Airline assessment feedback recommendation and lookup result mapping using artificial intelligence

Abstract

The invention discloses airline assessment feedback recommendation and lookup result mapping using artificial intelligence. Airline assessment, comment recommendations, and lookup result prediction mappings are provided. In response to evaluating a first user portion entry in the category entry field, a first Large Language Model (LLM) provides a cue of a suggested category. A user entry of a selection or alternative category of one of the prompts is received. In response to a partial entry of the task description, the first LLM provides a prompt suggesting the task description. User input of a selection or alternative task description of one of the prompts is received. In response to a partial entry of the comment, the first LLM provides a cue that includes a subset of words of the suggested comment. An input of a selection or alternative comment of one of the cues is then received. The second LLM provides predicted lookup results based on the assessment category, the task description, and the comments. A user input of a selection or alternative lookup result of one of the predicted lookup results is then received.

Description

Technical Field

[0001] The present disclosure relates generally to artificial intelligence systems and, more particularly, to large-scale language models for recommending and mapping user inputs regarding results for airline operations evaluation. Background Art

[0002] Airline Operational Safety Support (AOSS) is an initiative designed to improve operators' effectiveness in identifying and evaluating their preparedness for sustained safe aircraft operations. Under this initiative, an assessment of the airline's own processes, procedures, and procedures is performed in partnership with individual airlines. This assesses the airline's ability to sustainably operate and maintain its aircraft fleet at a safe, effective, and airworthy level, and identifies individual needs for further improvement.

[0003] As part of this process, the safety team defines their own reviews for individual items based on the assessment checklist descriptions, ratings, and more. These reviews detail the safety team pilots' understanding and observations of the operator's efficiency, airline processes, procedures, and safety risks. The reviews also identify the types of findings (hazards) that could occur in the near future if corrective action is not taken. Finding categorization helps the AOSS team identify the root cause of the safety issue and a mitigation plan. Summary of the Invention

[0004] An illustrative embodiment provides a computer-implemented method for mapping airline reviews, review recommendations, and predictive search results. The method includes, in response to a first partial user input in an evaluation category input field, providing, by a first large language model, a first plurality of prompts of suggested evaluation categories based on a plurality of predefined evaluation categories. User input is received for selecting or replacing one of the first plurality of prompts. In response to a second partial user input in a task description input field, the first large language model provides a second plurality of prompts of suggested complete task descriptions based on historical content related to the evaluation categories. User input is received for selecting or replacing one of the second plurality of prompts of the complete task description. In response to a third partial user input in a review input field, the first large language model provides a third prompt comprising an initial word subset of a suggested review based on historical content related to the task description. User input is then received for selecting or replacing one of the third prompts of a review. The second large language model provides a plurality of predicted search results based on the evaluation categories, the task description, and the reviews. User input is received for selecting or replacing one of the predicted search results.

[0005] Another illustrative embodiment provides a system for airline review, review recommendation, and search result prediction mapping. The system includes a storage device storing program instructions and one or more processors operably connected to the storage device and configured to execute the program instructions to cause the system to: in response to a first partial user input in an evaluation category input field, provide, by a first large language model, a first plurality of prompts of suggested evaluation categories based on a plurality of predefined evaluation categories; receive user input selecting or replacing the evaluation category for one of the first plurality of prompts; in response to a second partial user input in a task description input field, provide, by the first large language model, a second plurality of prompts of suggested complete task descriptions based on historical content associated with the evaluation categories; receive user input selecting or replacing the complete task description for one of the second plurality of prompts; in response to a third partial user input in a review input field, provide, by the first large language model, a third prompt including an initial word subset of a suggested review based on historical content associated with the task description; provide, by the second large language model, a plurality of predicted search results based on the evaluation category, the task description, and the review; and receive user input selecting or replacing one of the predicted search results.

[0006] Another illustrative embodiment provides a computer program product for airline review, review recommendation, and search result prediction mapping. The computer program product includes a computer-readable storage medium having program instructions embodied thereon for performing the following operations: in response to a first partial user entry in an evaluation category entry field, providing, by a first large language model, a first plurality of prompts of suggested evaluation categories based on a plurality of predefined evaluation categories; receiving user input selecting or replacing the evaluation category for one of the first plurality of prompts; in response to a second partial user entry in a task description entry field, providing, by the first large language model, a second plurality of prompts of suggested complete task descriptions based on historical content associated with the evaluation categories; receiving user input selecting or replacing the complete task description for one of the second plurality of prompts; in response to a third partial user entry in a review entry field, providing, by the first large language model, a third prompt including an initial word subset of suggested reviews based on historical content associated with the task description; receiving user input selecting or replacing the review for one of the third prompts; providing, by the second large language model, a plurality of predicted search results based on the evaluation categories, the task description, and the reviews; and receiving user input selecting or replacing the search result for one of the predicted search results.

[0007] The features and functions can be achieved independently in various embodiments of the present disclosure or may be combined in yet other embodiments, further details of which can be seen with reference to the following description and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The novel features which are believed to be characteristic of the illustrative embodiments are set forth in the appended claims. However, the illustrative embodiments together with their preferred modes of use, further objects and features will be best understood by reference to the following detailed description of illustrative embodiments of the disclosure when read in conjunction with the accompanying drawings, in which:

[0009] Figure 1 depicts a block diagram of an airline rating system according to an illustrative embodiment;

[0010] Figure 2 depicts a schematic diagram illustrating artificial intelligence architecture assessment review recommendation in accordance with an illustrative embodiment;

[0011] Figure 3 depicts a schematic diagram illustrating a review recommendation engine in accordance with an illustrative embodiment;

[0012] Figure 4A depicts an illustration of a user interface with a blank comment field in accordance with an illustrative embodiment;

[0013] Figure 4B depicts an illustration of a user interface with partial comment entry and suggested comments in accordance with an illustrative embodiment;

[0014] Figure 4C depicts an illustration of a user interface having suggested search results responsive to an entered review in accordance with an illustrative embodiment;

[0015] Figure 5 depicts a flow chart illustrating a process for airline evaluation, review recommendation, and search result prediction mapping in accordance with an illustrative embodiment; and

[0016] Figure 6 is an illustration of a block diagram of a data processing system in accordance with an illustrative embodiment. DETAILED DESCRIPTION

[0017] The illustrative embodiments recognize and take into account that the information collected by the security team members is manually written in an Excel spreadsheet with comments / reviews, and that the average word count for each feedback comment is approximately 50-60 words. This takes approximately 2-4 minutes per observation item, and there is a risk of human error such as spelling errors or punctuation errors, as well as the possibility of missing feedback.

[0018] The illustrative embodiments also recognize and take into account that once the comments / comments are manually entered for all observations, the AOSS team begins reading the observations, their comments, and begins writing the finding (hazard) type for each observation. This is a manual activity and may map non-standardized finding results. Mapping approximately 400 observations takes a significant amount of time.

[0019] The illustrative embodiments provide an artificial intelligence system that uses a trained GPT algorithm to predict and prompt suggested comments / review wording in response to a user inputting an initial word in an interface. The algorithm is trained on historical data observations and learns patterns from the historical data, finding relationships between words / symbols present in aviation safety data. A second GPT algorithm is trained to map comments / reviews to lookup results / potential hazards based on the historical data.

[0020] Figure 1 1 is a block diagram of an airline rating system according to an illustrative embodiment. The airline rating system 100 includes a web-based front-end application 102 having a user interface 104.

[0021] The user interface 104 includes several entry fields in which the user can enter key data. These include an assessment category entry field 106, a task description entry field 108, a comment entry field 110, and a search result entry field 112. In response to a user typing a few (2-3) words into the assessment category entry field 106, the task description entry field 108, or the comment entry field 110, the web-based front-end application 102 makes corresponding application programming interface (API) calls 114 to the Python / JAVA back-end 122 (see Figure 3 ). The API calls 114 include a checklist data population API call 116 , a review prediction API call 118 , and a search result prediction API call 120 .

[0022] The first large language model 124 is triggered by the API call 114 of the Python / JAVA backend 122, which generates suggested entries to be displayed in the user interface 104 to help the user complete the entry (see Figure 4B ). These recommendations are based on predefined evaluation categories 130, historical task description data 132, and historical review data 134 stored in a database 128, on which the first large language model 124 is trained.

[0023] In response to the entry in the comment entry field 110, the second large language model 126 generates a number of suggested search results by mapping the comment to the historical risk classification data 136. These suggested search results are displayed in the user interface 104 for selection and entry into the search result entry field 112 (see Figure 4C ).

[0024] Airline evaluation system 100 can be implemented using software, hardware, firmware, or a combination thereof. When software is used, the operations performed by airline evaluation system 100 can be implemented in program code configured to run on hardware (such as a processor unit). When firmware is used, the operations performed by airline evaluation system 100 can be implemented in program code and data and stored in persistent memory for execution on a processor unit. When hardware is used, the hardware can include circuitry that operates to perform the operations in airline evaluation system 100.

[0025] In the illustrative examples, the hardware can take the form of at least one selected from a circuit system, an integrated circuit, an application specific integrated circuit (ASIC), a programmable logic device, or some other suitable type of hardware configured to perform several operations. Using a programmable logic device, the device can be configured to perform several operations. The device can be reconfigured at a later time, or can be permanently configured to perform several operations. Programmable logic devices include, for example, programmable logic arrays, programmable array logic, field programmable logic arrays, field programmable gate arrays, and other suitable hardware devices. In addition, these processes can be implemented in organic components integrated with inorganic components, and can be composed entirely of organic components, excluding humans. For example, these processes can be implemented as circuits in organic semiconductors.

[0026] Computer system 150 is a physical hardware system that includes one or more data processing systems. When more than one data processing system is present in computer system 150, these data processing systems communicate with each other using a communication medium. The communication medium may be a network. The data processing system may be selected from at least one of a computer, a server computer, a mobile device (such as a tablet computer), or some other suitable data processing system.

[0027] As depicted, computer system 150 includes several processor units 152 that are capable of executing program code 154 that implements the processes in the illustrative example. As used herein, a processor unit in several processor units 152 is a hardware device and is composed of hardware circuits, such as those on an integrated circuit, that respond to and process instructions and program codes that operate a computer. When several processor units 152 execute program code 154 for a process, several processor units 152 are one or more processor units that can be on the same computer or on different computers. In other words, the process can be distributed between processor units on the same or different computers in the computer system. In addition, several processor units 152 can be processor units of the same type or different types. For example, several processor units can be selected from at least one of a single-core processor, a dual-core processor, a multi-processor core, a general-purpose central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), or some other type of processor unit.

[0028] The illustrative embodiments provide an automated AI-based tool that recommends suggested comments for safety team members to choose from and recommends the type of findings (hazards) to look for when they begin writing comments for individual checklist items. These recommendations can save significant time and reduce the occurrence of human errors such as spelling errors or punctuation errors, and avoid missed feedback points.

[0029] This solution uses a Generative Pre-Trained Transformer (GPT) algorithm to recommend descriptive comments and map search results for individual observations. GPT is a powerful natural language processing (NLP) algorithm that learns patterns in text data and uses these patterns to write sentences, summarize text content, Q&A, and categorize. The illustrative embodiment uses two pre-trained GPT algorithms with over 2 billion parameters. The first algorithm is fine-tuned using approximately 1,500 AOSS historical data observations (each observation is a small paragraph). The first algorithm learns patterns from historical data and finds relationships between words / symbols present in AOSS aviation safety data. When security team members begin typing their observations, these patterns help the algorithm recommend comments in the future. The second GPT pre-trained algorithm is fine-tuned using historical data with an additional feature "comments" to predict search results / hazards related to the observations.

[0030] The AOSS team has approximately 400 to 500 predefined checklists / tasks that are given to the airline during the assessment. For each task, the assessor observes the airline's performance and provides comments for each task performed. After the assessment is completed, the AOSS team begins mapping (classifying) findings / hazards for individuals based on the given comments for that task. Comments can be single sentences or small paragraphs. Manually writing comments can take a considerable amount of time and may contain errors such as missing comment points and spelling errors. Therefore, the illustrative embodiment automates comment writing by recommending comment wording to the assessor during the assessment and mapping of findings / hazards.

[0031] Figure 2 A schematic diagram illustrating an artificial intelligence architecture evaluation review recommendation according to an illustrative embodiment is depicted. Architecture 200 is Figure 1 An example implementation of the airline rating system 100 is provided.

[0032] The trained GPT model 1 202 is a GPT-based algorithm that is trained on historical review data entered in the past. The algorithm maps the relationship between the area of ​​evaluation, task description, and initial review input from the evaluation and the historically captured reviews. When the evaluator starts writing a few words (e.g., 3-4) for the review field 210, the trained GPT model 1 202 takes these initial inputs and finds the best historical reviews that are similar to the given input context and suggests these reviews to the evaluator (see Figure 4B ). Comments 214 are example suggested comments that the trained GPT model 1 202 might find in response to a user's partial word entry. The user can select the best recommended comment or enter an alternative comment.

[0033] Once a review is selected and entered, a trained GPT model 2 204 (another GPT-based algorithm trained on historical reviews and hazards) takes the input (the assessed area 206, the task description 208, and the AI-recommended feedback / review 210) and classifies the review as a type of hazard / finding, such as a finding 212 associated with an element contained in the review.

[0034] Trained GPT model 1 202 uses a transformer architecture to train on text data. It takes three inputs from historical data: checklists / assessments 206, task descriptions 208, and feedback / comments 210. The algorithm attempts to find relationships between words in a sentence or paragraph and to find the meaning of words based on their position in the sentence, which helps the algorithm find the contextual meaning of the sentence or paragraph.

[0035] Trained GPT model 2 204 also uses a transformer architecture. This model is trained on historical data, such as checklists / evaluations 206, task descriptions 208, feedback / reviews 210, and hazard / finding results, such as 212. It attempts to map the context of sentences / paragraphs to hazard / finding result types. Once a relationship between reviews and hazard types is found, this relationship is applied to new data to map hazard / finding results.

[0036] Figure 3 Depicted is a schematic diagram illustrating a review recommendation engine in accordance with an illustrative embodiment. Figure 3 Describes a high-level architectural diagram of a web application, where the front-end application will communicate with the back-end tier through multiple API calls.

[0037] The illustrative embodiment provides a web-based responsive application that can be hosted in a cloud computing system. The front-end interface 310 can use the Angular 15+ framework and interact with the Python / Java-based back-end 316 through several APIs that communicate with two GPT machine language models 312 and 314.

[0038] The HTTP API call 302 that loads filter information brings all the checklist information from the database 308 through the backend 316. The database 308 may include an Azure Structured Query Language (SQL) database. The information is displayed in a tabular format in the frontend interface 310, where each row represents an assessment item (see Figure 4A ). Users can edit this information to write comments and map findings / hazards.

[0039] The comment field for each assessment item includes an editable text box (see Figure 4B Once the user starts writing a review of 2-3 words, a review prediction API call 304 occurs, which feeds all the information of the evaluation item back to the review recommendation ML model 312 (equivalent to Figure 2 ), and brings back a list of suggested comments to the front-end interface 310.

[0040] After selecting a comment from the list of suggested comments, or entering a replacement comment manually entered by the user, a search result prediction API call 306 occurs, which retrieves all information including the selected / entered comment. The search result prediction API call 306 feeds this information to the search result prediction ML model 314 (equivalent to Figure 2 204 in the model) and retrieve the assigned search results for the assessment item.

[0041] Figures 4A-4CA front-end user interface is depicted according to an illustrative embodiment. In this example, front-end user interface 400 includes a table-formatted row item that divides item categories into different columns, including search results 402 , ratings 404 , checklists 406 , aircraft models 408 , and reviews 410 .

[0042] exist Figure 4A In the example shown, the comment entry field 412 is initially empty. Each row has a corresponding comment entry field that operates as a smart search box.

[0043] Once the user starts writing a partial entry 414 (e.g., 2-3 words) in the review entry field 412, the review recommendation ML model 312 will assess all other parameters such as rating 404, checklist 406, and aircraft model 408 and will send back a list of the most appropriate recommended reviews 416, 418 for the item. Figure 4B As shown, these recommended reviews 416, 418 are displayed as selectable options under the review entry field 412. The user will then be able to select the appropriate review simply by clicking on the correct option. Alternatively, if none of the recommended reviews sufficiently matches what the user wishes to convey, the user can manually enter the original review, which can be used to further train the review recommendation ML model 312.

[0044] Once the user selects the appropriate review, such as 416, and saves it, another API call is made to obtain the possible search results mapped from the search result prediction ML model 314. Figure 4C As shown, the predicted search result 420 is displayed in the search result field 402, opposite the evaluation item.

[0045] Figure 5 A flow chart illustrating a process for airline evaluation, review recommendation, and search result prediction mapping according to an illustrative embodiment is depicted. Process 500 may be performed at Figure 1 The airline evaluation system 100 is implemented in FIG.

[0046] In response to a first user partial entry in an evaluation category entry field, a first large language model provides a first plurality of prompts of suggested evaluation categories based on a plurality of predefined evaluation categories (operation 502). The predefined evaluation categories include flight operations, simulator operations, training, safety, crew dispatch, and crew scheduling.

[0047] The first large-scale language model includes a Generative Pre-Trained Transformer (GPT) algorithm that uses evaluation categories, task descriptions, and reviews to find relationships between words in a sentence or paragraph and to determine the meaning of words based on their position in the sentence to find the contextual meaning of the sentence or paragraph. The GPT algorithm is integrated with a web-based front-end application using an interface that provides a table format with several smart search fields that trigger front-end API calls in response to the entry of several words.

[0048] The system receives user entry of a selection or an alternate evaluation category for one of the first number of prompts (operation 504 ).

[0049] In response to a second user partial entry in the task description entry field, the first large language model provides a second number of hints for suggested complete task descriptions based on historical content related to the evaluation category (operation 506 ).

[0050] The system then receives user input selecting one of the second number of prompts or an alternative complete task description (operation 508 ).

[0051] In response to a third user partial entry in the feedback entry field, the first large language model provides a third prompt including an initial word subset (3 to 4 words) of a suggested comment based on historical content related to the task description (operation 510).

[0052] The system receives user input of a selection of one of the third prompts or an alternative comment (operation 512). The user entry of an alternative evaluation category, an alternative complete task description, or an alternative comment is used to retrain and update the first large language model.

[0053] The second large language model provides a number of predicted search results based on the evaluation categories, task descriptions, and reviews (operation 514). The second large language model includes a Generative Pre-Trained Transformer (GPT) algorithm that uses the evaluation categories, task descriptions, and reviews from the first large language model, or user-entered reviews, to map the context of sentences and paragraphs to the types of search results.

[0054] The system receives a user input of a selection of one of the predicted search results or an alternative search result (operation 516). The user input of the alternative search result is used to retrain and update the second large language model. Then, process 500 ends.

[0055] Now go to Figure 6 , depicts an illustration of a block diagram of a data processing system according to an illustrative embodiment. Data processing system 600 may be used to implement Figure 11. In this illustrative example, data processing system 600 includes communications framework 602, which provides communications between processor unit 604, memory 606, persistent storage 608, communications unit 610, input / output (I / O) unit 612, and display 614. In this example, communications framework 602 takes the form of a bus system.

[0056] Processor unit 604 is used to execute instructions of software that may be loaded into memory 606. Depending on the specific implementation, processor unit 604 may be a number of processors, a multi-processor core, or some other type of processor. In one embodiment, processor unit 604 includes one or more conventional general-purpose central processing units (CPUs). In an alternative embodiment, processor unit 604 includes one or more graphics processing units (GPUs).

[0057] Memory 606 and persistent storage 608 are examples of storage devices 616. A storage device is any piece of hardware capable of storing information, such as, but not limited to, at least one of data, program code in functional form, or other suitable information, temporarily, permanently, or both. In these illustrative examples, storage devices 616 may also be referred to as computer-readable storage devices. In these examples, memory 606 may be, for example, random access memory or any other suitable volatile or non-volatile storage device. Persistent storage 608 may take various forms, depending on the particular implementation.

[0058] For example, persistent storage 608 may include one or more components or devices. For example, persistent storage 608 may be a hard drive, flash memory, a rewritable optical disk, a rewritable magnetic tape, or some combination thereof. The media used by persistent storage 608 may also be removable. For example, a removable hard drive may be used for persistent storage 608. In these illustrative examples, communications unit 610 provides for communications with other data processing systems or devices. In these illustrative examples, communications unit 610 is a network interface card.

[0059] Input / output unit 612 allows for input and output of data with other devices that may be connected to data processing system 600. For example, input / output unit 612 may provide a connection for user input via at least one of a keyboard, a mouse, or some other suitable input device. Additionally, input / output unit 612 may send output to a printer. Display 614 provides a mechanism for displaying information to a user.

[0060] Instructions for at least one of an operating system, applications, or programs may be located in storage devices 616, which are in communication with processor unit 604 through communications framework 602. The processes of the different embodiments may be performed by processor unit 604 using computer-implemented instructions, which may be located in a memory, such as memory 606.

[0061] These instructions are referred to as program code, computer usable program code, or computer readable program code, which can be read and executed by a processor in processor unit 604. The program code in different embodiments may be embodied on different physical or computer-readable storage media, such as memory 606 or persistent storage 608.

[0062] Program code 618 is located in functional form on computer-readable media 620, which is selectively removable and can be loaded or transferred to data processing system 600 for execution by processor unit 604. In these illustrative examples, program code 618 and computer-readable media 620 form computer program product 622. In one example, computer-readable media 620 may be computer-readable storage media 624 or computer-readable signal media 626.

[0063] In these illustrative examples, computer-readable storage media 624 is a physical or tangible storage device used to store program code 618, rather than a medium that propagates or transmits program code 618. As used herein, computer-readable storage media 624 should not be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses passing through fiber optic cables), or electrical signals transmitted through wires.

[0064] Alternatively, program code 618 can be transmitted to data processing system 600 using computer-readable signal media 626. Computer-readable signal media 626 can be, for example, a propagated data signal containing program code 618. For example, computer-readable signal media 626 can be at least one of an electromagnetic signal, an optical signal, or any other suitable type of signal. These signals can be transmitted via at least one communication link, such as a wireless communication link, a fiber optic cable, a coaxial cable, an electrical wire, or any other suitable type of communication link.

[0065] The different components illustrated for data processing system 600 are not meant to provide architectural limitations to the manner in which different embodiments may be implemented. The different illustrative embodiments may be implemented in a data processing system including components in addition to or in place of those illustrated for data processing system 600. Figure 6 Other components shown in FIG614 may vary from the illustrative examples shown. The different embodiments may be implemented using any hardware device or system capable of running program code 618.

[0066] As used herein, when used with a list of items, the phrase "at least one of" means that different combinations of one or more of the listed items can be used, and only one of each item in the list may be required. In other words, "at least one of" refers to any combination of items and quantities that can be used in the list, but not all items in the list are required. An item can be a specific object, thing, or category.

[0067] For example, but not limited to, "at least one of Item A, Item B, or Item C" may include Item A, Item A and Item B, or Item B. This example may also include Item A, Item B, and Item C, or Item B and Item C. Of course, any combination of these items may be present. In some illustrative examples, "at least one" may include, for example, but not limited to, two of Item A; one of Item B; and ten of Item C; four of Item B and seven of Item C; or other suitable combinations.

[0068] As used herein, when used with reference to an item, "several" refers to one or more items. For example, "several different types of networks" refers to one or more different types of networks. In the illustrative examples, when used with reference to an item, "group / set" refers to one or more items. For example, a set of indicators refers to one or more indicators.

[0069] The descriptions of the different illustrative embodiments are presented for purposes of illustration and description and are not meant to be exhaustive or limited to the embodiments in the form disclosed. The different illustrative examples describe components that perform actions or operations. In the illustrative embodiments, the components can be configured to perform the described actions or operations. For example, the component can have a configuration or design for a structure that provides the component with the ability to perform the actions or operations described as being performed by the component in the illustrative examples. Furthermore, to the extent the terms "comprises," "includes," "has," "contains," and variations thereof are used herein, these terms are intended to be inclusive in a manner similar to the term "comprising" as an open transition word, without excluding any additional or other elements.

[0070] Additionally, this application includes examples under the following terms:

[0071] Clause 1. A system for airline rating, review recommendation, and search result prediction mapping, the system comprising:

[0072] a storage device (616) that stores program instructions;

[0073] One or more processors (604) operably connected to the storage device and configured to execute the program instructions to cause the system to:

[0074] In response to a first user partial entry in an evaluation category entry field (106), providing (502) by a first large language model (124) a first plurality of hints of suggested evaluation categories based on a plurality of predefined evaluation categories (130);

[0075] receiving (504) user entry of a selection or alternative evaluation category for one of the first plurality of prompts;

[0076] In response to a second user partial entry in the task description entry field (108), providing (506) by the first large language model a second plurality of suggested prompts for complete task descriptions based on historical content (132) associated with the evaluation category;

[0077] receiving (508) user input of a selection of one of the second plurality of prompts or an alternative complete task description;

[0078] In response to a third partial user entry in the comment entry field (110), providing (510) by the first large language model a third prompt comprising an initial subset of words for a suggested comment based on historical content (134) associated with the task description;

[0079] receiving (512) user input of a comment selecting or replacing one of the third prompts;

[0080] providing (514) a plurality of predicted search results based on the evaluation category, the task description, and the comments by a second large language model (126); and

[0081] User input is received (516) for a selection of one of the predicted search results or an alternative search result.

[0082] Clause 2. The system of clause 1, wherein the predefined evaluation categories include:

[0083] flight operations;

[0084] Simulator operation;

[0085] Training Department;

[0086] Ministry of Security;

[0087] Crew dispatch department; and

[0088] Crew dispatch department.

[0089] Clause 3. A system according to clause 1, wherein the first large language model includes a generative pre-trained transformer algorithm, namely a GPT algorithm, which uses evaluation categories, task descriptions, and comments to find relationships between words present in a sentence or paragraph, and determines the meaning of words based on their position in the sentence to find the contextual meaning of the sentence or paragraph.

[0090] Clause 4. A system according to clause 3, wherein the GPT algorithm is integrated with a web-based front-end application (102) using an interface (104), wherein the interface provides a table format with several smart search fields, and the smart search fields trigger a front-end application programming interface call in response to the entry of several words.

[0091] Clause 5. A system according to clause 1, wherein the second large language model includes a generative pre-trained transformer algorithm (GPT algorithm) that uses evaluation categories, task descriptions and comments from the first large language model, or manually entered comments from the user, to map the context of sentences and paragraphs to the type of search results.

[0092] Clause 6. The system of clause 1, wherein the first large language model is retrained and updated using user input of alternative evaluation categories, alternative complete task descriptions, or alternative reviews.

[0093] Clause 7. The system of clause 1, wherein the second large language model is retrained and updated using user input of alternative lookup results.

[0094] Clause 8. The system of clause 1, wherein the initial subset of words for the suggested review comprises 3 to 4 words.

[0095] Clause 9. A computer program product for airline ratings, review recommendations, and search result prediction mapping, the computer program product comprising:

[0096] A computer readable storage medium (624) having program instructions embodied thereon to:

[0097] In response to a first user partial entry in an evaluation category entry field (106), providing (502) by a first large language model a first plurality of hints of suggested evaluation categories based on a plurality of predefined evaluation categories (130);

[0098] receiving (504) user entry of a selection or alternative evaluation category for one of the first plurality of prompts;

[0099] In response to a second user partial entry in the task description entry field (108), providing (506) by the first large language model a second plurality of suggested prompts for complete task descriptions based on historical content (132) associated with the evaluation category;

[0100] receiving (508) user input of a selection of one of the second plurality of prompts or an alternative complete task description;

[0101] In response to a third partial user entry in the comment entry field (110), providing (510) by the first large language model a third prompt comprising an initial subset of words for a suggested comment based on historical content (134) associated with the task description;

[0102] receiving (512) user input of a comment selecting or replacing one of the third prompts;

[0103] providing (514) a plurality of predicted search results based on the evaluation category, the task description, and the comments by a second large language model (126); and

[0104] User input is received (516) for a selection of one of the predicted search results or an alternative search result.

[0105] Clause 10. The computer program product of clause 9, wherein the predefined evaluation categories include:

[0106] flight operations;

[0107] Simulator operation;

[0108] Training Department;

[0109] Ministry of Security;

[0110] Crew dispatch department; and

[0111] Crew dispatch department.

[0112] Clause 11. A computer program product according to clause 9, wherein the first large language model includes a generative pre-trained transformer algorithm (GPT algorithm) that uses evaluation categories, task descriptions, and reviews to find relationships between words present in a sentence or paragraph, and determines the meaning of words based on their position in the sentence to find the contextual meaning of the sentence or paragraph.

[0113] Clause 12. A computer program product according to clause 9, wherein the second large language model includes a generative pre-trained transformer algorithm (GPT algorithm) that uses evaluation categories, task descriptions and comments from the first large language model, or manually entered comments from the user, to map the context of sentences and paragraphs to the type of search results.

[0114] Many modifications and variations will be apparent to those skilled in the art. Furthermore, different illustrative embodiments may provide different features than other desired embodiments. The selected embodiment or embodiments are chosen and described in order to best explain the principles of the embodiments, their practical application, and to enable those skilled in the art to understand the disclosure of various embodiments with various modifications as are suitable for the particular use contemplated.

Claims

1. A computer-implemented method for airline rating, review recommendation, and search result prediction mapping, the method comprising: Use several processors to execute: In response to a first user partial entry in an evaluation category entry field (106), providing (502) by a first large language model (124) a first plurality of hints of suggested evaluation categories based on a plurality of predefined evaluation categories (130); receiving (504) user entry of a selection or alternative evaluation category for one of the first plurality of prompts; In response to a second user partial entry in the task description entry field (108), providing (506) by the first large language model a second plurality of suggested prompts for complete task descriptions based on historical content (132) associated with the evaluation category; receiving (508) user input of a selection of one of the second plurality of prompts or an alternative complete task description; In response to a third partial user entry in the comment entry field (110), providing (510) by the first large language model a third prompt comprising an initial subset of words for a suggested comment based on historical content (134) associated with the task description; receiving (512) user input of a comment selecting or replacing one of the third prompts; providing (514) a plurality of predicted search results based on the evaluation category, the task description, and the comments by a second large language model (126); as well as User input is received (516) for a selection of one of the predicted search results or an alternative search result.

2. The method according to claim 1, wherein the predefined evaluation categories include: flight operations; Simulator operation; Training Department; Ministry of Security; Crew dispatch department; as well as Crew dispatch department.

3. The method according to claim 1, wherein the first large language model includes a Generative Pre-Trained Transformer (GPT) algorithm, which uses evaluation categories, task descriptions, and reviews to find relationships between words in a sentence or paragraph, and determines the meaning of words based on their position in the sentence to find the contextual meaning of the sentence or paragraph.

4. The method according to claim 3, wherein the GPT algorithm is integrated with a web-based front-end application (102) using an interface (104), wherein the interface provides a table format with several smart search fields, and the smart search fields trigger a front-end application programming interface call in response to the entry of several words.

5. The method of claim 1, wherein the second large language model comprises a Generative Pre-Trained Transformer (GPT) algorithm that uses evaluation categories, task descriptions, and reviews from the first large language model, or manually entered reviews from the user, to map the context of sentences and paragraphs to the type of search results.

6. The method of claim 1, wherein the first large language model is retrained and updated using user input of alternative evaluation categories, alternative complete task descriptions, or alternative reviews.

7. The method of claim 1, wherein the second large language model is retrained and updated using user input of alternative search results. The method of claim 1 , wherein the initial subset of words for suggested reviews comprises 3 to 4 words.

9. A system for airline evaluation, review recommendation, and search result prediction mapping, the system comprising: a storage device (616) that stores program instructions; One or more processors (604) operably connected to the storage device and configured to execute the program instructions to cause the system to: In response to a first user partial entry in an evaluation category entry field (106), providing (502) by a first large language model (124) a first plurality of hints of suggested evaluation categories based on a plurality of predefined evaluation categories (130); receiving (504) user entry of a selection or alternative evaluation category for one of the first plurality of prompts; In response to a second user partial entry in the task description entry field (108), providing (506) by the first large language model a second plurality of suggested prompts for complete task descriptions based on historical content (132) associated with the evaluation category; receiving (508) user input of a selection of one of the second plurality of prompts or an alternative complete task description; In response to a third partial user entry in the comment entry field (110), providing (510) by the first large language model a third prompt comprising an initial subset of words for a suggested comment based on historical content (134) associated with the task description; receiving (512) user input of a comment selecting or replacing one of the third prompts; providing (514) a plurality of predicted search results based on the evaluation category, the task description, and the comments by a second large language model (126); as well as User input is received (516) for a selection of one of the predicted search results or an alternative search result.

10. The system of claim 9, wherein the predefined evaluation categories include: flight operations; Simulator operation; Training Department; Ministry of Security; Crew dispatch department; as well as Crew dispatch department.

Images