System and method for actionizing patient comments

1. A system for processing and actionizing patient experience data, the system comprising:

a server comprising a natural language processing (NLP) engine; and

a relational database;

wherein a plurality of communications is received at the server, each of the plurality of communications comprises comment data collected from publicly available data of an Internet web site or from one or more surveys, wherein the comment data comprises structured or unstructured patient experience data;

wherein the comment data from each of the plurality of communications is transformed to structured patient experience data and stored at the relational database in a response table that includes one or more records, wherein each record corresponds to the comment data from a communication, and wherein each record comprises the corresponding comment data and a timestamp;

wherein the comment data from each of the plurality of communications is parsed for individual phrases to generate a plurality of phrases;

wherein one or more phrases are selected from the plurality of phrases based on a predetermined parameter;

wherein the NLP engine is to predict one or more annotations for the one or more phrases based upon a score, wherein to predict the one or more annotations for the one or more phrases based upon the score comprises to (i) predict one or more annotations for the one or more phrases based upon a machine learning score, wherein the one or more annotations comprise a sentiment, a theme, or any named entity of the one or more phrases, (ii) determine whether the machine learning score is less than a predetermined threshold score, and (iii) predict the one or more annotations for the one or more phrases based upon a reference score in response to a determination that the machine learning score is less than the predetermined threshold score;

wherein the one or more annotations are stored at the relational database in an annotation table that includes one or more records in response to prediction of the one or more annotations, wherein each record corresponds to an annotation, and wherein each record includes the sentiment, the named entity, a primary tag indicative of a subject matter, or a secondary tag indicative of the theme; and

wherein the server is to generate a dashboard web page for a user that includes the one or more annotations in response to prediction of the one or more annotations.

2. The system according to claim 1 wherein the NLP engine predicts the sentiment of the phrase, wherein the sentiment comprises if the phrase is positive, negative or neutral.

3. The system according to claim 1 the NLP engine tags each phrase with a primary tag based on a subject matter of the phrase.

4. The system according to claim 1 further comprising an API for providing business intelligence for the user to act upon the inputted patient experience data.

5. The system according to claim 1 wherein the comment data of each of the plurality of communications is extracted at server, wherein the extracted comment data are transformed to a format compatible with a target, and wherein the transformed comment data are loaded into the response table of the relational database.

6. The system according to claim 1 wherein the comment data from each of the plurality of communications is parsed for individual phrases based upon punctuation or linguistic structure.

7. The system according to claim 1 wherein the plurality of communications is collected from publicly available data and uploaded from a hospital entity.

8. The system according to claim 1 wherein the server is configured to (i) determine whether a phrase is associated with a complete record in the annotations table, wherein the complete record comprises a sentiment, a primary tag, and a secondary tag, and (ii) generate an improvement action as a function of the phrase's annotation in response to a determination that the phrase is associated with a complete record.

9. The system of claim 1 , wherein to predict the one or more annotations for the one or more phrases based upon a reference score comprises to (i) store the one or more annotations at the relational database in a batching table that includes one or more records, wherein each record corresponds to an annotation, and wherein each record includes the sentiment, the named entity, the primary tag, the secondary tag, and an approval; and (ii) receive the approval in response to storing the one or more annotations at the relational database in the batching table.

10. The system of claim 1 , wherein the NLP engine is to predict the sentiment, the theme, and any named entity of the one or more phrases based upon the machine learning score in parallel.

11. A method according for processing and actionizing patient experience data, the method comprising:

receiving at a server a plurality of communications, each of the plurality of communications comprises comment data collected from publicly available data of an Internet web site or from one or more surveys, wherein the comment data comprises structured or unstructured patient experience data;

transforming the comment data of each of the plurality of communications to structured patient experience data;

storing the comment data of each of the plurality of communications at a relational database in a response table that includes one or more records, wherein each record corresponds to the comment data from a communication, and wherein each record comprises the corresponding comment data and a timestamp;

parsing the comment data for individual phrases to generate a plurality of phrases;

selecting one or more phrases from the plurality of phrases based on a predetermined parameter;

predicting at a NLP engine one or more annotations for the one or more phrases based upon a machine learning score wherein the one or more annotations comprise a sentiment, a theme, or any named entity of the one or more phrases;

determining whether the machine learning score is less than a predetermined threshold score;

predicting the one or more annotations for the one or more phrases based upon a reference score in response to determining that the machine learning score is less than the predetermined threshold score;

storing the one or more annotations at the relational database in an annotation table that includes one or more records in response to predicting the one or more annotations, wherein each record corresponds to an annotation, and wherein each record includes the sentiment, the named entity, a primary tag indicative of a subject matter, or a secondary tag indicative of the theme; and

generating at the server a dashboard web page for a user that includes the one or more annotations in response to predicting the one or more annotations.

12. The method according to claim 11 further comprising: extracting the comment data of each of the plurality of communications at the server;

transforming the comment data to a format compatible with a target; and

loading the transformed comment data into the response table of the relational database.

13. The method according to claim 11 further comprising tagging at the NLP engine each phrase with a primary tag based on a subject matter of the phrase.

14. The method according to claim 11 further comprising an API for providing business intelligence for the user to act upon the inputted patient experience data.

15. The method according to claim 11 wherein the comment data of each of the plurality of communications is extracted at server, wherein the extracted comment data are transformed to a format compatible with a target, and wherein the transformed comment data are loaded into the response table of the relational database.

16. The method according to claim 11 wherein the comment data from each of the plurality of communications is parsed for individual phrases based upon punctuation or linguistic structure.

17. The method according to claim 11 further comprising:

determining whether a phrase is associated with a complete record in the annotations table, wherein the complete record comprises a sentiment, a primary tag, and a secondary tag; and

generating an improvement action as a function of the phrase's annotation in response to determining that the phrase is associated with a complete record, and wherein if a named entity is detected in the phrase the improvement action is specific for the named entity.

18. The method according to claim 11 wherein predicting the sentiment of the phrase comprises predicting if the phrase is positive, negative or neutral.

19. The method of claim 11 , wherein predicting the one or more annotations for the one or more phrases based upon the reference score comprises predicting the one or more annotations based upon a manual review score.