# US 2019/0317957 A1 · Weather-based natural language text processing

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0317957 A1
- Document ID: US20190317957A1
- Status: Published application
- Kind: A1
- Application number: 16/456,549
- Filed: 2019-06-28
- Published: 2019-10-17
- Earliest claimed: 2017-05-30
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Shilpi Ahuja; Sheng Hua Bao; Rashmi Gangadharaiah
- Web page: https://patentgrep.com/patent/US20190317957A1

## CPC

- G06F 16/345: {Summarisation for human users}
- G06F 40/258: Heading extraction; Automatic titling; Numbering
- G06F 40/30: Semantic analysis
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients
- G16H 70/00: ICT specially adapted for the handling or processing of medical references
- G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Relationship extraction between descriptors in one or more lists of weather condition descriptors, and adverse event descriptors within unstructured data sources using natural language processing. Medical condition descriptor may be a descriptor that may be used to further extract relationships between weather condition descriptors and adverse event descriptors. A data object is generated, according to a data model, based on the extracted relationships between the descriptors. A set of candidate unstructured documents containing the extracted relationship between the descriptors is retrieved and filtered by selecting unstructured documents that include a precautionary measure descriptor. The filtered precautionary measure descriptors are presented to a user in a summarized message to a user device.

## Claims

### Claim 1

A method for performing electronic natural language processing on unstructured data, comprising:
extracting relationships between at least three descriptors stored in one or more lists of weather condition descriptors, adverse event descriptors, and medical condition descriptors, wherein the at least three descriptors have a corresponding extracted relationship,
wherein a weather condition descriptor comprises a span of electronic text characters associated with an atmospheric condition,
wherein an adverse event descriptor comprises a span of electronic text characters associated with an occurrence, and
wherein a medical condition descriptor comprises a span of electronic text characters associated with an illness or abnormality;
generating a candidate data object comprising descriptors identified as related;
presenting the candidate data object to a user for identification of known relationships;
receiving an input from the user identifying the candidate data object as comprising descriptors having a known relationship;
filtering from the extracted relationships the user-identified candidate data object comprising descriptors having a known relationship;
generating a data object according to a data model, based on the extracted relationships between the descriptors,
wherein generating the data object according to the data model excludes the filtered candidate data object;
retrieving a set of candidate unstructured documents comprising instances of the adverse event descriptor and a related weather condition descriptor;
filtering the retrieved set of candidate unstructured documents by selecting candidate unstructured documents that comprise a precautionary measure descriptor,
wherein a precautionary measure descriptor comprises a span of electronic text characters associated with an action relative to the adverse event descriptor;
receiving the weather condition descriptor and the medical condition descriptor associated with the user;
determining, based on the generated data object, the adverse event descriptor and the precautionary measure descriptor associated with the received weather condition descriptor and medical condition descriptor;
ranking the relationships between the weather condition descriptor, the adverse event descriptor, and the precautionary measure descriptor based on the received medical condition descriptor of the user; and
generating a message to the user based on the ranked relationships between the weather condition descriptor, the adverse event descriptor, the precautionary measure descriptor, and the received medical condition descriptor of the user.
