Medical processing systems and methods

A system for natural language processing of Source data from a plurality of sources via a network, the system comprising:

• processor hardware;
• and
• memory hardware coupled to the processor hardware,
• wherein the memory hardware stores instructions that, upon execution, cause processor hardware to:
• obtain source data from a plurality of data sources via the network, wherein:
• the source data includes two or more selected from (a) feedback concerning at least one service, product, or experience, (b) an employee complaint, (c) a telephone call transcript, (d) email correspondence, (e) text messaging, (f) a chat session, (g) a scan of hardcopy correspondence, (h) a social media posting, (i) electronic correspondence, (j) phone call data, (k) website feedback, (l) a website user ranking, (m) patient satisfaction data, or (n) a survey, and
• the plurality of data sources includes two or more selected from a pharmacy benefit manager, a client support center, a customer call center, a message center, a mobile application store, an email server, a text messaging server, a chat room server, and a social media website server,
• generate tagged source data by tagging a subset of terms in the source data, wherein the subset of terms excludes confidential information,
• perform a data substitution in the tagged source data by applying an ontology to the tagged source data, wherein:
• apply the ontology including removing one or more non-alphanumeric characters from the tagged source data, and
• the removing ignores the subset of terms,
• detect at least one instance of confidential information among the tagged source data based on an access level of a user associated with a user device, wherein the detecting ignores the subset of terms,
• transform the tagged source data into modified data by replacing the at least one instance of confidential information in the modified data with at least one contextual placeholder, wherein the transforming ignores the subset of terms,
• identify a set of terms and a set of phrases in the modified data based on the ontology,
• for each term of the set of terms, tabulate a corresponding count indicating a frequency of appearance in the modified data,
• identify a first term,
• identify a first phrase including the first term, wherein the first phrase is in a first level of the ontology,
• tabulate a first count indicating a frequency of appearance of the first phrase in the modified data,
• identify a second phrase in the set of phrases including the first term, wherein the second phrase is in a second level of the ontology,
• tabulate a second count indicating a frequency of appearance of the second pase in the modified data,
• reduce a first term count corresponding to the first term by the first count;
• reduce the first term count corresponding to the first term by the second count
• generate, based on the ontology applied, a user interface respectively reflecting each term of the set of terms;
• transform the user interface by adjusting a feature of each term of the set of terms to indicate a corresponding count of the corresponding term;
• and
• in response to receiving user input, modify the user interface to remove a portion of terms of the set of terms based on a parameter of the user input.

A method of natural language processing of source data including confidential patent health information from a plurality of data sources via a network, the method comprising:

• obtaining, from the plurality of data sources via the network, the source data, wherein:
• the source data includes two or more selected from (a) feedback concerning at least one service, product, or experience, (b) an employee complaint, (c) a telephone call transcript, (d) email correspondence, (e) text messaging, (f) a chat session, (g) a scan of hardcopy correspondence, (h) a social media posting, (i) electronic correspondence, (j) phone call data, (k) website feedback, (l) a website user ranking, (m) patient satisfaction data, an (n) a survey, and
• the plurality of data sources includes two or more selected from a pharmacy benefit manager, a client support center, a customer call center, a message center, a mobile application store, an email server, a text messaging server, a chat room server, and a social media website server;
• generating tagged source data by tagging a subset of terms in the source data, wherein the subset of terms excludes confidential information,
• performing data substitution in the tagged source data by applying an ontology to the source data, wherein:
• applying the ontology includes removing one or more non-alphanumeric characters from the source data, and
• the removing ignores the subset of terms;
• detecting at least one instance of confidential information among the tagged source data based on an access level of a user associated with a user device, wherein the detecting ignores the subset of terms;
• transforming the tagged source data into modified data by replacing the at least one instance of confidential information in the modified data with at least one contextual placeholder, wherein the transforming ignores the subset of terms;
• identifying a set of terms and a set of phrases in the modified data based on the ontology;
• for each term of the set of terms, tabulating a corresponding count indicating a frequency of appearance in the modified data;
• identifying a first term;
• identifying a first phrase in the set of phrase including the first term,
• tabulating a first count indicating a frequency of appearance of the first phrase in the modified data;
• identifying a second phrase in the set of phrases including the first term,
• tabulating a second count indicating a frequency of appearance of the second phrase in the modified data;
• modifying a first term count by subtracting the first count and the second count, the second phrase including more terms than the first phrase, the first phrase appearing in the second phrase, and the first term appearing within the first phrase and the second phrase;
• and
• generating, a user interface containing a word cloud respectively reflecting each term of the set of terms and corresponding count.