Semantic call notes
Customer relationship management (“CRM”) implemented in a computer system, including parsing a word, from call notes of a conversation between a tele-agent of a call center and a customer representative, into a parsed triple of a description logic; determining whether the parsed triple is recorded in a semantic CRM triple store of the computer system; if the parsed triple is not recorded in the semantic CRM triple store, recording the parsed triple as a call note in the semantic CRM triple store.
1. A method of customer relationship management (“CRM”) implemented in a computer system, the method comprising:
generating, by a natural language processing (“NLP”) engine by a parsing process and in dependence upon a grammar, digitized words from call notes of a tele-agent of a call center;
identifying from the digitized words, a plurality of word candidates for inclusion in a triple, the triple including a subject, a predicate, and an object;
presenting, on a user-interface of the NLP engine, a list of the identified words selectable as a subject for a triple;
receiving, through the user-interface, information indicative of user-selection of a word from the list of the identified words;
generating, by the NLP engine via the parsing process, the triple by associating a particular predicate and a particular object with the word;
storing the triple as a call note in a semantic CRM triple store.
2. The method of claim 1 wherein the semantic CRM triple store comprises a set of triples of description logic comprising knowledge regarding the customer that is available to the tele-agent through the computer system.
3. The method of claim 1 wherein further comprising:
determining whether the triple is recorded in the semantic CRM triple store by asserting against the semantic CRM triple store a query comprising the triple.
4. The method of claim 1 further comprising:
executing by a query engine of the computer system against the semantic CRM triple store a semantic query for a definition of a word recorded in the semantic CRM triple store; and
displaying results of the semantic query.
5. A computer system that implements customer relationship management (“CRM”), the computer system comprising a computer processor operatively coupled to computer memory, the computer processor configured to function by:
generating, by a natural language processing (“NLP”) engine by a parsing process and in dependence upon a grammar, digitized words from call notes of a tele-agent of a call center;
identifying from the digitized words, a plurality of word candidates for inclusion in a triple, the triple including a subject, a predicate, and an object;
presenting, on a user-interface of the NLP engine, a list of the identified words selectable as a subject for a triple;
receiving, through the user-interface, information indicative of user-selection of a word from the list of the identified words;
generating, by the NLP engine via the parsing process, the triple by associating a particular predicate and a particular object with the word;
storing the triple as a call note in a semantic CRM triple store.
6. The computer system of claim 5 wherein the semantic CRM triple store comprises a set of triples of description logic comprising knowledge regarding the customer that is available to the tele-agent through the computer system.
7. The computer system of claim 5 further configured to determine whether the triple is recorded in the semantic CRM triple store by asserting against the semantic CRM triple store a query comprising the triple.
8. The computer system of 5 wherein the computer processor is further configured to:
execute by a query engine of the computer system against the semantic CRM triple store a semantic query for a definition of a word recorded in the semantic CRM triple store; and
display results of the semantic query.
9. The method of claim 1 , wherein generating the triple further comprises:
receiving a second user-selection associated with at least one of: the particular predicate or the particular object; and
generating the triple based at least on the second user-selection.