Repository matrix—explore entities across information types

Finding the most relevant people, companies or other entities in a given technology field is a task that many of our users face all the time.

Now, the ‘most relevant person’, for instance, may not necessarily be the person who has the most scientific publications or the most patents.  Rather, a person may be more relevant if they are active in science as well as in business, and not ‘just’ in one of these areas.

Our newly released repository matrix helps you see such cross-sectional relations, and zoom in on the most relevant entities.  It also helps you see differences.  For instance, the most talked-about science conferences in a field may be different from the most popular business-oriented conferences in the same technology field.

Here are two examples of how you could use the repository matrix:

Find relevant people

As mentioned above, ‘relevant people’ may be those that are active not just in science but also in business.  This is particularly important in advanced technology areas: on the one hand you need the science background in order to be able to build a business; on the other hand actually building a business also pushes the boundaries of your research because ‘real life applications’ confront you with new scientific challenges.

As an example, we looked at CRISPR, a recent gene editing method (https://en.wikipedia.org/wiki/CRISPR).  We simply searched for ‘CRISPR*’ (the wildcard allows for CRISPR/CAS, CRISPR-CAS, etc.), and selected the ‘repository matrix’ view.  Here is the result, sorted by ‘Company News’:

crispr-people

In the matrix you can see several people who are active across more than one type of information, i.e. Jennifer Doudna, Emmanuelle Charpentier (long names are truncated but you can see the full name if you move your mouse over the name), Feng Zhang, Rodolphe Barrangou, and George Church.

Compare conferences

For our second example, we looked at events in machine learning.  Events in mergeflow include conferences, trade fairs, etc..  Here is the result, sorted by ‘Investor News’:

machine-learning-conferences

In the matrix you can see that the events in business-oriented areas (Investor News; Financial Market News; Company News; Industry News) rank differently compared to more science-oriented areas (Funded Research Projects; Scientific Publications).  From this you might infer that if you are interested in machine learning, you might want to attend different conferences, depending on whether you would like to meet more business or more science types.  Or, if you like to meet a mixture of people, perhaps IJCAI is a good bet (http://ijcai-16.org/ or http://ijcai-17.org/).


Also published on Medium.