Method and device for generating push information

Abstract

The present application discloses a method, device, electronic device and computer-readable storage medium for generating push information, and relates to the fields of artificial intelligence, natural language processing technology, knowledge graph technology and big data technology. The specific implementation scheme is: obtaining standard representation information corresponding to the representation information in the user's input information; determining at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, and the correspondence is extracted from the abstract information of the medical literature; generating a push information set based on the medical state entity, sending the push information set to the user, and using the knowledge graph constructed based on the abstract information of the medical literature to determine the push message pushed to the user, thereby reducing the push message determination cost while improving the quality of the push message.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, specifically to the field of natural language processing technology, knowledge graph technology and big data technology, and in particular to a method, device, electronic device and computer-readable storage medium for generating push information. Background Art

[0002] With the development of society, in order to better meet the search needs of users, more and more methods based on Internet big data and knowledge graphs are used to match the query information input by users and generate corresponding push messages to provide search services for users.

[0003] In the prior art, when determining a knowledge graph, it is usually constructed based on domain expert knowledge. Summary of the invention

[0004] The present application provides a method, device, electronic device and storage medium for generating push information.

[0005] In a first aspect, an embodiment of the present application provides a method for generating push information, comprising: obtaining standard representation information corresponding to the representation information in the user's input information; determining at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, and the correspondence is extracted from the summary information of the medical literature; generating a push information set based on the medical state entity, and sending the push information set to the user.

[0006] In some embodiments, the medical knowledge graph is determined based on the following steps: obtaining summary text information of multiple medical documents to obtain a summary text information set; using an entity recognition neural network to determine a hit entity set in the summary text information set; wherein the entity set includes the following information in the summary text information set: information related to representation information and medical status entities; performing medical language normalization matching on the entity set to obtain a normalized entity set; classifying and labeling the normalized entities in the normalized entity set to obtain a representation information set and a medical status entity set; and obtaining the medical knowledge graph based on the co-occurrence relationship between the representation information in the representation information set and the medical status entities in the medical status entity set.

[0007] In some embodiments, the entity recognition neural network includes: a bidirectional short-term memory network and a conditional random field.

[0008] In some embodiments, a push information set is generated based on the medical status entity, and sending the push information set to the user includes: sorting the medical status entities using a probabilistic graph model, selecting a preset number of the medical status entities to generate a push information set based on the sorting results; and sending the push information set to the user.

[0009] In some embodiments, the step of generating standard representation information includes: obtaining user input information; identifying the representation information contained in the input information to obtain a recognition result; and determining the standard representation information based on the normalized semantics of the recognition result.

[0010] In some embodiments, based on the normalized semantics of the recognition result, determining the standard representation information includes: performing expansion based on the normalized semantics of the recognition result to generate an extended representation information set; and using the extended representation information in the extended representation information set as the standard representation information.

[0011] In some embodiments, based on a preconstructed medical knowledge graph, determining at least one medical state entity hit by the standard characterization information includes: in response to determining that selection information for the standard characterization information is received, using the preconstructed medical knowledge graph to determine at least one medical state entity hit by the standard characterization information.

[0012] In the second aspect, an embodiment of the present application provides a push information generation device, comprising: a standard representation information acquisition unit, configured to acquire standard representation information corresponding to the representation information in the user's input information; a medical state entity determination unit, configured to determine at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, and the correspondence is extracted from the summary information of the medical literature; a push information sending unit, configured to generate a push information set based on the medical state entity, and send the push information set to the user.

[0013] In some embodiments, a medical knowledge graph determination unit is also included, which specifically includes: an initial information acquisition subunit, configured to acquire summary text information of multiple medical documents to obtain a summary text information set; an entity recognition subunit, configured to use an entity recognition neural network to determine a hit entity set in the summary text information set; wherein the entity set includes the following information in the summary text information set: information related to representation information and medical state entities; a standard matching subunit, configured to perform medical language standardization matching on the entity set to obtain a standardized entity set; a classification and annotation subunit, configured to perform classification and annotation on the standardized entities in the standardized entity set to obtain a representation information set and a medical state entity set; a medical knowledge graph generation subunit, configured to obtain the medical knowledge graph based on the co-occurrence relationship between the representation information in the representation information set and the medical state entities in the medical state entity set.

[0014] In some embodiments, the entity recognition neural network in the entity recognition subunit includes: a bidirectional short-term memory network and a conditional random field.

[0015] In some embodiments, the push information sending unit is further configured to: sort the medical status entities using a probabilistic graph model, select a preset number of the medical status entities according to the sorting result to generate a push information set; and send the push information set to the user.

[0016] In some embodiments, a standard information generation unit is also included, including: an initial information acquisition subunit, configured to obtain the user's input information; an information identification subunit, configured to identify the representation information contained in the input information and obtain a recognition result; a standard representation information determination subunit, configured to determine the standard representation information based on the normalized semantics of the recognition result.

[0017] In some embodiments, the standard representation information determination subunit is further configured to: perform expansion based on the normalized semantics of the recognition result to generate an extended representation information set; and use the extended representation information in the extended representation information set as the standard representation information.

[0018] In some embodiments, the medical state entity determination unit is further configured to: in response to determining that selection information for the standard representation information is received, use a pre-constructed medical knowledge graph to determine at least one medical state entity hit by the standard representation information.

[0019] In a third aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for generating push information described in any implementation manner.

[0020] In a fourth aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium storing computer instructions, including: the computer instructions are used to enable the computer to execute the method for generating push information described in any implementation manner.

[0021] After obtaining standard representation information corresponding to the representation information in the user's input information, the present application determines at least one medical status entity hit by the standard representation information based on a pre-constructed medical knowledge graph, wherein the medical knowledge graph records the correspondence between the representation information and the medical status entity, and the correspondence is extracted from the abstract information of the medical literature. A push information set is generated based on the medical status entity, and the push information set is sent to the user. The knowledge graph constructed based on the abstract information of the medical literature is used to determine the push message to be pushed to the user, thereby reducing the push message determination cost while improving the quality of the push message.

[0022] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application.

[0024] Figure 1 is an exemplary system architecture in which embodiments of the present application may be applied;

[0025] Figure 2 is a flowchart of an embodiment of a method for generating push information according to the present application;

[0026] Figure 3 It is a flowchart of an implementation method for determining a medical knowledge graph in the method for generating push information according to the present application;

[0027] Figure 4 is a flowchart of another embodiment of a method for generating push information according to the present application;

[0028] Figure 5 is a structural diagram of an embodiment of a device for generating push information according to the present application;

[0029] Figure 6 It is a block diagram of an electronic device suitable for implementing the method for generating push information according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0031] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0032] Figure 1 An exemplary system architecture 100 is shown to which embodiments of the method, apparatus, electronic device, and computer-readable storage medium for generating push information of the present application can be applied.

[0033] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0034] The user can use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to achieve the purpose of sending the user's input information, etc. The terminal devices 101, 102, 103 can be installed with search applications, such as navigation applications, encyclopedia query applications, online consultation applications, etc.

[0035] Terminal devices 101, 102, 103 can be hardware or software. When it is hardware, it can be various electronic devices with display screens, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, etc. When terminal devices 101, 102, 103 are software, they can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules (for example, sending user input information, etc.), or it can be implemented as a single software or software module. No specific limitation is made here.

[0036] The server 105 may be a server that provides various services, such as a server that provides retrieval services and generates push information for the terminal devices 101, 102, and 103. For example, the server may obtain standard representation information corresponding to the representation information in the user's input information; determine at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph is determined based on the relationship between the representation information and the medical state entity in the medical literature; generate a push information set based on the medical state entity, and send the push information set to the user.

[0037] It should be noted that the method for generating push information provided in the embodiments of the present application is generally executed by the server 105 , and accordingly, the device for generating push information is generally disposed in the server 105 .

[0038] It should be noted that the server can be hardware or software. When the server is hardware, it can be implemented as a distributed server cluster consisting of multiple servers, or it can be implemented as a single server. When the server is software, it can be implemented as multiple software or software modules for providing distributed services, or it can be implemented as a single software or software module. No specific limitation is made here.

[0039] In addition, the method for generating push information may also be executed by the terminal devices 101, 102, 103, and accordingly, the apparatus for generating push information may also be arranged in the terminal devices 101, 102, 103. In this case, the exemplary system architecture 100 may also not include the server 105 and the network 104.

[0040] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0041] Continue to refer Figure 2 , which shows a process 200 of an embodiment of a method for generating push information according to the present application. The method for generating push information includes the following steps:

[0042] Step 201: Acquire standard representation information corresponding to the representation information in the user's input information.

[0043] In this embodiment, the execution subject of the method for generating push information (for example Figure 1 The server 105 shown in the figure can be used from a local or non-local human-computer interaction device (e.g. Figure 1 The terminal devices 101, 102, 103) shown in the figure obtain the user's input information and the standard representation information corresponding to the representation information in the user's input information, which is not limited in this application.

[0044] Among them, representational information is the way information is presented in thinking systems such as the brain and computer systems, the way information is recorded or expressed, a formal system that can clearly express certain entities or certain types of information, and several rules that explain how the system performs its functions. Therefore, we can understand that representational information refers to symbols or signals that can refer to something, that is, when a certain thing is absent, it represents the relevant information of the thing. Entities generally refer to entities with specific meanings or strong referentiality in the text, usually including names of people, places, names of organizations, dates and times, proper nouns, etc. Therefore, the concept of entity can be very broad, as long as it is a special text fragment required by the business, it can be called an entity.

[0045] It should be understood that the user's input information is usually input information obtained based on the user's own cognitive level and cultural differences, and contains one or more representation information of the user's true meaning. After obtaining these representation information, it is necessary to convert this information into standard representation information that can be recognized and understood by the above-mentioned execution subject. Because when the above-mentioned execution subject is pre-trained, the training model usually used is a standard expression form provided by an authoritative official in the corresponding field, so the above-mentioned standard representation information is a standard expression form provided by an authoritative official in each field. For example, in the medical field, the user's input information is stomach pain, and this content is not a standard expression in the medical field. Therefore, it is converted into standard expressions in the medical field such as abdominal pain, stomach pain, gastric colic, and gastric cramps to obtain standard representation information.

[0046] Similarly, the executor of the method for generating push information can process the user's input information locally after obtaining it to obtain standard representation information corresponding to the representation information in the input information, or can directly obtain the corresponding standard representation information obtained by other non-local terminal devices after processing the representation information in the user's input information.

[0047] Step 202, based on a pre-constructed medical knowledge graph, determine at least one medical state entity hit by the standard representation information; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, and the correspondence is extracted from the abstract information of the medical literature.

[0048] In this embodiment, the standard representation information obtained in the above step 201 is matched according to a pre-constructed medical knowledge graph that records the correspondence between the representation information and the medical state entities to determine one or more medical state entities hit by the standard representation information.

[0049] Among them, the correspondence between the representation information and medical status entities recorded in the medical knowledge graph is extracted based on the abstract information in multiple medical documents. Taking the abstract document of a medical document as an example, if the first representation information and the first medical status entity appear at the same time, it is considered that there is a correspondence between the first representation information and the first medical status entity. The medical knowledge graph is obtained based on the correspondence between multiple representation information and medical status entities existing in the abstract information of multiple medical documents.

[0050] It should be understood that there may be multiple representation information and medical status entities in the abstract information of a medical document. In this case, there may be multiple corresponding relationships between one representation information and multiple medical status entities and / or there may be multiple corresponding relationships between multiple representation information and one medical status entity. These corresponding relationships are also recorded in the process of generating the medical knowledge graph.

[0051] Step 203: Generate a push information set based on the medical status entity, and send the push information set to the user.

[0052] In this embodiment, based on the above step 203, one or more medical status entities hit by standard representation information can be obtained according to the pre-constructed medical knowledge graph. When there are multiple hit medical status entities, a medical status entity that meets the requirements can be selected from the multiple hit medical status entities according to a predetermined screening rule, and the multiple hit medical status entities can be sorted, and a push information set is generated according to the obtained one or more medical status entities, that is, the push information set contains one or more medical status entities, and then the push information set is sent to the user who inputs information in step 201 to determine the final push content pushed to the user, so that the user can obtain the push information generated based on the input information.

[0053] The method for generating push information provided in the embodiment of the present application, after obtaining standard representation information corresponding to the representation information in the user's input information, determines at least one medical status entity hit by the standard representation information based on a pre-constructed medical knowledge graph, wherein the medical knowledge graph records the correspondence between the representation information and the medical status entity, and the correspondence is extracted from the abstract information of the medical literature. A push information set is generated based on the medical status entity, and the push information set is sent to the user. The knowledge graph constructed based on the abstract information of the medical literature is used to determine the push message to be pushed to the user, thereby reducing the push message determination cost while improving the quality of the push message.

[0054] In some optional implementations of this embodiment, reference Figure 3, which shows a process 300 of determining steps of a medical knowledge graph, specifically including:

[0055] Step 301, obtain summary text information of multiple medical documents to obtain a summary text information set.

[0056] Specifically, a large amount of medical literature can be obtained through medical literature retrieval libraries such as Pubmed and Chinese Biomedical Literature. After obtaining the medical literature, abstract text information is extracted to obtain an abstract text information set.

[0057] The summary text information can be in English or Chinese. It is preferred to use English summary information for extraction because the use of English summary information can avoid the problem of word segmentation on the text content when using Chinese summary text information, thereby further improving the generation efficiency of the summary text information set.

[0058] It should be understood that when the abstract text information of the acquired medical literature is extracted, the title information of the medical literature may also be extracted for subsequent reference.

[0059] Step 302, using an entity recognition neural network to determine a hit entity set in the summary text information set; wherein the entity set includes the following information in the summary text information set: information related to the representation information and the medical status entity.

[0060] Specifically, the entity recognition neural network extracts entities from unstructured input text and can identify more categories of entities according to business needs, such as product names, models, prices, etc. For example, DeepWeb, NER and other entity recognition neural networks perform entity recognition on the content in the summary text information set to determine the hit entities in the summary text information set, where the entity information includes information related to representation information and medical status entities, and finally obtains the entity set based on the hit entities.

[0061] It should be understood that entities satisfying the preset rules may be selected from the hit entities according to the preset rules, and then an entity set may be obtained based on the entities satisfying the preset rules.

[0062] Step 303: Perform medical language normalization matching on the entity set to obtain a normalized entity set.

[0063] Specifically, although the entity recognition neural network can identify medical entities, the recognized entity names are not necessarily standardized (for example, "diarrhea" is not standardized, "diarrhea" is standardized), and the categories are not necessarily correct (for example, the category of "amoxicillin" is identified as "drug"). After the corpus passes through the entity recognition neural network, a set of entity candidates is obtained. In order to automatically standardize the entities, an approximate search is performed on the entity names and their synonyms in medical knowledge databases such as the UMLS library. The matched entity is a standardized entity, and then the matching entity is added with the corresponding entity code, such as using the CUI code to uniquely identify this standardized candidate entity, and a standardized entity set is obtained. In addition, existing methods such as key-value pair encoding can also be used to achieve this purpose.

[0064] Step 304 , classify and label the normalized entities in the normalized entity set to obtain a representation information set and a medical status entity set.

[0065] Specifically, since there is no comprehensive definition of the types of entities (representational information, medical status entity) in the medical knowledge database, it is impossible to distinguish whether the obtained entity is representational information or a medical status entity. Therefore, matching can be performed based on the encoding information in the above steps, and the obtained standardized entities can be classified and labeled through existing medical knowledge information, medical knowledge graphs, etc. to determine whether the entity is representational information, medical status entity, or neither.

[0066] Step 305: obtaining the medical knowledge graph based on the co-occurrence relationship between the representation information in the representation information set and the medical state entities in the medical state entity set.

[0067] Specifically, the number of times the medical state entity and the representation information appear together in the abstract text information set can be determined based on the number of times the medical state entity appears and the number of times the medical state entity and the representation information appear in the same medical literature abstract to determine whether there is a corresponding relationship between the representation information and the medical state entity. For example, whether the probability of medical state entity A appearing in the abstract text information set is similar to whether the co-occurrence probability of the medical state entity and the representation information B in the abstract text information is determined to determine whether there is a corresponding relationship between the medical state entity A and the representation information B. Or, for example, a co-occurrence threshold condition is predetermined, and when the co-occurrence number of representation information D and the medical state entity C meets the predetermined threshold condition, it is considered that there is a corresponding relationship between the representation information D and the medical state entity C, and the medical knowledge graph is obtained based on the corresponding relationship between the collected representation information and the medical state entities.

[0068] In this implementation, the method of generating a medical knowledge graph based on the summary information of medical literature can firstly achieve a wide range of representation information and entity information coverage through a large amount of medical literature, avoiding the problem of narrow information coverage in the prior art method of generating knowledge graphs based on expert knowledge, and can simplify medical literature through summary information, avoiding the technical problem of low recognition efficiency of medical literature due to excessive content information in medical literature, thereby improving the efficiency of generating knowledge graphs.

[0069] In some optional implementations of this embodiment, the entity recognition neural network includes: a bidirectional short-term memory network and a conditional random field.

[0070] Specifically, the biggest difference between the bidirectional short-term memory network, namely the Bi-LSTM (Bi-directional Long Short-Term Memory, referred to as Bi-LSTM) memory network and the traditional neural network is that the input of the hidden layer not only includes the output of the input layer, but also includes the output of the hidden layer at the previous moment. Its main feature is that it can store information at previous moments. Although RNN (Recurrent Neural Network, referred to as RNN) can theoretically retain all the above information, as the number of hidden layers increases, there is a phenomenon of gradient disappearance or gradient explosion. LSTM (Long Short-Term Memory, referred to as LSTM) can effectively solve the problem of long-term dependency, including forgetting gate, input gate and output gate. In order to make the network express more information and make the inference more accurate, the study adopted a bidirectional network structure, namely Bi-LSTM. Bi-LSTM is composed of two LSTM spliced ​​together, including a forward input sequence and a reverse input sequence, and considers past features and future features at the same time.

[0071] The advantage of LSTM is that it can learn the dependencies between observation sequences (input words) through a bidirectional setting. During the training process, LSTM can automatically extract the features of the observation sequence according to the goal (such as identifying entities). However, its disadvantage is that it cannot learn the relationship between state sequences (output labels). You must know that in the named entity recognition task, there is a certain relationship between labels. For example, an E-type label (indicating the beginning of an entity) will not be followed by another E-type label. Therefore, when solving the sequence labeling task, although LSTM can save very complicated feature engineering, it also has the disadvantage of not being able to learn the labeling context.

[0072] When Bi-LSTM is used for named entity recognition, the output of Bi-LSTM is the score of the entity label, and the label corresponding to the highest score is selected. However, sometimes Bi-LSTM cannot get the correct entity label. At this time, it is necessary to add conditional random field, namely CRF (Conditional Random Field, referred to as CRF). CRF combines the characteristics of maximum entropy model and hidden Markov model, can model the hidden state, and learn the characteristics of state sequence, but its disadvantage is that it needs to extract sequence features manually.

[0073] Therefore, when the bidirectional short-term memory network and conditional random field are used together, the shortcomings that exist when they are used separately can be avoided, so as to obtain the advantages of both at the same time and achieve higher-quality entity recognition.

[0074] In some optional implementations of the present embodiment, a push information set is generated based on the medical status entity, and sending the push information set to the user includes: sorting the medical status entities using a probabilistic graph model, and selecting a preset number of the medical status entities to generate a push information set based on the sorting results; and sending the push information set to the user.

[0075] Specifically, after obtaining the entity category, the correlation between entities can be measured by the number of co-occurrences between the representation information and the medical status entity, for example, using the formula:

[0076]

[0077] Among them, P(Dis i |Sym j ) represents the probability of correlation between the representation information and the medical status entity, P(Dis i , Sym j ) represents the probability based on the co-occurrence count between the representation information and the medical status entity, P(Sym j ) represents the probability of the representation information appearing. More specifically, the above formula describes the contribution of the jth representation information (Sym) to the ith medical status entity (Dis). For example, if "cough" appears 100 times in 10 million documents, and "tuberculosis" and "cough" appear together (#co_occurrence) 5 times in an abstract, then the contribution of cough to tuberculosis is: P(tuberculosis|cough) = 0.05. To simplify the calculation, it is assumed that symptoms are independent, so the representation information probability graph model (naive Bayes) is obtained, and the posterior probability of the medical status entity is as follows:

[0078] P(Dis i |Sym j ,Sym j+1 ,...)=Ρ(Disi )·P(Dis i |Sym j )·P(Dis i |Sym j+1 )...

[0079] Where P(Dis) is the prior probability of the medical status entity. In order to avoid the result being too small due to continuous multiplication, we can take the logarithm of both sides of the above formula and change it to an accumulation operation. The above formula can be used to calculate the posterior probability of each medical status entity when some characterization information is known.

[0080] After obtaining the posterior probability of each medical status entity, the medical status entities can be sorted according to the size of the probability and a preset number of medical status entities can be selected to generate a push information set, so that users can obtain medical status entities with a higher hit probability, further improving the quality of push information.

[0081] In some optional implementations of this embodiment, the step of generating standard characterization information includes:

[0082] The user's input information is obtained; the representation information contained in the input information is identified to obtain a recognition result; and the standard representation information is determined based on the normalized semantics of the recognition result.

[0083] Specifically, the word "gui" can be used to represent the common form of expression of some similar descriptions, and the normalized semantics can be to replace the template word in the entry information with the word "gui" to unify the form of expressions with the same different semantics. Therefore, a standard representation information database can be built in advance based on the existing information, and then after obtaining the recognition result of the user input information, the recognition result is normalized to obtain standard representation information to prevent the above-mentioned execution entity from being unable to understand the user's ideas and needs when the user is unable to use a more standardized and standard description language to accurately express his or her ideas and needs, so as to smoothly generate push information that meets the user's needs based on his or her ideas and needs.

[0084] Continue to refer Figure 4 , which shows a process 400 of another embodiment of a method for generating push information, which specifically includes the following steps:

[0085] Step 401: Obtain user input information.

[0086] Step 402: Identify the representation information contained in the input information to obtain a recognition result.

[0087] Specifically, the above Figure 3 The entity recognition neural network in the corresponding implementation recognizes the user's input information to determine the representation information present therein.

[0088] Step 403: Expand based on the normalized semantics of the recognition result to generate an extended representation information set.

[0089] Specifically, since it is difficult for the user's input information to include standard representation information, based on the above Figure 2 After the normalized semantic results are obtained in some implementations of the illustrated embodiments, similar expansions can be performed based on the semantic results. For example, if the normalized semantic result obtained is "gastric colic", it can be similarly expanded to the same type of "gastric cramps" based on the content to obtain more reference information related to the user's input information, that is, to expand the representation information, so as to facilitate the subsequent representation information set obtained based on these representation information to obtain more medical status entities, thereby improving the quality of the generated push information set.

[0090] Among them, in the use of Figure 2 In some implementations of the embodiments shown in the figure, when determining the technical solution of the push set based on the probabilistic graph model, the probabilistic graph model may also be used:

[0091]

[0092] P(Sym i |Sym j ,Sym j+1 ,...)=Ρ(Sym i )·P(Sym i |Sym j )·P(Sym i |Sym j+1 )...

[0093] To determine the extended characterization information, where P(Sym j 、Sym j+1 …) represents the probability of occurrence of different representation information, which is equivalent to calculating whether there is a correlation between different representation information according to the co-occurrence of different representation information. The specific principle is similar to the above-mentioned process of determining the correspondence between the representation information and the medical status entity based on the co-occurrence relationship between the representation information and the medical status entity, and will not be repeated here.

[0094] Step 404: Use the extended representation information in the extended representation information set as standard representation information.

[0095] Step 405 , in response to determining that selection information for the standard representation information is received, using a pre-constructed medical knowledge graph to determine at least one medical state entity hit by the standard representation information.

[0096] It should be understood that the number of standard characterization information here may be one message or multiple messages. When there are multiple pieces of standard characterization information, these standard characterization information can be presented to the user who inputs the information, and the selection information generated by the user based on the selected standard characterization information can be obtained. The standard characterization information included in the selection information, that is, the standard characterization information selected by the user, is determined, and the standard characterization information expected by the user is obtained in a human-computer interaction manner. The standard characterization information based on the user's selection can better meet the user's needs, so as to improve the quality of the subsequently generated push information.

[0097] Step 406: Generate a push information set based on the medical status entity, and send the push information set to the user.

[0098] Specifically, if the standard representation information in the selection information sent by the user in step 405 hits multiple medical status entities, a push information set is generated based on the multiple medical status entities and sent to the user. When the push information set is generated based on the multiple medical status entities, the medical status entities can be sorted and filtered according to predetermined rules, for example Figure 2 In the implementation of the illustrated embodiment, sorting is performed based on the probability of the characterization information hitting the medical status entity.

[0099] In this embodiment, part of the contents in steps 404 and 405 are the same as Figure 2 Steps 202-203 in the illustrated embodiment are similar, and repeated contents are not repeated here. In the present embodiment, after obtaining the user's input information, an extended characterization information set is determined according to the recognition result of the characterization information included in the input information, and the extended characterization set determines the standard characterization information. Then, based on the result of human-computer interaction with the user, the final selected standard characterization information is determined, and the medical status entity is determined accordingly, and a push information set is generated and pushed to the user, so as to provide the user with push information of higher quality and closer to the user's actual needs according to the user's actual needs.

[0100] To deepen understanding, this application also provides a specific implementation scheme in combination with a specific application scenario. In this specific application scenario, the user's input information is "diarrhea last night and this morning", and the predetermined number of medical status entity extractions is three.

[0101] After obtaining the information input by the user, an entity recognition neural network is used to identify the user's input information to determine the representation information "diarrhea" therein, and then the normalized semantic "diarrhea" is performed based on the recognition result, and the extended representation information "abdominal pain", "abdominal bloating", "indigestion" and "gastric colic" is expanded to present to the user.

[0102] In response to the user's selection information for the standard representation information "diarrhea", which includes the extended representation information "abdominal pain" and "indigestion", the medical status entities of "diarrhea", "abdominal pain" and "indigestion" are determined based on the pre-constructed knowledge graph. The medical knowledge graph records the correspondence between the representation information and the medical status entities, and the correspondence is extracted from the summary information of the medical literature.

[0103] The hit medical status entities are "irritable bowel syndrome", "lactose intolerance", "gastroparesis", "celiac disease", "gastritis" and "peptic ulcer". After sorting these medical status entities using the probabilistic graph model, the sorting relationship is: "gastritis", "peptic ulcer", "irritable bowel syndrome", "lactose intolerance", "celiac disease" and "gastroparesis".

[0104] Therefore, three medical status entities, namely "gastritis", "peptic ulcer" and "lactose intolerance" are extracted, a push information set is generated, and pushed to the user.

[0105] It can be seen from this application scenario that the method for generating push information provided in the embodiment of the present application, after obtaining standard representation information corresponding to the representation information in the user's input information, determines at least one medical status entity hit by the standard representation information based on a pre-constructed medical knowledge graph, wherein the medical knowledge graph records the correspondence between the representation information and the medical status entity, and the correspondence is extracted from the abstract information of the medical literature. A push information set is generated based on the medical status entity, and the push information set is sent to the user. The knowledge graph constructed based on the abstract information of the medical literature is used to determine the push message to be pushed to the user, thereby reducing the push message determination cost while improving the quality of the push message.

[0106] like Figure 5 As shown, the push information generation device 500 of this embodiment may include: a standard representation information acquisition unit 501, configured to obtain standard representation information corresponding to the representation information in the user's input information; a medical state entity determination unit 502, configured to determine at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, and the correspondence is extracted from the abstract information of the medical literature; a push information sending unit 503, configured to generate a push information set based on the medical state entity, and send the push information set to the user.

[0107] In some optional implementations of the present embodiment, the above-mentioned push information generation device also includes: a medical knowledge graph determination unit, including: an initial information acquisition subunit, configured to obtain summary text information of multiple medical documents to obtain a summary text information set; an entity recognition subunit, configured to use an entity recognition neural network to determine a hit entity set in the summary text information set; wherein the entity set includes the following information in the summary text information set: information related to representation information and medical state entities; a standard matching subunit, configured to perform medical language standardization matching on the entity set to obtain a normalized entity set; a classification and annotation subunit, configured to perform classification and annotation on the normalized entities in the normalized entity set to obtain a representation information set and a medical state entity set; a medical knowledge graph generation subunit, configured to obtain the medical knowledge graph based on the co-occurrence relationship between the representation information in the representation information set and the medical state entities in the medical state entity set.

[0108] In some optional implementations of this embodiment, the entity recognition neural network in the entity recognition subunit includes: a bidirectional short-term memory network and a conditional random field.

[0109] In some optional implementations of this embodiment, the push information sending unit is further configured to: sort the medical status entities using a probabilistic graph model, select a preset number of the medical status entities according to the sorting results to generate a push information set; and send the push information set to the user.

[0110] In some optional implementations of the present embodiment, the push information generation device further includes: a standard information generation unit, including: an initial information acquisition subunit, configured to acquire the user's input information; an information identification subunit, configured to identify the representation information contained in the input information and obtain an identification result; and a standard representation information determination subunit, configured to determine the standard representation information based on the normalized semantics of the identification result.

[0111] In some optional implementations of this embodiment, the standard representation information determination subunit is further configured to: perform expansion based on the normalized semantics of the recognition result to generate an extended representation information set; and use the extended representation information in the extended representation information set as the standard representation information.

[0112] In some optional implementations of this embodiment, the medical state entity determination unit is further configured to: in response to determining that selection information for the standard characterization information is received, use a pre-constructed medical knowledge graph to determine at least one medical state entity hit by the standard characterization information.

[0113] This embodiment exists as a device embodiment corresponding to the above method embodiment. For the same content, refer to the description of the above method embodiment, which will not be repeated here. Through the push information generation device provided in the embodiment of the present application, the knowledge graph constructed based on the summary information of medical literature is used to determine the push message pushed to the user, thereby reducing the push message determination cost and improving the quality of the push message.

[0114] like Figure 6 , is a block diagram of an electronic device according to a method for generating push information according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0115] like Figure 6 As shown, the electronic device includes: one or more processors 601, a memory 602, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 6 A processor 601 is taken as an example.

[0116] The memory 602 is a non-transient computer-readable storage medium provided in the present application. The memory stores instructions executable by at least one processor to enable the at least one processor to execute the method for generating push information provided in the present application. The non-transient computer-readable storage medium of the present application stores computer instructions, which are used to enable a computer to execute the method for generating push information provided in the present application.

[0117] The memory 602 is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, such as program instructions / modules corresponding to the method for generating push information in the embodiment of the present application (for example, Figure 5 The processor 601 executes various functional applications and data processing of the server by running the non-transient software programs, instructions and modules stored in the memory 602, that is, the method for generating push information in the above method embodiment is implemented.

[0118] The memory 602 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device generating the push information, etc. In addition, the memory 602 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 602 may optionally include a memory remotely arranged relative to the processor 601, and these remote memories may be connected to the electronic device generating the push information via 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 combinations thereof.

[0119] The electronic device for executing the method for generating push information may further include: an input device 603 and an output device 604. The processor 601, the memory 602, the input device 603 and the output device 604 may be connected via a bus or other means. Figure 6 The example of connecting through bus is taken in the following.

[0120] The input device 603 can receive input digital or character information, and generate key signal input related to user settings and function control of the electronic device generating the push information, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator rod, one or more mouse buttons, a trackball, a joystick and other input devices. The output device 604 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display and a plasma display. In some embodiments, the display device may be a touch screen.

[0121] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the above storage system, the above at least one input device, and the above at least one output device.

[0122] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0123] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0124] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0125] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0126] According to the technical solution of the embodiment of the present application, after obtaining standard representation information corresponding to the representation information in the user's input information, at least one medical status entity hit by the standard representation information is determined based on a pre-constructed medical knowledge graph, wherein the medical knowledge graph records the correspondence between the representation information and the medical status entity, and the correspondence is extracted from the summary information of the medical literature. A push information set is generated based on the medical status entity, and the push information set is sent to the user. The knowledge graph constructed based on the summary information of the medical literature is used to determine the push message to be pushed to the user, thereby reducing the push message determination cost while improving the quality of the push message.

[0127] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0128] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for generating push information, comprising: Acquire standard representation information corresponding to the representation information in the user's input information; Based on a pre-constructed medical knowledge graph, at least one medical state entity hit by the standard representation information is determined; wherein the medical knowledge graph records the correspondence between the representation information and the medical state entity, the correspondence is extracted from the abstract information of the medical literature, and the correspondence includes the correspondence between one representation information and multiple medical state entities and / or the correspondence between multiple representation information and one medical state entity; Generate a push information set based on the medical status entity, and send the push information set to the user; The medical knowledge graph is determined based on the following steps: Obtaining summary text information of multiple medical documents to obtain a summary text information set; An entity recognition neural network is used to determine a hit entity set in the summary text information set; wherein the entity set includes the following information in the summary text information set: information related to the representation information and the medical status entity; Performing medical language normalization matching on the entity set to obtain a normalized entity set; Classifying and labeling the standardized entities in the standardized entity set to obtain a representation information set and a medical status entity set; Based on the co-occurrence times of the representation information in the representation information set and the medical state entities in the medical state entity set satisfying a preset threshold, it is determined that there is a corresponding relationship between the representation information and the medical state entities, and the medical knowledge graph is obtained.

2. The method according to claim 1, wherein: The entity recognition neural network includes: a bidirectional short-term memory network and a conditional random field.

3. The method according to claim 1, wherein: The generating a push information set based on the medical status entity and sending the push information set to the user comprises: Sorting the medical status entities using a probabilistic graphical model, and selecting a preset number of the medical status entities according to the sorting result to generate a push information set; Send the push information set to the user.

4. The method according to claim 1, wherein: The step of generating the standard characterization information comprises: Get user input information; Identify the representation information contained in the input information to obtain a recognition result; The standard representation information is determined based on the normalized semantics of the recognition result.

5. The method according to claim 4, wherein determining the standard representation information based on the normalized semantics of the recognition result comprises: Expanding based on the normalized semantics of the recognition result to generate an extended representation information set; The extended representation information in the extended representation information set is used as standard representation information.

6. The method according to claim 5, wherein: The determining, based on the pre-constructed medical knowledge graph, at least one medical state entity hit by the standard representation information comprises: In response to determining that selection information for the standard characterization information is received, a pre-constructed medical knowledge graph is used to determine at least one medical state entity hit by the standard characterization information.

7. A device for generating push information, comprising: A standard representation information acquisition unit, configured to acquire standard representation information corresponding to the representation information in the user's input information; A medical state entity determination unit is configured to determine at least one medical state entity hit by the standard representation information based on a pre-constructed medical knowledge graph; wherein the medical knowledge graph records a correspondence between the representation information and the medical state entity, the correspondence is extracted from the abstract information of the medical literature, and the correspondence includes a correspondence between one representation information and multiple medical state entities and / or a correspondence between multiple representation information and one medical state entity; A push information sending unit, configured to generate a push information set based on the medical status entity, and send the push information set to the user; Also includes: The medical knowledge graph determination unit includes: The initial information acquisition subunit is configured to acquire summary text information of a plurality of medical documents to obtain a summary text information set; The entity recognition subunit is configured to determine a hit entity set in the summary text information set using an entity recognition neural network; wherein the entity set includes the following information in the summary text information set: information related to the representation information and the medical status entity; A standard matching subunit is configured to perform medical language standard matching on the entity set to obtain a standardized entity set; A classification and annotation subunit is configured to classify and annotate the normalized entities in the normalized entity set to obtain a representation information set and a medical status entity set; The medical knowledge graph generation subunit is configured to determine the existence of a corresponding relationship between the representation information and the medical state entities based on the co-occurrence times of the representation information in the representation information set and the medical state entities in the medical state entity set satisfying a preset threshold, so as to obtain the medical knowledge graph.

8. The device according to claim 7, wherein: The entity recognition neural network in the entity recognition subunit includes: a bidirectional short-term memory network and a conditional random field.

9. The device according to claim 7, wherein: The push information sending unit is further configured to: sort the medical status entities using a probabilistic graphical model, and select a preset number of the medical status entities according to the sorting result to generate a push information set; Send the push information set to the user.

10. The apparatus according to claim 7, further comprising: Standard information generation unit, including: An initial information acquisition subunit is configured to acquire user input information; An information recognition subunit is configured to recognize the representation information contained in the input information and obtain a recognition result; The standard representation information determination subunit is configured to determine the standard representation information based on the normalized semantics of the recognition result.

11. The device according to claim 10, wherein: The standard characterization information determination subunit is further configured to: Expanding based on the normalized semantics of the recognition result to generate an extended representation information set; The extended representation information in the extended representation information set is used as standard representation information.

12. The device according to claim 11, wherein The medical status entity determination unit is further configured to: In response to determining that selection information for the standard characterization information is received, a pre-constructed medical knowledge graph is used to determine at least one medical state entity hit by the standard characterization information.

13. An electronic device comprising: at least one processor; as well as A memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor so that the at least one processor can execute the method for generating push information according to any one of claims 1 to 6.

14. A non-transitory computer-readable storage medium storing computer instructions, comprising: The computer instructions are used to enable the computer to execute the method for generating push information according to any one of claims 1 to 6.

Images