Information processing method and apparatus, storage medium, and processor

Abstract

The application discloses an information processing method and device, a storage medium and a processor. The method comprises the following steps: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and the correlation degree of the target subtext with the test information is higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; and outputting the second test result. The application solves the technical problem that the prediction of the test result consumes a lot of resources due to the dependence on manual labeling.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information processing, and in particular, to an information processing method and device, a storage medium and a processor. BACKGROUND

[0002] Currently, when predicting test results, structured evidence can be explicitly extracted for learning to train a model to predict test results through the trained model.

[0003] Although the above method can achieve the purpose of predicting test results, the extraction of structured evidence in the above method relies on a large amount of manual annotation, which consumes a lot of resources.

[0004] In view of the problem that the prediction of test results relies on manual annotation and consumes a lot of resources, no effective solution has been proposed so far. SUMMARY

[0005] The embodiments of the present application provide an information processing method, device, storage medium and processor to at least solve the technical problem that the prediction of test results relies on manual annotation and consumes a lot of resources.

[0006] According to an aspect of the embodiments of the present application, an information processing method is provided. The method can include: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result and has a higher correlation with the test information than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; and outputting the second test result.

[0007] According to an aspect of the embodiments of the present application, another information processing method is also provided. The method can include: displaying test information on an operation interface, wherein the test information is used to obtain target text associated with the test information in a target database, the target text includes at least one target subtext, the target subtext is used to represent a first test result and has a higher correlation with the test information than a target threshold; and displaying a second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is predicted based on the target subtext.

[0008] According to an aspect of an embodiment of the present application, there is further provided another information processing method. The method can include: obtaining first target language information, and extracting test information from the first target language information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and has a relevance to the test information higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; and outputting second target language information carrying the second test result.

[0009] According to an aspect of an embodiment of the present application, there is further provided an information processing apparatus. The apparatus can include: a first obtaining unit configured to obtain test information; a second obtaining unit configured to obtain target text associated with the test information in a target database; a first extracting unit configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and has a relevance to the test information higher than a target threshold; a first predicting unit configured to predict a second test result corresponding to the test information based on the target subtext; and a first output unit configured to output the second test result.

[0010] According to an aspect of an embodiment of the present application, there is further provided another information processing apparatus. The apparatus can include: a first display unit configured to display test information on an operation interface, wherein the test information is used to obtain target text associated with the test information in a target database, the target text includes at least one target subtext, the target subtext is used to represent a first test result, and has a relevance to the test information higher than a target threshold; and a second display unit configured to display a second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is predicted based on the target subtext.

[0011] According to an aspect of an embodiment of the present application, there is further provided another information processing apparatus. The apparatus can include: a third obtaining unit configured to obtain first target language information, and extract test information from the first target language information; a fourth obtaining unit configured to obtain target text associated with the test information in a target database; a second extracting unit configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and has a relevance to the test information higher than a target threshold; a second predicting unit configured to predict a second test result corresponding to the test information based on the target subtext; and a second output unit configured to output second target language information carrying the second test result.

[0012] According to an aspect of the embodiments of the present application, a computer readable storage medium is also provided. The computer readable storage medium includes a stored program, wherein the program, when executed by a processor, controls the device where the computer readable storage medium is located to perform the information processing method of the embodiments of the present application.

[0013] According to an aspect of the embodiments of the present application, a processor is also provided. The processor is configured to execute a program, wherein the program, when executed, performs the information processing method of the embodiments of the present application.

[0014] According to an aspect of the embodiments of the present application, an information processing system is also provided. The system can include: a processor; a memory connected to the processor, configured to provide the processor with instructions for processing the following processing steps: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result and has a correlation degree with the test information higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; and outputting the second test result.

[0015] In the information processing method of the embodiments of the present application, the test information is obtained; the target text associated with the test information is obtained in the target database; at least one target subtext is extracted from the target text, wherein the target subtext is used to represent a first test result and has a correlation degree with the test information higher than a target threshold; a second test result corresponding to the test information is predicted based on the target subtext; and the second test result is output. That is, the present application finds ready-made relevant target text in the target database as judgment information, so as to use the target subtext with a correlation degree with the test information higher than a target threshold in the target text as a judgment information reasoning result. This process is similar to the process of expert judgment of test results, avoids the cost brought by structured explicit evidence labeling, and provides interpretability for the prediction of test results for test information. Then, the second test result corresponding to the test information is predicted based on the target subtext, and the second test result is output, thereby solving the technical problem that the prediction of test results consumes a lot of resources due to the dependence on artificial labeling, and achieving the technical effect of reducing the resources used in the prediction of test results. BRIEF DESCRIPTION OF DRAWINGS

[0016] The accompanying drawings, which are included to provide a further understanding of the present application and constitute a part of this application, illustrate certain illustrative embodiments of the present application and together with the description serve to explain the present application. In the drawings:

[0017] Figure 1 is a hardware structure block diagram of a computer terminal (or mobile device) for implementing the information processing method according to the embodiments of the present application;

[0018] Figure 2 is a flowchart of an information processing method according to an embodiment of the present application;

[0019] Figure 3 is a flowchart of another information processing method according to an embodiment of the present application;

[0020] Figure 4 is a flowchart of another information processing method according to an embodiment of the present application;

[0021] Figure 5 is a schematic diagram of test information according to an embodiment of the present application;

[0022] Figure 6A is a schematic diagram of test result prediction based on search and evidence reasoning according to an embodiment of the present application;

[0023] Figure 6B is a schematic diagram of an interactive interface of an information processing method according to an embodiment of the present application;

[0024] Figure 6C is a schematic diagram of an application scenario of an information processing method according to an embodiment of the present application;

[0025] Figure 7 is a schematic diagram of an information processing device according to an embodiment of the present application;

[0026] Figure 8 is a schematic diagram of another information processing device according to an embodiment of the present application;

[0027] Figure 9 is a schematic diagram of another information processing device according to an embodiment of the present application; and

[0028] Figure 10 is a structural block diagram of a computer terminal according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] In order to make the personnel in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0030] It should be noted that the terms "first", "second", and the like in the description and in the claims of the present application and above mentioned drawings are intended to identify similar objects, and are not necessarily intended to describe a particular sequential or chronological order. It is to be understood that data so designated can be interchanged, where appropriate, to refer to an embodiment of the application described herein in other than the order described. Furthermore, the terms "comprising" and "including" and any variations thereof are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or apparatus that comprises a list of steps or units can not necessarily be limited to those steps or units that are clearly recited, but can include other not expressly recited steps or units, or steps or units that are inherent to such process, method, product or apparatus.

[0031] First, some of the nouns or terms that appear in the description of the embodiments of the present application are applicable to the following explanations:

[0032] Clinical trial refers to an experimental method of randomized grouping with patients or healthy volunteers as research objects, and individual as a unit, which aims to evaluate the efficacy of a new drug or new therapy for a certain disease, and also includes whether it can prevent recurrence and prolong life;

[0033] Lucene is an open source library for full-text retrieval and search, commonly used for text retrieval;

[0034] Term Frequency-Inverse Document Frequency (TF-IDF) is a weighting method for vectorizing representation of text in retrieval;

[0035] Transformer represents a sequence modeling model;

[0036] Pre-trained model (BERT) is a Transformer encoder for language model pre-training, which can include BioBERT pre-trained with biomedical text data;

[0037] Fine-tuning refers to the process of updating the parameters of the pre-trained model with task-specific data, and the fine-tuned pre-trained model is more suitable for the corresponding task.

[0038] Search engine (Pubmed) database is a database containing about 30 million English biomedical article abstracts;

[0039] Digital repository (Pubmed Central) database is a database containing about 6 million English biomedical article full texts.

[0040] Example 1

[0041] According to an embodiment of the present application, an embodiment of an information processing method is also provided. It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0042] The method embodiment provided by the embodiment of the present application can be executed in a mobile terminal, a computer terminal or similar computing device. Figure 1 is a hardware structure block diagram of a computer terminal (or mobile device) for implementing an information processing method according to an embodiment of the present application. As shown in Figure 1 , the computer terminal 10 (or mobile device 10) can include one or more processors 102 (the processor 102 can include but not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions. In addition, it can also include a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which can be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can include more or less components than those shown in Figure 1 , or have a different configuration from that shown in Figure 1 .

[0043] It should be noted that the one or more processors 102 and / or other data processing circuits described above can be referred to as "data processing circuits" herein. The data processing circuit can be embodied in whole or in part as software, hardware, firmware or any other combination. In addition, the data processing circuit can be a single independent processing module, or any one of the other elements combined into the computer terminal 10 (or mobile device) in whole or in part. As referred to in the embodiments of the present application, the data processing circuit serves as a processor to control (for example, selection of a variable resistance terminal path connected to an interface).

[0044] The memory 104 can be used to store software programs of application software and modules, such as program instructions / data storage means corresponding to the information processing method in the embodiments of the present application. The processor 102 executes various functional applications and data processing, i.e., implements the information processing method described above, by running the software programs and modules stored in the memory 104. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include memories remotely arranged with respect to the processor 102, which can be connected to the computer terminal 10 through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0045] The transmission device 106 is used to receive or send data via a network. Specific examples of the network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network interface controller (NIC) that can be connected to other network devices through a base station so as to be able to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module used to communicate with the Internet in a wireless manner.

[0046] The display can be, for example, a touch screen type liquid crystal display (LCD) that can enable a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0047] It should be noted that, in some optional embodiments, the above Figure 1 The computer device (or mobile device) shown can include hardware elements (including circuitry), software elements (including computer code stored on a computer readable medium), or a combination of both hardware and software elements. It should be noted that, Figure 1 is only one example of a particular implementation and is intended to illustrate the types of components that can be present in the above-described computer device (or mobile device).

[0048] In the above-described operating environment, the present application provides an information processing method as shown in Figure 2 . Figure 2 is a flowchart of an information processing method according to an embodiment of the present application. As described above, Figure 2 the method can include the following steps:

[0049] In step S202, test information is acquired.

[0050] In the technical solution provided in the step S202 of the present application, the test information can be information related to the prediction task of the test result, can be a set of element information of the target test, can also be referred to as target test element information, and can include but is not limited to at least one of the following: background information of the target test; object information of the target test; treatment information of the target test; reference information of the target test; and measurement index of the target test. The target test can be a clinical test, which is a randomized grouping experimental method taking patients or healthy volunteers as research objects and taking individuals as units, and the purpose is to evaluate the efficacy of a new drug or a new treatment on a certain disease, or whether it can prevent recurrence and prolong life. Alternatively, the test information of this embodiment is clinical test information, the test result is a clinical test result, and the target test element information is clinical test element information.

[0051] In this embodiment, the background information of the target test can be used to indicate the research background (Background) of the target test, and can be represented by B; the object information of the target test can be used to indicate the research population (Population), and can be represented by P; the treatment information of the target test can be used to indicate the method of intervention (Intervention), and can be represented by I; the reference information of the target test can be used to indicate the comparison method (Comparison) of the target test, and can be represented by C; and the measurement index (Outcom) of the target test can be used to indicate the standard to be reached by the measured item, and can be represented by O, so that the PICO form of the target test content indicates the object information of the target test, the treatment information of the target test, the reference information of the target test, and the measurement index of the target test.

[0052] In step S204, target text associated with the test information is obtained from a target database.

[0053] In the technical solution provided in the step S204 of the present application, after obtaining the test information, the target text associated with the test information is obtained from the target database.

[0054] In this embodiment, the target database can be a medical database, which can be a search engine (Pubmed) database for providing biomedical paper search and abstract, can be a database containing about 30 million English biomedical article abstracts, and can also be a digital warehouse (Pubmed Central) database for archiving biomedicine, which is a database containing about 6 million English biomedical article full texts.

[0055] In this embodiment, the target text associated with the trial information can be obtained from the target database using a tool for text retrieval, such as a Lucene, which is an open source library for full-text search and search, and the sorting is based on the background information of the target trial, the object information of the target trial, the treatment information of the target trial, the reference information of the target trial, the measurement index of the target trial, and the TF-IDF vector distance determined between PubMed and PubMed Central. Wherein, TF-IDF is a weighting method for vector representation of text in retrieval. This embodiment can use Lucene to obtain the target text associated with the trial information in the target database, such as the target text being PubMed articles, PubMed Central articles, which can be denoted as a1, a2, a3,..., an. n Wherein, n is used to represent the number of articles to be retrieved.

[0056] Optionally, when the target text in PubMed and PubMed Central is found using the method of retrieval, other weighting processing methods can also be used, such as BM25 algorithm, wherein BM25 is an algorithm for evaluating the relevance between search words and documents, which is an algorithm based on a probabilistic retrieval model, which is not limited here.

[0057] Step S206, at least one target subtext is extracted from the target text, wherein the target subtext is used to represent the first trial result.

[0058] In the technical solution provided by the above step S206 of the present application, after obtaining the target text associated with the trial information in the target database, at least one target subtext is extracted from the target text, wherein the target subtext is used to represent the first trial result, and the relevance with the trial information is higher than the target threshold. Optionally, the first trial result of this embodiment is the first clinical trial result.

[0059] In this embodiment, the target text can be split to obtain at least one target subtext, such as a1, a2, a3,..., an. nThe splitting is performed to obtain at least one target subtext, which can be a sentence, a sentence for prompting a relevant test result, a sentence for representing the first test result, and has a relevance to the test information higher than a target threshold, wherein the relevance is a degree of association (relevance degree) between the target subtext and the test information, so that the first test result is a similar test result of a final test result for predicting the test information, wherein the target threshold is related to a number of the at least one target subtext to be determined, for example, m target subtexts are to be determined, the relevance of the m target subtexts to the test information is higher than the relevance of any subtext other than the m target subtexts to the test information in the plurality of subtexts, and the m target subtexts include a target subtext with the minimum relevance to the test information, so that the target threshold is less than the minimum relevance and greater than the relevance of any subtext other than the m target subtexts to the test information, in which case, the number of at least one target subtext with a relevance to the test information greater than the target threshold is m, for example, the m target subtexts are s1, s2, s3,..., s m .

[0060] This embodiment avoids the cost of structurally explicitly labeling evidence by finding ready-made at least one target subtext in a target database (such as PubMed and PubMed Central) to use as a judgment information reasoning result, and the process is similar to that of expert judgment of test results, and also provides interpretability for prediction of test results for test information.

[0061] In step S208, a second test result corresponding to the test information is predicted based on the target subtext.

[0062] In the technical solution provided by the above step S208 of the present application, after at least one target subtext is extracted from the target text, a second test result corresponding to the test information can be predicted based on the target subtext. Optionally, the second test result of this embodiment is a second clinical test result.

[0063] In this embodiment, a second test result corresponding to the test information can be determined based on the target subtext and the test information, and the second test result is a final prediction test result for the test information, because the target subtext is used to represent the first test result, and the first test result is a similar test result of a test result for predicting the test information, so that this embodiment predicts the result of a target test that has not been carried out based on the result of a relevant completed target test, overcoming the disadvantage that the result of a deep learning model is difficult to explain.

[0064] In step S210, the second test result is output.

[0065] In the technical solution provided in the above step S210 of the present application, after the second test result corresponding to the test information based on the target subtext is predicted, the second test result can be output, which can be represented by. Alternatively, the embodiment can output the second test result to the display screen for display, or play the second test result through a voice player to assist the selection of the target test and help the medical researchers to select the appropriate target test.

[0066] In this embodiment, the test information (B, PICO) is input, the predicted second test result is R, and the relationship between the control group generally has three possibilities of rising, falling and remaining unchanged.

[0067] Through the above steps S202 to S210 of the present application, the test information is obtained; the target text associated with the test information is obtained in the target database; at least one target subtext is extracted from the target text, wherein the target subtext is used to represent the first test result and has a correlation degree with the test information higher than a target threshold; the second test result corresponding to the test information based on the target subtext is predicted; and the second test result is output. That is, the embodiment finds ready-made relevant target text in the target database as judgment information, so as to use the target subtext with a correlation degree with the test information higher than the target threshold in the target text as the judgment information reasoning result. The process is similar to the process of expert judging the test result, avoids the cost brought by structurally explicitly labeling evidence, and provides interpretability for the prediction of the test result for the test information. Then, the second test result corresponding to the test information based on the target subtext is predicted, and the second test result is output, which solves the technical problem that the prediction of the test result consumes a lot of resources due to the dependence on artificial labeling, and achieves the technical effect of reducing the resources used in the prediction of the test result.

[0068] It should be noted that the information processing method of the embodiment can be applied to medical clinics, but the information processing method of the embodiment is not limited to the scene of clinical medicine. Any scene that can use the information processing method of the embodiment is within the scope of the present application, such as biological, chemical, physical and other test scenes.

[0069] For example, in a biological test scene, biological test information is obtained; target text associated with the biological test information is obtained in a biological database; at least one target subtext is extracted from the target text, wherein the target subtext is used to represent a first biological test result and has a correlation degree with the biological test information higher than a target threshold; a second biological test result corresponding to the biological test information based on the target subtext is predicted; and the second biological test result is output.

[0070] For example, in a chemical test scenario, obtain chemical test information; obtain target text associated with the chemical test information in a chemical database; extract at least one target subtext from the target text, wherein the target subtext is used to represent a first chemical test result and has a relevance to the chemical test information higher than a target threshold; predict a second chemical test result corresponding to the chemical test information based on the target subtext; and output the second chemical test result.

[0071] For example, in a physical test scenario, obtain physical test information; obtain target text associated with the physical test information in a physical database; extract at least one target subtext from the target text, wherein the target subtext is used to represent a first physical test result and has a relevance to the physical test information higher than a target threshold; predict a second physical test result corresponding to the physical test information based on the target subtext; and output the second physical test result.

[0072] The above method of this embodiment is further described below.

[0073] As an optional implementation, step S206, extracting at least one target subtext from the target text, comprises: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the relevance of the sentence to the test information; and determining the at least one target subtext based on the first values of the plurality of sentences.

[0074] In this embodiment, the target text is composed of sentences, when extracting at least one target subtext from the target text, the target text associated with the input test information can be split to obtain a plurality of sentences, the first value of each sentence can be determined based on each sentence and the test information, the first value is used to indicate the relevance of the sentence to the test information, and the relevance score used to represent the relevance of each sentence to the test information. After determining the first value of the sentence based on the sentence and the test information, the at least one target subtext can be determined based on the first values of the plurality of sentences. The determination of the first value of the sentence based on the sentence and the test information of this embodiment is further described below.

[0075] As an optional implementation, determining the first value of the sentence based on the sentence and the test information comprises: analyzing the sentence and the test information based on an evidence ranking model to obtain the first value of the sentence, wherein the evidence ranking model is trained by a test information sample, a first subtext sample associated with the test information sample, and a second subtext sample not associated with the test information sample.

[0076] In this embodiment, the evidence ranking model (BioBERT F), which can be obtained by a language model pre-trained Transformer encoder (BERT), BioBERT is used to represent BERT pre-trained with biomedical text data, wherein the first subtext sample associated with the trial information sample can be the relevant evidence sentence of the trial information sample, and the second subtext sample not associated with the trial information sample can be the irrelevant sentence data set of the trial information sample, so that the evidence ranking model of the embodiment can be used to represent a model fine-tuned with the trial information sample and the relevant evidence sentence and the irrelevant sentence data set, and the evidence ranking model can be used to determine the evidence sentence related to the trial information with a higher first value, so that in the embodiment, when the first value of the sentence based on the sentence and the trial information is determined, the sentence of the target text and the trial information are input into the evidence ranking model, the sentence of the target text and the trial information are analyzed by the evidence ranking model, and the first value of the sentence of the target text is obtained, and then the first value of the sentence of the target text is output to determine the relevance of the sentence of the target text and the trial information. Wherein, fine-tuning refers to the process of updating the parameters of the pre-trained model with task-specific data, and the fine-tuned pre-trained model is more suitable for the corresponding task.

[0077] In this embodiment, the evidence ranking model can be a convolutional neural network (CNN), a long short-term memory network (LSTM), or other sequence modeling models (Transformer) to determine the target subtext (to find the relevant evidence sentence).

[0078] As an optional implementation, determining at least one target subtext based on the first values of the plurality of sentences includes: determining at least one first value greater than a target threshold from the first values of the plurality of sentences; and determining at least one sentence corresponding to the at least one first value as the at least one target subtext.

[0079] In this embodiment, the first value of the sentence is used to indicate the relevance of the sentence and the trial information, which can be a relevance score, and the embodiment can determine at least one first value greater than a target threshold from the first values of the plurality of sentences of the target text, for example, the highest m relevance scores can be r1, r2, r3,..., r m Based on r1, r2, r3,..., r m Select the m sentences with the highest relevance scores as the relevant evidence of the trial information, and record them as s1, s2, s3,..., s m Corresponding to r1, r2, r3,..., r m , which can be determined as at least one target subtext.

[0080] As an optional implementation, the second test result corresponding to the target subtext pre-test information is predicted based on the target subtext and the test information.

[0081] In this embodiment, when the second test result corresponding to the target subtext pre-test information is implemented, the second test result corresponding to the test information can be predicted by the target subtext and the test information together.

[0082] As an optional implementation, the second test result is predicted based on the target subtext and the test information, which includes: in the case of multiple target subtexts, the multiple target subtexts and the test information are analyzed based on an evidence reasoning model respectively, to obtain multiple sub-test results, wherein the evidence reasoning model is obtained by training a test information sample, a first subtext sample associated with the test information sample, and a test result sample; and the second test result is determined based on the multiple sub-test results.

[0083] In this embodiment, the evidence reasoning model (BioBERT C ) can be trained in advance, the first subtext sample associated with the test information sample can be the relevant evidence sentence of the test information sample, and the test result sample can be a dataset labeled by the test result. The evidence reasoning model of this embodiment can represent a model fine-tuned by the target test element information and the relevant evidence sentence and the dataset labeled by the test result. Therefore, when the second test result is predicted based on the target subtext and the test information in this embodiment, the target subtext and the test information can be input into the evidence reasoning model, and the multiple target subtexts and the test information are analyzed by the evidence reasoning model respectively to obtain multiple sub-test results, such as This step is evidence reasoning. Each target subtext corresponds to a sub-test result, which is a completed test result similar to the final test result predicted for the test information. After obtaining the sub-test results corresponding to the multiple target subtexts respectively, the second test result can be determined based on the multiple sub-test results in this embodiment, thereby realizing the prediction of the result of the target test that has not been carried out based on the results of the similar completed target test, to help medical researchers select appropriate target tests.

[0084] Optionally, in this embodiment, the evidence ranking model can be a CNN, LSTM, or other sequence modeling model (Transformer) for evidence reasoning to determine the sub-test result corresponding to the target subtext.

[0085] As an optional implementation, determining the second test result based on the multiple sub-test results comprises: determining the second test result based on the multiple sub-test results and the second values of the multiple target subtexts, wherein the second value of the target subtext is used to indicate the relevance of the target subtext to the test information.

[0086] In this embodiment, each target subtext corresponds to a second value, which is the relevance score of the clinical evidence. For example, the second values of the multiple target subtexts are r1, r2, r3,..., r m Therefore, when determining the second test result based on the multiple sub-test results, a certain algorithm can be used to determine the second test result based on the multiple sub-test results and the second values of the multiple target subtexts. For example, the multiple sub-test results are and the second values of the multiple target subtexts are r1, r2, r3,..., r m and A certain algorithm is used to determine the second test result, which is used as the final prediction test result for predicting the test information.

[0087] As an optional implementation, determining the second test result based on the multiple sub-test results and the second values of the multiple target subtexts comprises: performing weighted processing on the multiple sub-test results according to the second values of the multiple target subtexts to obtain the second test result.

[0088] In this embodiment, when determining the second test result based on the multiple sub-test results and the second values of the multiple target subtexts, weighted processing can be performed on the multiple sub-test results according to the second values of the multiple target subtexts to obtain the second test result. For example, the product of each sub-test result and the corresponding second value of the target subtext is obtained to obtain multiple products, the first sum of the multiple products is obtained, the second sum of the second values of the multiple target subtexts is obtained, and then the quotient of the first sum and the second sum is determined as the second test result. For example, the multiple sub-test results corresponding to the multiple target subtexts are The second values of the multiple target subtexts are r1, r2, r3,..., r m Therefore, the second test result can be

[0089] It should be noted that the above weighted processing on the multiple sub-test results according to the second values of the multiple target subtexts is only an example of the embodiment of the present application, and is not limited to the weighted processing on the multiple sub-test results according to the second values of the multiple target subtexts in the embodiment of the present application.

[0090] The embodiment of the present application also provides another information processing method.

[0091] Figure 3 is a flowchart of another information processing method according to an embodiment of the present application. As shown in Figure 3 the method can include:

[0092] In step S302, display the trial information on the operation interface, wherein the trial information is used to obtain the target text associated with the trial information in the target database, the target text includes at least one target subtext, the target subtext is used to represent the first trial result, and the relevance degree of the target subtext to the trial information is higher than the target threshold.

[0093] In the technical solution provided by the above step S302 of the present application, the trial information is obtained, and then the trial information is displayed on the operation interface. The trial information can be target trial element information, which can include but is not limited to at least one of the following: background information of the target trial; object information of the target trial; treatment information of the target trial; reference information of the target trial; and measurement index of the target trial.

[0094] In this embodiment, the trial information is used to obtain the target text associated with the trial information in the target database, wherein the target database can be PubMed, and can also be PubMed Central.

[0095] In this embodiment, the target text associated with the trial information can be obtained in the target database by using a text retrieval tool, such as a Lucene tool. Alternatively, the target text of this embodiment is a PubMed article or a PubMed Central article, which can be denoted as a1, a2, a3,..., an. n wherein n is used to represent the number of articles that need to be retrieved.

[0096] In this embodiment, the target text includes at least one target subtext, the target subtext is used to represent the first trial result, and the relevance degree of the target subtext to the trial information is higher than the target threshold. In this embodiment, the target text can be split to obtain at least one target subtext, such as a1, a2, a3,..., an. n The target text is split to obtain at least one target subtext, which can be a sentence, a sentence for prompting a related trial result, used to represent the first trial result, and the relevance degree of the target subtext to the trial information is higher than the target threshold, that is, the first trial result is a similar trial result of the trial result predicted by the trial information.

[0097] The embodiment finds at least one target subtext in a target database (such as PubMed and PubMed Central) to serve as the judgment information reasoning result, which is similar to the process of expert judgment test result, avoids the cost of structured explicit evidence labeling, and provides interpretability for the prediction of test results for test information.

[0098] In step S304, the second test result corresponding to the test information is displayed in the browsing area of the operation interface, wherein the second test result is predicted based on the target subtext.

[0099] In the technical solution provided in step S304 of the present application, after displaying the test information on the operation interface, the second test result corresponding to the test information can be displayed in the browsing area of the operation interface. The second test result can be output to the display screen for display to help medical researchers select appropriate target tests.

[0100] In this embodiment, the second test result corresponding to the test information can be determined based on the target subtext and the test information. The second test result is the final prediction of the test result for the test information. Since the target subtext is used to represent the first test result, the first test result is a similar test result of the test result predicted for the test information. Therefore, this embodiment predicts the result of the target test that has not been carried out based on the result of the related completed target test, which overcomes the disadvantage that the result of the deep learning model is difficult to explain.

[0101] As an optional implementation, before displaying the test information on the operation interface, the method further includes: in response to a selection instruction acting on the operation interface, obtaining a set of element information of the target test, wherein each element information in the set of element information is used to match at least one text in the target database; and displaying the test information on the operation interface includes: displaying the set of element information on the operation interface.

[0102] In this embodiment, before displaying the test information on the operation interface, the user can trigger a selection instruction on the operation interface, which is used to select a set of element information of the target test from a plurality of element information on the operation interface. Therefore, in response to the selection instruction acting on the operation interface, a set of element information of the target test is obtained, which includes a plurality of element information. The plurality of element information can include background information of the target test, object information of the target test, treatment information of the target test, reference information of the target test, and measurement index of the target test. Each element information can be used to match at least one text in the medical knowledge graph. After obtaining the set of element information of the target test, the set of element information can be displayed on the operation interface.

[0103] Figure 4 This is a flowchart of another information processing method according to an embodiment of the present invention. Figure 4 As shown, the method may include the following steps:

[0104] Step S402: Obtain the first target language information and extract the test information from the first target language information.

[0105] In the technical solution provided in step S402 of the present invention, the first target language information can be input information in any natural language, and can be used to model a wider range of interventional treatments. This embodiment acquires the input first target language information, which carries experimental information, and the experimental information can be extracted from the first target language information. The experimental information in this embodiment can be information related to the prediction task of the experimental results, or a set of element information of the target experiment, including but not limited to at least one of the following: background information for conducting the target experiment; object information for conducting the target experiment; treatment information for conducting the target experiment; reference information for conducting the target experiment; and measurement indicators for conducting the target experiment.

[0106] Step S404: Obtain the target text associated with the test information from the target database.

[0107] The technical solution provided in step S404 of the present invention can be implemented according to the method in step S204, which will not be elaborated here.

[0108] Step S406: Extract at least one target subtext from the target text, wherein the target subtext is used to characterize the result of the first experiment.

[0109] In the technical solution provided by step S406 of the present invention, the correlation with the test information is higher than the target threshold.

[0110] The technical solution provided in step S406 of the present invention can be carried out according to the method of step S206, which will not be elaborated here.

[0111] Step S408: Predict the second experimental result corresponding to the experimental information based on the target subtext.

[0112] The technical solution provided in step S408 of the present invention can be carried out according to the method in step S208, which will not be elaborated here.

[0113] Step S410: Output the second target language information carrying the second test results.

[0114] In the technical solution provided in the step S410 of the present application, after the second test result corresponding to the target subtext pre-test information is obtained, the second target language information carrying the second test result can be determined, the second target language information can be any natural language, and the second target language information carrying the second test result is output, for example, the second test result is played through a voice player to assist the selection of the target test.

[0115] As an optional implementation, the step S406 of extracting at least one target subtext from the target text comprises: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the relevance of the sentence and the test information; and determining the at least one target subtext based on the first values of the plurality of sentences.

[0116] In this embodiment, the target text is composed of sentences, when extracting at least one target subtext from the target text, the target text associated with the input test information can be split to obtain a plurality of sentences, the first value of each sentence can be determined based on each sentence and the test information, the first value is used to indicate the relevance of the sentence and the test information, which can be a relevance score used to represent the relevance of each sentence and the test information. After determining the first value of the sentence based on the sentence and the test information, the at least one target subtext can be determined based on the first values of the plurality of sentences, for example, at least one first value greater than a target threshold is determined from the first values of the plurality of sentences, and at least one sentence corresponding to the at least one first value is determined as the at least one target subtext.

[0117] As an optional implementation, the step S408 of obtaining the second test result corresponding to the target subtext pre-test information comprises: predicting the second test result based on the target subtext and the test information.

[0118] In this embodiment, the second test result corresponding to the test information can be predicted by the target subtext and the test information together, in the case that the at least one target subtext is a plurality of target subtexts, the plurality of target subtexts and the test information can be analyzed based on an evidence reasoning model to obtain a plurality of sub-test results, wherein the evidence reasoning model is obtained by training a test information sample, a first subtext sample associated with the test information sample, and a test result sample, and the second test result is determined based on the plurality of sub-test results.

[0119] In this embodiment, the method of retrieval is used to find ready-made target texts in the target database, at least one target subtext is extracted from the target text as judgment information, the process is similar to that of the expert judgment test result, and the model prediction provides interpretability, thereby avoiding the cost of structurally explicitly labeling evidence; the target subtext retrieved in this embodiment and the numerical value corresponding thereto can be used as an explanation for the final prediction of the test information corresponding to the test result, thereby overcoming the disadvantage that the deep learning model result is difficult to explain; this embodiment predicts the result of the target test that has not been carried out by retrieving similar completed first test results, thereby achieving the purpose of helping medical researchers to select suitable target tests.

[0120] Embodiment 2

[0121] The above method of the embodiment of the present application will be further illustrated below in combination with preferred embodiments, and specific examples are taken from the target test in clinical medicine as a clinical test, the test result as a clinical test result, the test information as clinical test information, the target test element information as clinical test element information, and the target database as a medical database.

[0122] A clinical test refers to any systematic study of a drug on human beings (patients or healthy volunteers) to confirm or reveal the effects, adverse reactions and / or absorption, distribution, metabolism and excretion of the test drug, with the purpose of determining the efficacy and safety of the test drug. Clinical tests are an important part of evidence-based medicine. The most core part of a clinical test is a randomized controlled clinical test, which is generally considered to be the gold standard for evaluating the safety and effectiveness of a drug. However, when the research conclusions of a randomized controlled clinical test are extrapolated to clinical practical applications, challenges may be faced, or there may be problems such as the difficulty of implementing a traditional drug clinical test or the high time cost. The above-mentioned RCT has strict standards for patient screening and drug management. On the other hand, some rare disease drugs and children's drugs are difficult to obtain due to enrollment, and the cost is huge.

[0123] In the related art, structured clinical evidence can be extracted explicitly for learning, but this method requires a large amount of manual annotation when extracting structured clinical evidence, which is very resource-consuming.

[0124] To solve the above problems, in order to assist the design of clinical experiments, select the clinical trials that are easier to succeed, save costs, and make full use of limited resources, the embodiment proposes to use the method of retrieval and evidence reasoning to predict the results of clinical trials, by retrieving the results of similar completed clinical trials to predict the results of clinical trials that have not been carried out, so as to better help medical researchers to select suitable clinical trials. And by finding ready-made relevant evidence sentences in PubMed and PubMed Central as judgment information reasoning results, the process is also similar to the process of experts judging the results of clinical trials, thereby providing interpretability for the prediction of the model. The above method of the embodiment is further introduced as follows.

[0125] Figure 5 is a schematic diagram of clinical trial information according to an embodiment of the present application. As shown in Figure 5 When a proposal for a new clinical trial is proposed, two parts of content need to be filled in: first, the research background of the clinical trial needs to be filled in, as shown in B (Background) in Figure 5 ; second, the content of the clinical trial in PICO form needs to be filled in, that is, the research population P (Population), the intervention I (Intervention), the comparison method C, and the measurement index O, R is used to represent the output of the clinical trial prediction result.

[0126] In this embodiment, B, PICO are taken as inputs, the result R of the predicted clinical trial, and the relationship between the control group generally has three possibilities of rising, falling, and remaining unchanged.

[0127] Figure 6A is a schematic diagram of a trial result prediction based on retrieval and evidence reasoning according to an embodiment of the present application. As shown in Figure 6A The following methods can be involved:

[0128] S1, use Lucene to preliminarily find PubMed and PubMed Central articles related to the input clinical trial information.

[0129] In this embodiment, Lucene is used to search articles related to the input clinical trial background B and trial element information PICO in PubMed and PubMed Central, and the searched articles are recorded as a1, a2, a3,..., an, where n is used to represent the number of articles retrieved in advance, Lucene is an open source library for full-text retrieval and search, and the sorting is determined based on the TF-IDF vector distance between the above B, PICO and PubMed, PubMed Central articles a n . i ​

[0130] S2, using the evidence ranking model BioBERT F Find sentences in the articles retrieved in S1 that indicate relevant clinical trial results.

[0131] In this embodiment, the articles a1, a2, a3, ..., a1 found in S1 that are related to the input clinical trial information are... n The evidence is broken down into sentences, and each sentence is fed into the BioBERT evidence ranking model along with the PICO (Personal Information Framework) of the clinical trial. F This yields a relevance score indicating the correlation between clinical evidence and the input clinical trial information. The m sentences with the highest relevance scores can be selected as relevant evidence and denoted as s1, s2, s3, ..., s m The corresponding relevance scores can be r1, r2, r3, ..., r m Among them, BioBERT F This represents a BioBERT dataset fine-tuned using clinical trial element information, relevant evidence sentences, and irrelevant sentences. F Evidence sentences that are relevant to the elements of a clinical trial will receive higher scores.

[0132] S3, using the evidence reasoning model BioBERT C The evidence sentences found in S2 are reasoned and integrated to predict the results of the input clinical trial.

[0133] In this embodiment, the evidence s1, s2, s3, ..., s found in S2 that is relevant to the input clinical trial is... m The PICO elements, representing elements of clinical trials, are input into the evidence reasoning model BioBERT. C The clinical use prediction results corresponding to the input clinical trial information are obtained. This step is evidence reasoning. Among them, BioBERT... C This represents a fine-tuned BioBERT dataset annotated with clinical trial element information, relevant evidence sentences, and clinical trial results.

[0134] This embodiment uses relevance scores to weight the clinical trial prediction results corresponding to each evidence sentence, and determines the weighted result as the final clinical trial prediction result:

[0135]

[0136] Figure 6B This is a schematic diagram of an interactive interface for an information processing method according to an embodiment of the present invention. Figure 6BAs shown, the user can drag the clinical trial element information B, P, I, C, O into the "add" text box, click the "clinical trial result prediction" button, obtain the target text associated with the clinical trial information in the medical database, extract at least one target subtext from the target text, wherein the target subtext is used to represent the first clinical trial result, and the relevance of the target subtext to the clinical trial information is higher than the target threshold, predict the second clinical trial result corresponding to the clinical trial information based on the target subtext, and output the second clinical trial result. Finally, the second clinical trial result corresponding to the first target node is displayed in the browsing area of the operation interface, thereby solving the technical problem that the prediction of the test result consumes a lot of resources due to the dependence on manual annotation.

[0137] Figure 6C is a schematic diagram of an application scenario of an information processing method according to an embodiment of the present application. As shown in Figure 6C The clinical trial information is input into the computing device, and the clinical trial information can be displayed on the operation interface of the computing device. The clinical trial information is used to obtain target text associated with the clinical trial information in a medical database. The target text includes at least one target subtext. The target subtext is used to represent a first clinical trial result, and the relevance of the target subtext to the clinical trial information is higher than a target threshold.

[0138] In this embodiment, the clinical trial information is used to obtain target text associated with the clinical trial information in a medical database. The medical database can be PubMed, and can also be PubMed Central.

[0139] In this embodiment, the target text associated with the clinical trial information can be obtained in the medical database by using a text retrieval tool, such as a Lucene tool for text retrieval. Alternatively, the target text of this embodiment is a PubMed article or a PubMed Central article, which can be denoted as a1, a2, a3,..., an. n Wherein n is used to represent the number of articles that need to be retrieved in advance.

[0140] In this embodiment, the target text includes at least one target subtext, which is used to represent the first clinical trial result and has a relevance to the clinical trial information higher than a target threshold. In this embodiment, the target text can be split to obtain at least one target subtext, such as a1, a2, a3,..., an. nThe splitting is performed to obtain at least one target subtext, which can be a sentence, a sentence for prompting a related clinical trial result, a sentence for representing the first clinical trial result, and has a higher relevance to the clinical trial information than a target threshold, that is, the first clinical trial result is a similar clinical trial result of the clinical trial result predicted by the clinical trial information.

[0141] After displaying the clinical trial information on the operation interface, the second clinical trial result corresponding to the clinical trial information can be displayed in the browsing area of the operation interface, and the second clinical trial result can be output to the display screen for display, so as to help the medical researchers to select appropriate clinical trials.

[0142] It should be noted that the embodiment can use other weighting processing methods such as BM25 when finding the existing relevant evidence sentences in PubMed and PubMed Central using the retrieval method, and the present disclosure is not limited in this regard.

[0143] It should be noted that the embodiment can use other sequence modeling models such as CNN and LSTM to find the relevant evidence sentences and the evidence reasoning when finding the existing relevant evidence sentences in PubMed and PubMed Central using the retrieval method.

[0144] It should be noted that the embodiment can use other weighting methods to integrate the target trial prediction results corresponding to each evidence sentence.

[0145] The above method of the scheme can be applied to related products of the medical research platform.

[0146] It should be noted that the embodiment is described by taking the information processing method in clinical medicine as an example, but the scope of the embodiment is not limited to the scenario of the information processing method in clinical medicine. Any scenario in which the information processing method in the present disclosure can be used is within the scope of the embodiment, such as various test scenarios in biology, chemistry, physics, etc., and the present disclosure will not be repeated here.

[0147] The method of the above test results of this embodiment can be implemented by a model, which can accept input of any natural language, and thus can model a wider range of interventions; this embodiment uses the retrieval method to find ready-made relevant evidence sentences in PubMed and PubMed Central as judgment information, avoiding the cost of structured explicit evidence labeling, and the process is similar to the process of expert judgment of test results, thereby providing interpretability for the prediction of the model; the evidence sentences retrieved by this embodiment and the relevance scores thereof can be used as an explanation of the final target test prediction results, thereby overcoming the disadvantage that the results of deep learning models are difficult to interpret; this embodiment proposes predicting the results of the target test to be carried out by retrieving the results of similar completed target tests to help medical researchers select appropriate target tests.

[0148] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all expressed as a combination of a series of actions, but those skilled in the art should know that the present application is not limited to the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0149] From the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software and the necessary general hardware platform, and of course, it can also be implemented by hardware, but in many cases, the former is a better implementation. Based on such understanding, the technical solutions of the present application or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, or optical disk) and includes a number of instructions to make a terminal device (which can be a mobile phone, computer, server, or network device) execute the methods described in various embodiments of the present application.

[0150] Embodiment 3

[0151] According to the embodiments of the present application, an information processing device for implementing the above information processing method is also provided. It should be noted that the information processing device of this embodiment can be used to execute the information processing method of the embodiments of the present application Figure 2 The information processing method of the embodiments shown.

[0152] Figure 7 is a schematic diagram of an information processing device according to an embodiment of the present application. As shown in Figure 7As shown in the figure, the information processing apparatus 70 can include a first acquisition unit 71, a second acquisition unit 72, a first extraction unit 73, a first prediction unit 74, and a first output unit 75.

[0153] The first acquisition unit 71 is configured to acquire test information.

[0154] The second acquisition unit 72 is configured to acquire target text associated with the test information in a target database.

[0155] The first extraction unit 73 is configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent a first test result and has a relevance to the test information higher than a target threshold.

[0156] The first prediction unit 74 is configured to predict a second test result corresponding to the test information based on the target subtext.

[0157] The first output unit 75 is configured to output the second test result.

[0158] It should be noted that the first acquisition unit 71, the second acquisition unit 72, the first extraction unit 73, the first prediction unit 74, and the first output unit 75 correspond to steps S202 to S210 in Embodiment 1, and the four modules have the same instances and application scenarios as the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules as part of the apparatus can run in the computer terminal 10 provided in Embodiment 1.

[0159] The present application also provides another information processing apparatus for implementing the above information processing method. It should be noted that the information processing apparatus of this embodiment can be used to execute the information processing method of the present application Figure 3 The information processing method of the embodiment shown.

[0160] Figure 8 is a schematic diagram of another information processing apparatus according to an embodiment of the present application. As shown in the figure, Figure 8 The information processing apparatus 80 can include a first display unit 81 and a second display unit 82.

[0161] The first display unit 81 is configured to display test information on an operation interface, wherein the test information is used to acquire target text associated with the test information in a target database, the target text includes at least one target subtext, the target subtext is used to represent a first test result, and has a relevance to the test information higher than a target threshold.

[0162] The second display unit 82 is configured to display a second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is obtained based on the target subtext.

[0163] The embodiment of the present application further provides another information processing device for implementing the information processing method. Figure 4 The information processing method shown in the embodiment.

[0164] Figure 9 is a schematic diagram of another information processing device according to an embodiment of the present application. As shown in the figure, the information processing device 90 can include a third acquisition unit 91, a fourth acquisition unit 92, a second extraction unit 93, a second prediction unit 94 and a second output unit 95. Figure 9

[0165] The third acquisition unit 91 is configured to acquire first target language information and extract test information from the first target language information.

[0166] The fourth acquisition unit 92 is configured to acquire target text associated with the test information in a target database.

[0167] The second extraction unit 93 is configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent the first test result and has a higher correlation with the test information than a target threshold.

[0168] The second prediction unit 94 is configured to predict a second test result corresponding to the test information based on the target subtext.

[0169] The second output unit 95 is configured to output second target language information carrying the second test result.

[0170] In the information processing device of the embodiment, the present application finds ready-made relevant target text in the target database as judgment information, so as to infer the judgment information through the target subtext in the target text, which has a higher correlation with the test information than the target threshold, and the process is similar to the process of expert judgment of the test result, thereby avoiding the cost caused by the structured explicit annotation of evidence, and providing interpretability for the prediction of the test result for the test information. Furthermore, the second test result corresponding to the test information is predicted based on the target subtext, and the second test result is output, thereby solving the technical problem that the prediction of the test result consumes a lot of resources due to the dependence on manual annotation, and achieving the technical effect of reducing the resources used in the prediction of the test result.

[0171] Embodiment 4

[0172] ​The embodiment of the present application can provide an information processing system, which can include a computer terminal, which can be any one of the computer terminal devices in a computer terminal group. Alternatively, in the embodiment, the computer terminal can be replaced by a mobile terminal or other terminal device.

[0173] Alternatively, in the embodiment, the computer terminal can be located in at least one of the network devices in a computer network.

[0174] In the embodiment, the computer terminal can execute the program code of the following steps in the information processing method of the application program: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result and has a higher relevance to the test information than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; and outputting the second test result.

[0175] Alternatively, Figure 10 is a structural block diagram of a computer terminal according to an embodiment of the present application. As shown in Figure 10 , the computer terminal A can include one or more (only one is shown in the figure) processor 1002, memory 1004, and transmission device 1006.

[0176] The memory can be used to store software programs and modules, such as program instructions / modules corresponding to the information processing method and device in the embodiment of the present application. The processor executes various functions and data processing by running the software programs and modules stored in the memory, i.e., implements the above-mentioned information processing method. The memory can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some examples, the memory can further include a memory remotely arranged with respect to the processor, which can be connected to the computer terminal A through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0177] The processor can call the information and application programs stored in the memory through the transmission device to execute the following steps: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result; predicting a second test result corresponding to the test information based on the target subtext; and outputting the second test result, and the relevance to the test information is higher than a target threshold.

[0178] Optionally, the processor can further execute program codes of the following steps: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the relevance of the sentence to the test information; and determining at least one target subtext based on the first values of the plurality of sentences.

[0179] Optionally, the processor can further execute program codes of the following steps: analyzing the sentence and the test information based on the evidence ranking model to obtain the first value of the sentence, wherein the evidence ranking model is trained by the test information sample, the first subtext sample associated with the test information sample, and the second subtext sample not associated with the test information sample.

[0180] Optionally, the processor can further execute program codes of the following steps: determining at least one first value greater than a target threshold value from the first values of the plurality of sentences; and determining at least one sentence corresponding to the at least one first value as the at least one target subtext.

[0181] Optionally, the processor can further execute program codes of the following steps: predicting a second test result based on the target subtext and the test information.

[0182] Optionally, the processor can further execute program codes of the following steps: in the case that the at least one target subtext is a plurality of target subtexts, analyzing the plurality of target subtexts and the test information based on an evidence reasoning model to obtain a plurality of sub-test results, wherein the evidence reasoning model is trained by the test information sample, the first subtext sample associated with the test information sample, and the test result sample; and determining the second test result based on the plurality of sub-test results.

[0183] Optionally, the processor can further execute program codes of the following steps: determining the second test result based on the plurality of sub-test results and second values of the plurality of target subtexts, wherein the second value of the target subtext is used to indicate the relevance of the target subtext to the test information.

[0184] Optionally, the processor can further execute program codes of the following steps: weighting the plurality of sub-test results according to the second values of the plurality of target subtexts to obtain the second test result.

[0185] Optionally, the processor can further execute program codes of the following steps: obtaining a set of element information of the target test, wherein each element information in the set of element information is used to match at least one text in a target database.

[0186] As an optional implementation, the processor can call information and application programs stored in the memory through the transmission device to execute the following steps: displaying test information on the operation interface, wherein the test information is used to obtain target text associated with the test information in the target database, the target text includes at least one target subtext, the target subtext is used to represent the first test result, and the correlation between the target subtext and the test information is higher than a target threshold; and displaying the second test result corresponding to the test information in the browsing area of the operation interface, wherein the second test result is obtained based on the target subtext.

[0187] Optionally, the processor can further execute the program code of the following steps: obtaining a set of element information of the target test in response to the selection instruction acting on the operation interface before displaying the test information on the operation interface, wherein each element information in the set of element information is used to match at least one text in the target database; and displaying the set of element information on the operation interface.

[0188] As an optional implementation, the processor can call information and application programs stored in the memory through the transmission device to execute the following steps: obtaining first target language information, and extracting test information from the first target language information; obtaining target text associated with the test information in the target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent the first test result, and the correlation between the target subtext and the test information is higher than a target threshold; predicting the second test result corresponding to the test information based on the target subtext; and outputting second target language information carrying the second test result.

[0189] Optionally, the processor can further execute the program code of the following steps: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the correlation between the sentence and the test information; and determining the at least one target subtext based on the first values of the plurality of sentences.

[0190] Optionally, the processor can further execute the program code of the following steps: predicting the second test result based on the target subtext and the test information.

[0191] This invention provides an information processing method. It involves acquiring experimental information; obtaining target text associated with the experimental information from a target database; extracting at least one target subtext from the target text, wherein the target subtext represents a first experimental result and has a correlation with the experimental information higher than a target threshold; predicting a second experimental result corresponding to the experimental information based on the target subtext; and outputting the second experimental result. In other words, this application finds readily available relevant target text in the target database as judgment information, and uses target subtexts in the target text with a correlation higher than a target threshold to infer the result. This process is similar to the process of expert judgment of experimental results, avoiding the cost of explicitly annotating structured evidence. It also provides interpretability for predicting experimental results based on experimental information, and then predicts the second experimental result corresponding to the experimental information based on the target subtext and outputs the second experimental result. This solves the technical problem that predicting experimental results relies heavily on manual annotation, leading to high resource consumption, and achieves the technical effect of reducing the resources used in predicting experimental results.

[0192] Those skilled in the art will understand that Figure 10 The structure shown is for illustrative purposes only. The computer terminal can also be a smartphone (such as an Android phone, an iOS phone, etc.), a tablet computer, a mobile internet device (MID), a PAD, and other terminal devices. Figure 10 This does not limit the structure of the aforementioned computer terminal. For example, computer terminal A may also include components that are more complex than those described above. Figure 10 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 10 The different configurations shown.

[0193] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0194] Example 4

[0195] Embodiments of the present invention also provide a computer-readable storage medium. Optionally, in this embodiment, the computer-readable storage medium can be used to store the program code executed by the information processing method provided in Embodiment 1.

[0196] Optionally, in the embodiment, the computer readable storage medium can be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.

[0197] Optionally, in the embodiment, the computer readable storage medium is configured to store program code for performing the following steps: obtaining the test information; obtaining the target text associated with the test information in the target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent the first test result and has a higher relevance to the test information than a target threshold; predicting the second test result corresponding to the test information based on the target subtext; and outputting the second test result.

[0198] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the relevance of the sentence to the test information; and determining at least one target subtext based on the first values of the plurality of sentences.

[0199] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: analyzing the sentence and the test information based on an evidence ranking model to obtain the first value of the sentence, wherein the evidence ranking model is trained by a test information sample, a first subtext sample associated with the test information sample, and a second subtext sample not associated with the test information sample.

[0200] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: determining at least one first value greater than the target threshold from the first values of the plurality of sentences; and determining at least one sentence corresponding to the at least one first value as the at least one target subtext.

[0201] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: predicting the second test result based on the target subtext and the test information.

[0202] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: in the case that the at least one target subtext is a plurality of target subtexts, analyzing the plurality of target subtexts and the test information based on an evidence reasoning model to obtain a plurality of sub-test results, wherein the evidence reasoning model is trained by a test information sample, a first subtext sample associated with the test information sample, and a test result sample; and determining the second test result based on the plurality of sub-test results.

[0203] Optionally, the computer readable storage medium is further arranged to store program code for performing the following step: determining the second test result based on the plurality of sub-test results and a second number of target sub-texts, wherein the second number of target sub-texts is used to indicate a relevance of the target sub-texts to the test information.

[0204] Optionally, the computer readable storage medium is further arranged to store program code for performing the following step: weighting the plurality of sub-test results according to the second number of target sub-texts to obtain the second test result.

[0205] Optionally, the computer readable storage medium is further arranged to store program code for performing the following step: obtaining a set of element information of the target test, wherein each element information in the set of element information is used to match at least one text in the target database.

[0206] As an optional implementation, the computer readable storage medium is arranged to store program code for performing the following steps: displaying the test information on the operation interface, wherein the test information is used to obtain a target text associated with the test information in the target database, the target text comprises at least one target sub-text, the target sub-text is used to represent the first test result, and the relevance of the target sub-text to the test information is higher than a target threshold; displaying the second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is predicted based on the target sub-text.

[0207] Optionally, the computer readable storage medium is further arranged to store program code for performing the following steps: before displaying the test information on the operation interface, obtaining a set of element information of the target test in response to a selection instruction acting on the operation interface, wherein each element information in the set of element information is used to match at least one text in the target database; displaying the set of element information on the operation interface.

[0208] As an optional implementation, the computer readable storage medium is arranged to store program code for performing the following steps: obtaining the first target language information, and extracting the test information from the first target language information; obtaining a target text associated with the test information in the target database; extracting at least one target sub-text from the target text, wherein the target sub-text is used to represent the first test result, and the relevance of the target sub-text to the test information is higher than a target threshold; predicting a second test result corresponding to the test information based on the target sub-text; and outputting the second target language information carrying the second test result.

[0209] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: splitting the target text to obtain a plurality of sentences; determining a first value of the sentences based on the sentences and the test information, wherein the first value of the sentences is used to indicate a relevance of the sentences to the test information; and determining at least one target subtext based on the first values of the plurality of sentences.

[0210] Optionally, the computer readable storage medium is further configured to store program code for performing the following steps: predicting a second test result based on the target subtext and the test information.

[0211] The above-mentioned sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages or disadvantages of the embodiments.

[0212] In the above-mentioned embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0213] In several embodiments provided in the present application, it should be understood that the disclosed technical contents can be implemented by other manners. Among them, the above-mentioned device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or modules shown or discussed can be indirect coupling or communication connection through some interfaces, units or modules, which can be electrical or other forms.

[0214] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the present embodiment.

[0215] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0216] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0217] The above is only the preferred embodiment of the present application, it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. An information processing method characterized by comprising: The method comprises: obtaining test information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result and has a relevance to the test information higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; outputting the second test result; wherein the obtaining test information comprises: obtaining a set of element information of a target test, wherein each element information in the set of element information is used to match at least one text in the target database.

2. The method of claim 1, wherein, The extracting at least one target subtext from the target text comprises: splitting the target text to obtain a plurality of sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence is used to indicate the relevance of the sentence to the test information; determining at least one target subtext based on the first values of the plurality of sentences.

3. The method of claim 2, wherein, The determining a first value of the sentence based on the sentence and the test information comprises: analyzing the sentence and the test information based on an evidence ranking model to obtain the first value of the sentence, wherein the evidence ranking model is trained by a test information sample, a first subtext sample associated with the test information sample, and a second subtext sample not associated with the test information sample.

4. The method of claim 2, wherein, The determining at least one target subtext based on the first values of the plurality of sentences comprises: determining at least one first value greater than the target threshold from the first values of the plurality of sentences; determining at least one sentence corresponding to the at least one first value as the at least one target subtext.

5. The method of claim 1, wherein, The predicting a second test result corresponding to the test information based on the target subtext comprises: predicting the second test result based on the target subtext and the test information.

6. The method of claim 5, wherein, The predicting a second test result corresponding to the test information based on the target subtext comprises: in a case that the at least one target subtext is a plurality of target subtexts, analyzing the plurality of target subtexts and the test information based on an evidence reasoning model respectively to obtain a plurality of sub-test results, wherein the evidence reasoning model is trained by a test information sample, a first subtext sample associated with the test information sample, and a test result sample; determining the second test result based on the plurality of sub-test results.

7. The method of claim 6, wherein, The determining the second test result based on the plurality of sub-test results comprises: determining the second test result based on the plurality of sub-test results and second values of the plurality of target subtexts, wherein the second values of the target subtexts are used to indicate the relevance of the target subtexts to the test information.

8. The method of claim 7, wherein, The determining the second test result based on the plurality of sub-test results and second values of the plurality of target subtexts comprises: weighting the plurality of sub-test results according to the second values of the plurality of target subtexts to obtain the second test result.

9. The method of claim 1, wherein, The set of element information includes at least one of the following: background information of the target test; subject information of the target test; treatment information of the target test; reference information of the target test; reference index of the target test.

10. An information processing method characterized by comprising: The method comprises: displaying test information on an operation interface, wherein the test information is used to obtain target text associated with the test information in a target database, the target text includes at least one target subtext, the target subtext is used to represent a first test result, and the relevance of the target subtext to the test information is higher than a target threshold; displaying a second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is predicted based on the target subtext; wherein the test information is a set of element information of a target test, and each element information in the set of element information is used to match at least one text in the target database.

11. The method of claim 10, wherein, before displaying the test information on the operation interface, the method further comprises: obtaining a set of element information of a target test in response to a selection instruction acting on the operation interface; displaying the test information on the operation interface comprises: displaying the set of element information on the operation interface.

12. An information processing method characterized by comprising: The method comprises: obtaining first target language information and extracting test information from the first target language information; obtaining target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and the relevance of the target subtext to the test information is higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; outputting second target language information carrying the second test result; wherein the test information is a set of element information of a target test, and each element information in the set of element information is used to match at least one text in the target database.

13. The method of claim 12, wherein, extracting at least one target subtext from the target text comprises: splitting the target text into multiple sentences; determining a first value of the sentence based on the sentence and the test information, wherein the first value of the sentence indicates the relevance of the sentence to the test information; determining at least one target subtext based on the first values of the multiple sentences.

14. The method of claim 12, wherein, predicting a second test result corresponding to the test information based on the target subtext comprises: predicting the second test result based on the target subtext and the test information.

15. An information processing apparatus comprising: The method comprises: a first obtaining unit configured to obtain test information; a second obtaining unit configured to obtain target text associated with the test information in a target database; a first extracting unit configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and the relevance of the target subtext to the test information is higher than a target threshold; a first predicting unit configured to predict a second test result corresponding to the test information based on the target subtext; The first output unit is configured to output the second test result. The first obtaining unit is configured to obtain test information by the following steps: obtaining a set of element information of a target test, wherein each element information in the set of element information is used to match at least one text in the target database.

16. An information processing apparatus comprising: The method comprises the following steps: The first display unit is configured to display test information on an operation interface, wherein the test information is used to obtain a target text associated with the test information in a target database, the target text comprises at least one target subtext, the target subtext is used to represent a first test result, and the relevance between the target subtext and the test information is higher than a target threshold; The second display unit is configured to display a second test result corresponding to the test information in a browsing area of the operation interface, wherein the second test result is predicted based on the target subtext. The test information is a set of element information of a target test, and each element information in the set of element information is used to match at least one text in the target database.

17. An information processing apparatus comprising: The method comprises the following steps: The third obtaining unit is configured to obtain first target language information and extract test information from the first target language information; The fourth obtaining unit is configured to obtain a target text associated with the test information in a target database; The second extracting unit is configured to extract at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and the relevance between the target subtext and the test information is higher than a target threshold; The second predicting unit is configured to predict a second test result corresponding to the test information based on the target subtext; The second output unit is configured to output second target language information carrying the second test result; The test information is a set of element information of a target test, and each element information in the set of element information is used to match at least one text in the target database.

18. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored program, wherein the program controls the device where the computer readable storage medium is located to execute the method of any one of claims 1-14 when the program is run by the processor.

19. A processor, comprising: The processor is configured to run a program, wherein the program executes the method of any one of claims 1-14 when the program is run.

20. An information processing system, characterized by comprising: The method comprises the following steps: A processor; A memory connected to the processor, configured to provide the processor with instructions for processing the following processing steps: obtaining test information; Obtaining a target text associated with the test information in a target database; extracting at least one target subtext from the target text, wherein the target subtext is used to represent a first test result, and the relevance between the target subtext and the test information is higher than a target threshold; predicting a second test result corresponding to the test information based on the target subtext; outputting the second test result; wherein the test information comprises a set of element information of a target test, and each element information in the set of element information is used to match at least one text in the target database.

Images