The Keygroups Report

Introducing the higher concepts based on the keywords

After the introduction of the Publications Report and Publisher Networks, it is now time for another report: the Keygroups Report. In the platform I have developed the keygroups represent higher concepts that are based on the keywords. The readers of my previous posts would remember that keywords are generated 95% automatically from the press releases (the system actually creates a lot keywords and some need to be eliminated, which is currently done manually, hence it is not 100% automatic). The keygroups, however, are created only manually, using software, since those concepts are actually groupings that make sense only together for a reader.

There are a total of 721 press releases caught by the system so far from a pool of less than 50 companies (due to the ongoing development process) since May 15, 2022.

Here is the Keygroups Report created on September 4, 2022:

Many of the keygroups you see above are self-explanatory, but some need an explanation, such as “Disease/Disorder/Syndrome”, the keygroup most referred (402 times) in the press releases. This keygroup represents the concept of the target of a drug or therapy, that is, it is the representation of anything for which a drug or therapy can be developed. Note that the “referral” here is not a direct mentioning of the concept, but the keygroup is created indirectly using the keywords, which are lower level concepts for the platform.

“Therapy/Treatment/Care”, referred 228 times, is also another broad keygroup that needs explanation as it unifies all concept of any treatment, irrespective of whether a drug is used or not for the treatment.

Another broad keygroup is “Organizational/Corporate Action”, which represents any decision by an organization. For example, when a company announces an acquisition, such an action is assigned to this keygroup. Similarly, the announcement of a capital increase, spin-off transaction or corporate restructuring is covered by this keygroup as well as the announcement of a business update, a conference call or donation.

The second keygroup most referred (383 times) is the “Medicinal Drug”, which does not include the keygroup of “Vaccine”. The keygroup “Vaccine” is referred a total of 60 times. Taken together these concepts are the most referred ones.

For those interested: there are currently a total of 2926 keywords in the system while the number of keygroups is only 49. The keywords are created from noun phrases found in sentences. As such they include all possible nouns and noun phrases, including  the names of drugs, medicines, diseases, disorders, corporate actions, etc., as well as the names of organizations, locations, countries and persons. To provide just some examples, the system has both the single word “Tremelimumab”, a specific drug, and the complex phrase “Metastatic Non-Small Cell Lung Cancer”, a specific form of lung cancer, as keywords. The total number of keywords assigned to the keygroup “Medicinal Drug” is currently 454 while the number of keywords for the keygroup “Disease/Disorder/Syndrome” is 659.