research grants back in time.

    Many research foundations have been in business for many years. These days, they are digitizing their application and reporting information from grantees.  The aim is to be better at assessing possibilities of success, monitor output and predict outcome. However in research some impacts might not show before after 15 or 20 years. Therefore, it is necessary to apply older data from older grants to measure impact and create prediction models. This data is regularly stored in different formats like Word, Pdf and scanned image files. .

    In this case, we collected and systemized the data by applying big analytics techniques in order to read and understand unstructured text. The project consisted of reading and transforming vast numbers of image files and digital documents with different formats of unstructered text to structured files. In order to structure the information, we applied natural language processing machine leaning algorithms to identify academic publications being published during the project. Finally, the data was verified by connecting and comparing the results with global journal databases, like Scopus and PubMed.

    The outcome of the project consisted of a large and structured dataset enabling the foundation to perform analytics on the grants they had given and the outcome in terms of academic and research excellence.

  • Collecting and optimizing data for analytics

    Client: Three Nordic Universities

    Goal: To merge data sets enabling to analyse the results of industry-university collaboration

    Many universities want to improve their strategic collaboration with industry in order to maximize impact and be relevant as partner university. Furthermore, universities want to show their return to society given their important societal role. DAMVAD Analytics assist universities in the Nordics to measure their societal impact. In order to do this, we first collect university data on their collaboration with private companies and enrich it with other data sources – patentdata, publication data and DAMVAD Analytics proprietary data on public funded research and innovation processes. It generates a unique vast data set on the collaboration with specific companies during the last ten years.

    The data is applied in an advanced time series econometric model to  apply a quasi-experimental design. The aim is to generate the contra factual situation by using a relevant control group.

    We have previously conducted this kind of study for University of Copenhagen, Swedish Knowledge Foundation, Danish Technical University, Danish Ministry of Science, Technology and Innovation, Bergen University among others.