Answer sequence discovery and generation

Patent metadata

Field	Value
Publication number	US-2018322121-A1
Application number	US-201816035813-A
Country	US
Kind code	A1
Filing date	Jul 16, 2018
Priority date	Nov 5, 2014
Publication date	Nov 8, 2018
Grant date	—

Abstract

Aspects of the present disclosure are directed toward discovering and generating answer sequences. Aspects are directed toward parsing, by a natural language processing technique configured to analyze syntactic and semantic content, a corpus of data for a subject matter. Aspects are also directed toward detecting a first set of answers including a first answer corresponding to a first answer category and a second set of answers including a second answer corresponding to a second answer category. Both the first and the second answer categories may relate to the subject matter. Aspects are also directed toward identifying a first set of ordering data for the first set of answers and the second set of answers. Aspects are also directed toward determining a first answer sequence corresponding to an order of the first set of answers and the second set of answers.

First claim

What is claimed is:

1 . A method for treating a human patient having a particular disease, the method comprising:
providing medical history information about the patient a processor of medical question-answering system; parsing, by the processor using a natural language processing technique configured to analyze syntactic and semantic content, the medical history information; performing, based on the parsed medical history information, a first search of a medical corpus containing a plurality of sources having information about the particular disease; detecting, by the processor and based on the first search, a group of medical treatments relevant to the particular disease, wherein the detected medical treatments are identified from different sources of the plurality of sources from each other; analyzing, by the processor and after the group of medical treatments have been detected, similarities and difference among the detected medical treatments; sorting, by the processor and based on the analysis, the detected medical treatments into at least three treatment categories, including sorting a first plurality of the detected medical treatments into a first treatment category, a second plurality of the detected medical treatments into a second treatment category, and a third plurality of the detected medical treatments into a third treatment category, wherein each treatment category includes medical treatments that are similar in type to each other as determined based on the analysis; performing, by the processor and based on the sorting of the detected medical treatments, a second search of the medical corpus; identifying, by the processor and based on the second search, ordering data indicating a particular arrangement of a first medical treatment of the first treatment category, a second medical treatment of the second treatment category, and a third medical treatment of the third treatment category relative to each other, wherein the ordering data is identified based on a particular at least one source that is different from any of the sources from which the first medical treatment, second medical treatment, and third medical treatment were identified based on the first search; generating, by the processor and based on the ordering data, a medical treatment plan for the patient that includes a combination of the first medical treatment, the second medical treatment, and the third medical treatment ordered within the medical treatment plan as indicated by the ordering data; and treating the patient for the particular disease with the first medical treatment, the second medical treatment, and the third medical treatment in an order dictated by the medical treatment plan.

2 . A computer program product for treating a human patient having a particular disease, the computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed on a processor, causes the processor to perform a method comprising:
receiving, by the processor which is part of a medical question-answering system, medical history information about the patient; parsing, by the processor using a natural language processing technique configured to analyze syntactic and semantic content, the medical history information; performing, based on the parsed medical history information, a first search of a medical corpus containing a plurality of sources having information about the particular disease; detecting, by the processor and based on the first search, a group of medical treatments relevant to the particular disease, wherein the detected medical treatments are identified from different sources of the plurality of sources from each other; analyzing, by the processor and after the group of medical treatments have been detected, similarities and difference among the detected medical treatments; sorting, by the processor and based on the analysis, the detected medical treatments into at least three treatment categories, including sorting a first plurality of the detected medical treatments into a first treatment category, a second plurality of the detected medical treatments into a second treatment category, and a third plurality of the detected medical treatments into a third treatment category, wherein each treatment category includes medical treatments that are similar in type to each other as determined based on the analysis; performing, by the processor and based on the sorting of the detected medical treatments, a second search of the medical corpus; identifying, by the processor and based on the second search, ordering data indicating a particular arrangement of a first medical treatment of the first treatment category, a second medical treatment of the second treatment category, and a third medical treatment of the third treatment category relative to each other, wherein the ordering data is identified based on a particular at least one source that is different from any of the sources from which the first medical treatment, second medical treatment, and third medical treatment were identified based on the first search; and generating, by the processor and based on the ordering data, a medical treatment plan for the patient that includes a combination of the first medical treatment, the second medical treatment, and the third medical treatment ordered within the medical treatment plan as indicated by the ordering data, wherein the patient is treated for the particular disease with the first medical treatment, the second medical treatment, and the third medical treatment in an order dictated by the medical treatment plan.

3 . A computer system for treating a human patient having a particular disease, the computer system comprising:
a memory; and a processor coupled to the memory, the processor configured to obtain instructions from the memory that cause the processor to perform a method comprising: receiving medical history information about the patient; parsing, using a natural language processing technique configured to analyze syntactic and semantic content, the medical history information; performing, based on the parsed medical history information, a first search of a medical corpus containing a plurality of sources having information about the particular disease; detecting, based on the first search, a group of medical treatments relevant to the particular disease, wherein the detected medical treatments are identified from different sources of the plurality of sources from each other; analyzing, after the group of medical treatments have been detected, similarities and difference among the detected medical treatments; sorting, based on the analysis, the detected medical treatments into at least three treatment categories, including sorting a first plurality of the detected medical treatments into a first treatment category, a second plurality of the detected medical treatments into a second treatment category, and a third plurality of the detected medical treatments into a third treatment category, wherein each treatment category includes medical treatments that are similar in type to each other as determined based on the analysis; performing, based on the sorting of the detected medical treatments, a second search of the medical corpus; identifying, based on the second search, ordering data indicating a particular arrangement of a first medical treatment of the first treatment category, a second medical treatment of the second treatment category, and a third medical treatment of the third treatment category relative to each other, wherein the ordering data is identified based on a particular at least one source that is different from any of the sources from which the first medical treatment, second medical treatment, and third medical treatment were ident

Assignees

• IBM

Inventors

• Barker Kevin S
• Eggebraaten Thomas J
• Megerian Mark G
• Petri John E

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G16H10/20

  for electronic clinical trials or questionnaires · CPC title

• G06F16/24522 Primary

  Translation of natural language queries to structured queries · CPC title

• G06F16/955

  using information identifiers, e.g. uniform resource locators [URL] · CPC title

• G06F16/24578

  using ranking · CPC title