Machine learning of keywords

A computer program product comprising:

• a computer readable storage medium readable by one or more processors and storing instructions for execution by the one or more processors for performing a method comprising:
• progressively obtaining, by the one or more processors, textual content from a meeting hosting system, wherein the textual content is hosted on the meeting hosting system during a pre-defined interval;
• progressively parsing, by the one or more processors, the textual content to identify potential keywords from the textual content;
• controlling, by the one or more processors, utilizing an adaptive growth and self-pruning (AGSP) framework, a growth rate of a data structure, wherein the data structure facilitates interpretation of keywords, the controlling comprising:
• iteratively cognitively analyzing, by the one or more processors, the potential keywords to determine which potential keywords comprise seed keywords and to determine which existing seed keywords in a past iteration of data structure are no longer seed keywords, wherein the iteratively cognitively analyzing comprises determining that the seed keywords cannot be interpolated utilizing a dictionary, wherein the seed keywords are determined based on applying a learning rate function, wherein the learning rate function is a relationship between learning rate and maturity threshold of a keyword, wherein the maturity threshold corresponds to a minimum number of times that the keyword is referenced to make the keyword a seed keyword, wherein the learning rate corresponds to a pre-defined threshold of how frequently a given keyword should appear during the pre-defined interval for capture by an AGSP algorithm of the AGSP framework, wherein a seed keyword is an important concept for the one or more processors to learn and include in the data structure, wherein the iterative cognitive analysis of each potential keyword of the potential keywords is repeated a pre-defined number of times, and wherein the iteratively cognitively analyzing comprises generating and updating the data structure;
• outputting, by the one or more processors, based on completing the pre-defined number of times, the data structure comprising the seed keywords, wherein the outputting comprises pruning the existing seed keywords in the past iteration of the data structure which are no longer seed keywords from the data structure;
• determining, by the one or more processors, an interpretation for each seed keyword of the seed keywords, based on accessing one or more applications and one or more data sources via a connector of the AGSP framework, wherein the connector comprises an integration bus, wherein the one or more applications and one or more data sources are external to the meeting hosting system, wherein the connector is a single node that encapsulates interactions between the one or more processors and the one or more application and interactions between the one or more processors and the one or more data sources;
• and
• applying, by the one or more processors, the data structure to facilitate real-time interpretation of the keywords on the meeting hosting system during the pre-defined interval.

A system comprising:

• a memory;
• one or more processors in communication with the memory;
• program instructions executable by the one or more processors via the memory to perform a method, the method comprising:
• progressively obtaining, by the one or more processors, textual content from a meeting hosting system, wherein the textual content is hosted on the meeting hosting system during a pre-defined interval;
• progressively parsing, by the one or more processors, the textual content to identify potential keywords from the textual content;
• controlling, by the one or more processors, utilizing an adaptive growth and self-pruning (AGSP) framework, a growth rate of a data structure, wherein the data structure facilitates interpretation of keywords, the controlling comprising:
• iteratively cognitively analyzing, by the one or more processors, the potential keywords to determine which potential keywords comprise seed keywords and to determine which existing seed keywords in a past iteration of data structure are no longer seed keywords, wherein the iteratively cognitively analyzing comprises determining that the seed keywords cannot be interpolated utilizing a dictionary, wherein the seed keywords are determined based on applying a learning rate function, wherein the learning rate function is a relationship between learning rate and maturity threshold of a keyword, wherein the maturity threshold corresponds to a minimum number of times that the keyword is referenced to make the keyword a seed keyword, wherein the learning rate corresponds to a pre-defined threshold of how frequently a given keyword should appear during the pre-defined interval for capture by an AGSP algorithm of the AGSP framework, wherein a seed keyword is an important concept for the one or more processors to learn and include in the data structure, wherein the iterative cognitive analysis of each potential keyword of the potential keywords is repeated a pre-defined number of times, and wherein the iteratively cognitively analyzing comprises generating and updating the data structure;
• outputting, by the one or more processors, based on completing the pre-defined number of times, the data structure comprising the seed keywords, wherein the outputting comprises pruning the existing seed keywords in the past iteration of the data structure which are no longer seed keywords from the data structure;
• determining, by the one or more processors, an interpretation for each seed keyword of the seed keywords, based on accessing one or more applications and one or more data sources via a connector of the AGSP framework, wherein the connector comprises an integration bus, wherein the one or more applications and one or more data sources are external to the meeting hosting system, wherein the connector is a single node that encapsulates interactions between the one or more processors and the one or more application and interactions between the one or more processors and the one or more data sources;
• and
• applying, by the one or more processors, the data structure to facilitate real-time interpretation of the keywords on the meeting hosting system during the pre-defined interval.