Cross-domain entity recognition method and device for entity linking, equipment and medium

Abstract

The application is suitable for the technical field of big data mining, and particularly relates to a cross-domain entity recognition method, device and equipment for entity linking and a medium. The method pre-trains a representation model through embedded representation of source data and target data, constructs supervised training samples and self-supervised training samples through data expansion and enhancement, and respectively uses supervised training and self-supervised training to optimize and train the representation model, so as to provide an accurate expression model for data recognition of a target domain, can recognize entities of data in different domains, and thus improves the adaptive ability of the entity linking task.

Description

Technical Field

[0001] The present application is applicable to the field of big data mining technology, and in particular relates to a cross-domain entity recognition method, device, equipment and medium for entity linking. Background Art

[0002] With the continuous growth of data scale and the widespread application of machine learning, data deduplication (Entity Resolution, ER) has become an indispensable part of data processing, information integration, and data cleaning. Especially when faced with large-scale datasets, the deduplication task becomes particularly complex and computationally expensive. To improve data processing efficiency, it is necessary to rely on entity resolution to cluster and classify data. Traditional entity resolution methods rely on entity linking (EL), which divides the dataset into several smaller blocks to reduce the number of entity tuple pairs that need to be matched later. However, existing entity linking technologies still face the following challenges:

[0003] 1. Insufficient matching recall: When performing entity linking, it is crucial to ensure that the tuple pairs after segmentation cover all possible matches. Therefore, a high recall rate is key to ensuring the completeness of entity resolution. If a large number of matching tuple pairs are missed, subsequent entity matching steps will be affected, and even the overall resolution accuracy will be affected.

[0004] 2. Low computational efficiency: Existing entity linking typically relies on Cartesian product operations or predefined matching rules. This results in extremely high computational overhead when the dataset is large. This is especially true for applications that process datasets of tens of millions or more. Efficiently screening candidate tuple pairs to avoid unnecessary computation has become a pressing issue.

[0005] 3. Poor domain adaptability: Existing entity linking focuses on optimizing datasets in specific fields and lacks cross-domain adaptability. Especially when integrating information from data sources in different fields, the large differences in data distribution across fields make it difficult for general entity linking methods to achieve their desired effects.

[0006] With the rise of neural network-based representation learning, research in entity linking has made significant progress. In this approach, all tuples are converted into dense embeddings, and a K-Nearest Neighbor (KNN) search is performed to retrieve the top K most similar tuples for each tuple, where K is a predefined hyperparameter. Embeddings rely on a representation model, typically implemented using a Transformer encoder, such as the BERT-based model, to convert tuples into vectors. Furthermore, a specific loss function, such as the Contrastive Loss (CL), is employed to bring the embeddings of tuples referring to the same entity closer together and push the embeddings of mismatched tuples further apart. Finally, after the tuples are converted into dense semantic vectors, efficient vector retrieval tools (such as the Faiss retrieval tool) can be used to find the top K most similar results as candidate sets.

[0007] Due to the lack of manual annotation, existing entity linking models (such as DeepBlocker and Sudowoodo) automatically generate training tuple pairs through data augmentation techniques. Positive examples are generated by randomly inserting, deleting, or replacing tags within individual tuples. However, these models cannot accurately produce results when processing cross-domain data and are unable to adapt to specific domains.

[0008] Therefore, how to accurately identify cross-domain entities in cross-domain data in the entity linking task and thus improve the domain adaptation performance of the entity linking task has become an urgent problem to be solved. Summary of the Invention

[0009] In view of this, the embodiments of the present application provide a cross-domain entity recognition method, apparatus, device and medium for entity linking to solve the problem of how to accurately perform cross-domain entity recognition on cross-domain data in entity linking tasks, thereby improving the domain adaptation performance of entity linking tasks.

[0010] In a first aspect, an embodiment of the present application provides a cross-domain entity recognition method for entity linking, the cross-domain entity recognition method comprising:

[0011] Obtain source data in a source domain and target data in a target domain, embed the source data and target data using an initial representation model to obtain a source embedding representation and a target embedding representation, perform adversarial learning training on the representation model with the goal of aligning the source embedding representation and the target embedding representation, and obtain a pre-trained representation model;

[0012] In the source domain, performing attribute extension on the source data to obtain source extended data corresponding to the source data, and associating the source extended data with a label corresponding to the source data to obtain a first supervised training sample;

[0013] In the target domain, performing attribute extension on the target data to obtain target extended data corresponding to the target data, enhancing the target extended data to obtain enhanced target data, constructing pseudo labels on the enhanced target data to obtain corresponding pseudo labels, and associating the enhanced target data with the corresponding pseudo labels to obtain a second supervised training sample;

[0014] Using the first supervised training sample and the second supervised training sample, the pre-trained representation model is trained by supervised contrastive learning, and using the enhanced target data, the pre-trained representation model is trained by self-supervised contrastive learning, to obtain a trained representation model;

[0015] The trained representation model is used to embed the data to be identified in the target domain to obtain an embedded expression to be identified, and entity linking is performed on the data to be identified based on the embedded expression to be identified.

[0016] In a second aspect, an embodiment of the present application provides a cross-domain entity recognition device for entity linking, the cross-domain entity recognition device comprising:

[0017] An alignment training module is used to obtain source data in a source domain and target data in a target domain, embed the source data and target data using an initial representation model to obtain a source embedding representation and a target embedding representation, and perform adversarial learning training on the representation model with the goal of aligning the source embedding representation and the target embedding representation to obtain a pre-trained representation model;

[0018] a source data extension module, configured to perform attribute extension on the source data in the source domain to obtain source extended data corresponding to the source data, and associate the source extended data with a label corresponding to the source data to obtain a first supervised training sample;

[0019] a target data expansion module, configured to perform attribute expansion on the target data in the target domain to obtain target expansion data corresponding to the target data, enhance the target expansion data to obtain enhanced target data, construct pseudo labels on the enhanced target data to obtain corresponding pseudo labels, and associate the enhanced target data with the corresponding pseudo labels to obtain a second supervised training sample;

[0020] a contrastive training module, configured to perform supervised contrastive learning training on the pre-trained representation model using the first supervised training sample and the second supervised training sample, and to perform self-supervised contrastive learning training on the pre-trained representation model using the enhanced target data, to obtain a trained representation model;

[0021] The entity linking module is used to use the trained representation model to embed the data to be identified in the target domain to obtain the embedded expression to be identified, and perform entity linking on the data to be identified based on the embedded expression to be identified.

[0022] In a third aspect, an embodiment of the present application provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the cross-domain entity recognition method as described in the first aspect is implemented.

[0023] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the cross-domain entity recognition method as described in the first aspect is implemented.

[0024] Compared with the prior art, the embodiments of the present application have the following beneficial effects: the present application obtains source data in a source domain and target data in a target domain, uses an initial representation model to embed the source data and the target data to obtain a source embedding representation and a target embedding representation, performs adversarial learning training on the representation model with the goal of aligning the source embedding representation and the target embedding representation to obtain a pre-trained representation model, performs attribute extension on the source data in the source domain to obtain source extended data corresponding to the source data, associates the source extended data with a label corresponding to the source data to obtain a first supervised training sample, performs attribute extension on the target data in the target domain to obtain target extended data corresponding to the target data, and performs attribute extension on the target data in the target domain to obtain target extended data corresponding to the target data. The target extended data is enhanced to obtain enhanced target data, a pseudo-label is constructed for the enhanced target data to obtain a corresponding pseudo-label, the enhanced target data is associated with the corresponding pseudo-label to obtain a second supervised training sample, the first supervised training sample and the second supervised training sample are used to perform supervised contrastive learning training on the pre-trained representation model, and the enhanced target data is used to perform self-supervised contrastive learning training on the pre-trained representation model to obtain a trained representation model, the trained representation model is used to embed the data to be identified in the target field to obtain an embedded expression to be identified, and entity linking is performed on the data to be identified based on the embedded expression to be identified.

[0025] The representation model is obtained by pre-training the embedded representation of the source data and the target data, and then the supervised training samples and the self-supervised training samples are constructed through data expansion and enhancement. The above representation model is optimized and trained using supervised training and self-supervised training respectively, thereby providing an accurate expression model for data recognition in the target field, which can be applied to entity recognition of data in different fields, thereby improving the adaptability of the entity linking task. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0027] Figure 1 This is a schematic diagram of an application environment of a cross-domain entity recognition method for entity linking provided in Example 1 of the present application;

[0028] Figure 2 This is a flow chart of a cross-domain entity recognition method for entity linking provided in Example 2 of the present application;

[0029] Figure 3 This is a flow chart of a cross-domain entity recognition method for entity linking provided in Example 3 of the present application;

[0030] Figure 4 This is a flow chart of a cross-domain entity recognition method for entity linking provided in Example 4 of the present application;

[0031] Figure 5 This is a flowchart of a cross-domain entity recognition method for entity linking provided in Example 5 of the present application;

[0032] Figure 6 This is a structural diagram of a cross-domain entity recognition device for entity linking provided in Example 6 of the present application;

[0033] Figure 7 This is a structural diagram of a computer device provided in Example 7 of the present application. DETAILED DESCRIPTION

[0034] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0035] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0036] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0037] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0038] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0039] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0040] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence refers to the theories, methods, techniques, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results.

[0041] Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0042] It should be understood that the size of the serial numbers of the steps in the following embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0043] In order to illustrate the technical solution of the present application, specific embodiments are provided below.

[0044] The cross-domain entity recognition method for entity linking provided in the first embodiment of the present application can be applied in Figure 1 In an application environment, a server connects to a client and a corresponding database, obtaining data from the database and the client. The client may include, but is not limited to, a PDA, desktop computer, laptop computer, ultra-mobile personal computer (UMPC), netbook computer, cloud computing device, personal digital assistant (PDA), and other computer devices. The server may be implemented as a standalone server or an integrated server. The database may be a relational database such as MySQL or SqlServe, or a non-relational database.

[0045] See also Figure 2 , is a flow chart of a cross-domain entity recognition method for entity linking provided in the second embodiment of the present application. The cross-domain entity recognition method is applied to Figure 1 The server in . Figure 2 As shown, the cross-domain entity recognition method may include the following steps:

[0046] In step S201, source data in the source domain and target data in the target domain are obtained, and the source data and target data are embedded using the initial representation model to obtain source embedded representation and target embedded representation. With the goal of aligning the source embedded representation and the target embedded representation, adversarial learning training is performed on the representation model to obtain a pre-trained representation model.

[0047] In this embodiment, the source domain and the target domain are different domains, where the domain can refer to a data acquisition scenario. For example, in a shopping scenario, the products on platform A and the products on platform B are data from two different domains. In a text classification scenario, language A and language B are data from two different domains.

[0048] For data in the source domain, task models designed to perform tasks like classification and clustering are trained using this data. However, in the target domain, the data may have different distributions than the source data, which may result in the source domain task model being unable to accurately process the target domain data. Therefore, it is necessary to design a representation model that ensures that the input to the task model has the same representation as the source domain data, thereby improving the accuracy of the task model's processing of the target domain data.

[0049] The representation model can realize the embedded representation of data, so as to be used in subsequent task models. The representation model can align the source embedded representation of the source data with the target embedded representation of the target data, so that when the target data to be identified is obtained, the target data can be embedded in the representation to obtain content that can be recognized by the task model.

[0050] In this embodiment, the task model corresponds to the entity linking task. Before inputting data into the entity linking task model, the cross-domain data is embedded using this representation model, and then recognized when input into the task model. An entity is an object represented by data, and the data corresponding to the entity is the data representation and encapsulation of that object.

[0051] Step S202 : In the source domain, attribute extension is performed on the source data to obtain source extended data corresponding to the source data, and the source extended data is associated with a label of the corresponding source data to obtain a first supervised training sample.

[0052] In this embodiment, in cross-domain tasks, the data distributions of the source and target domains often differ, resulting in decreased model performance in the target domain. While aligning feature distributions to alleviate this problem is effective for complex data structures or scenarios lacking rich attributes, the alignment is limited. Therefore, attribute expansion processing is performed on the source and target data to generate data that improves model alignment.

[0053] After the source data is expanded, source extended data is obtained. At this time, for each source extended data, the label of the corresponding source data is the label of the source extended data, thereby forming a first supervised training sample for subsequent supervised training of the representation model.

[0054] Step S203: In the target domain, the target data is attribute-extended to obtain target extended data corresponding to the target data; the target extended data is enhanced to obtain enhanced target data; pseudo-labels are constructed for the enhanced target data to obtain corresponding pseudo-labels; the enhanced target data is associated with the corresponding pseudo-labels to obtain a second supervised training sample.

[0055] In this embodiment, the attribute extension method of the target data corresponds to the attribute extension method of the source data. The target extended data can retain the attribute characteristics of the target data while being closer to the distribution of the source data.

[0056] Since the target data in the target domain has no labels, enhancing the target extended data can improve the effect of using the target data for supervised training. At the same time, constructing pseudo labels for the enhanced target data can obtain higher-quality pseudo labels and improve the effect of using the target data for self-supervised training.

[0057] In one embodiment, a large language model (LLM) is used as a powerful generation and reasoning tool to generate more structured and domain-invariant attribute data, providing new possibilities for cross-domain adaptation.

[0058] Optionally, in the source field, attribute extension is performed on the source data to obtain source extended data corresponding to the source data, including:

[0059] Inputting source data and a basic attribute set corresponding to the source data into a large language model, enabling the large language model to generate a first candidate attribute set according to a first preset instruction;

[0060] Obtaining the attribute frequency of each attribute in the first candidate attribute set, filtering out attributes whose attribute frequencies are greater than a preset frequency or whose attribute frequencies are invalid from the first candidate attribute set, to obtain a first filtered attribute set;

[0061] Based on the normalized mutual information evaluation index, a first extended attribute subset is screened out from the first filter attribute set, and the first extended attribute subset is used to perform attribute extension on the source data to obtain source extended data corresponding to the source data.

[0062] Optionally, in the target field, the target data is attribute-extended to obtain target extended data corresponding to the target data, including:

[0063] Inputting the target data and a basic attribute set corresponding to the target data into the large language model, enabling the large language model to generate a second candidate attribute set according to a second preset instruction;

[0064] Obtaining the attribute frequency of each attribute in the second candidate attribute set, filtering out attributes whose attribute frequencies are greater than a preset frequency or whose attribute frequencies are invalid from the second candidate attribute set, to obtain a second filtered attribute set;

[0065] The second filter attribute set is screened using a maximum average difference optimization algorithm to obtain a second extended attribute subset, and the second extended attribute subset is used to extend the attributes of the target data to obtain target extended data corresponding to the target data.

[0066] First, a large language model is used in the source domain to generate a set of candidate attributes, and the Normalized Mutual Information (NMI) indicator is used to screen out an extended attribute subset that can significantly improve the performance of the source domain task. Second, in the target domain, a large language model is used to generate a set of candidate attributes, and the Maximum Mean Discrepancy (MMD) optimization algorithm is used to select the attribute subset, ensuring that the target domain data retains key features while being closer to the distribution of the source domain data.

[0067] The specific implementation steps include: the first step is to input the source domain data and its basic attribute set into the large language model, generate a candidate attribute set according to the preset instructions, filter out low-frequency or invalid attributes by analyzing the attribute frequency, and select the attribute subset that can significantly improve the performance of the source domain task based on the normalized mutual information, and finally update the attribute set of the source domain data; the second step is to perform the same processing on the target domain data and the basic attribute set, and screen the extended attribute set through the maximum average difference minimization algorithm, so that the target domain data is aligned with the source domain distribution while enhancing the representation ability, forming an optimized target domain attribute set.

[0068] This embodiment uses a large language model to automatically generate rich and discriminative attribute data, reducing manual intervention and improving efficiency. Through dual optimization of normalized mutual information and maximum mean difference, it ensures the efficiency and robustness of extended attributes. It is suitable for different types of cross-domain tasks and datasets and has good versatility and scalability.

[0069] Step S204 , using the first supervised training sample and the second supervised training sample, performs supervised contrastive learning training on the pre-trained representation model, and using the enhanced target data, performs self-supervised contrastive learning training on the pre-trained representation model to obtain a trained representation model.

[0070] In this embodiment, for the representation model used for data processing before the entity linking task, supervised contrastive learning and self-supervised contrastive learning are jointly trained to obtain a better representation model, thereby improving the accuracy of cross-domain data in the entity linking task.

[0071] In step S205 , the trained representation model is used to embed the data to be identified in the target domain to obtain the embedded expression to be identified, and entity linking is performed on the data to be identified based on the embedded expression to be identified.

[0072] The embodiment of the present application obtains a representation model by pre-training the embedded representation of the source data and the target data, and then constructs supervised training samples and self-supervised training samples through data expansion and enhancement. The above-mentioned representation model is optimized and trained using supervised training and self-supervised training respectively, thereby providing an accurate expression model for data recognition in the target field, which can be applied to entity recognition of data in different fields, thereby improving the adaptability of entity linking tasks.

[0073] See also Figure 3 , is a flow chart of a cross-domain entity recognition method for entity linking provided in Example 3 of this application. Figure 3 As shown, in the above step S201, the source data and the target data are embedded using the initial representation model to obtain the source embedding representation and the target embedding representation. With the goal of aligning the source embedding representation and the target embedding representation, the representation model is subjected to adversarial learning training to obtain a pre-trained representation model, including the following steps:

[0074] Step S301: construct a training sample set based on source data and target data.

[0075] Each sample in the training sample set includes a source data tuple in the source data, a target data tuple matching the source data tuple, and a target data tuple not matching the source data tuple, and the target data tuple belongs to the target data.

[0076] Step S302: For any sample, use the initial representation model to embed the source data tuple and any target data tuple of the sample respectively to obtain the corresponding source embedding representation and target embedding representation, and use the initial feature extractor to extract the source embedding representation and target embedding representation to obtain the similarity embedding representation.

[0077] In step S303, the initial classifier is used to predict the matching relationship between the source embedding representation and the target embedding representation to obtain a matching prediction result, and the initial discriminator is used to predict the source of the similarity embedding representation to obtain a domain prediction result.

[0078] In step S304, with the matching prediction result being the true matching relationship between the source data tuple and any target data tuple and the domain prediction result not being the source domain as the goal, the initial representation model, the initial feature extractor and the initial discriminator are subjected to adversarial learning training to obtain a pre-trained representation model.

[0079] Among them, the goal of representation learning is to represent the input entity tuple as a high-dimensional vector embedding for subsequent tasks to be calculated.

[0080] The specific implementation is as follows:

[0081] Input serialization: Each tuple t contains multiple attributes A1, A2, ..., Am, and is serialized according to the following rule: serial(t) = [COL] A1 [VAL]t[A1] ... [COL]Am [VAL] t[Am], where [COL] and [VAL] are special separators used to structurally identify attributes and values;

[0082] Embedding generation: Use the serialized tuple of the representation model M (e.g., SentenceBERT) for embedding to obtain a high-dimensional vector representation emb_t = M(serial(t));

[0083] Contrastive learning optimization: The labeled data DS of the source domain is used to generate a contrastive learning training sample set CL-S. Each sample contains a source data tuple and its positive example set (matching target data tuple) Pa and negative example set (unmatching target data tuple) Na, which are optimized using the cross-entropy loss function.

[0084] In cross-domain scenarios, there are differences in data distribution between the source domain and the target domain. Representation learning that relies solely on the source domain is difficult to adapt to the characteristics of the target domain.

[0085] This application generates cross-domain similarity representations through the following methods:

[0086] (a) Feature extractor design: A multi-layer perceptron (MLP) is used as the feature extractor F, which accepts the embedding representation of two tuples and generates their similarity embedding: emb(a, b) = ReLU(NN(emb(a)⊕emb(b))), where ⊕ is the row-by-row concatenation operation of the embedding and NN is a fully connected layer;

[0087] (b) Classifier: Based on similarity embedding, a binary classifier G is designed to predict the matching relationship of tuple pairs: G(F(emb(a)⊕emb(b)));

[0088] (c) Adversarial learning for cross-domain alignment: In order to align similarity representations in the source and target domains, a binary discriminator D is used to predict whether the input similarity embedding is from the source domain. Then, an optimization objective is constructed to allow the representation model M and feature extractor F to learn cross-domain feature embeddings that can confuse the classification discriminator D. The overall loss function is L = L_CE +λ* L_adv, where L_CE is the cross entropy loss used to optimize the classifier G, and L_adv is the adversarial loss used to optimize M and F to confuse the discriminator D.

[0089] (d) Optimization step: Fix the classification discriminator D, optimize the representation model M and feature extractor F, so that the generated similarity embedding confuses the classification discriminator D as much as possible. Secondly, fix the representation model M and feature extractor F, and optimize the classification discriminator D to improve the discrimination ability. Alternate training between the two steps to finally achieve cross-domain distribution alignment.

[0090] The embodiment of the present application efficiently captures the similarity information between tuples through contrastive learning and serialized embedding generation. Secondly, adversarial learning is introduced in the entity linking task, and unsupervised domain adaptation is achieved through game optimization. The use of lightweight feature extractors and classification discriminators significantly reduces the computational complexity while improving cross-domain adaptability.

[0091] See also Figure 4 , is a flow chart of a cross-domain entity recognition method for entity linking provided in the fourth embodiment of the present application. Figure 4 As shown, in the above step S203, the target extended data is enhanced to obtain enhanced target data, which may include the following steps:

[0092] Step S401 : performing data enhancement on each target data tuple in the target extended data to obtain a first enhancement result corresponding to the target data tuple.

[0093] Step S402 : obtaining a preset enhanced attribute set, and for any target data tuple in the target extended data, randomly selecting N enhanced attribute subsets from the enhanced attribute set as a second enhanced result of the target data tuple.

[0094] Wherein, N is an integer greater than zero.

[0095] Step S403: Combine each target data tuple with the corresponding first enhancement result and second enhancement result to obtain enhanced target data.

[0096] In this embodiment, data augmentation and LLM augmentation are used to improve the training effect of the target domain, as follows:

[0097] Data augmentation: Extract all individual tuples from the left and right tables (R^T_l and R^T_r) of the target data and apply data augmentation techniques to each tuple. Specifically, data augmentation strategies (such as L_aug) are used, including random insertions, deletions, and substitutions, to generate multiple augmented tuples. This data augmentation makes the target domain data more diverse, which helps the model generalize better.

[0098] LLM augmentation: First, we define the set of all augmented attributes in the target domain (B^T_all). Then, we randomly sample multiple subsets from this set and combine these subsets with the attributes of the target tuple to generate new augmented data. This further enhances the robustness of the target domain data. For example, for a tuple t, if C_i and C_j are two different subsets sampled from the augmented attribute set, then t[C_i] and t[C_j] will represent the same entity. Through LLM augmentation, the generated set of positive tuples (P_t^LLM) can provide more diverse training data.

[0099] Optionally, use augmented target data to train the pre-trained representation model with self-supervised contrastive learning, including:

[0100] For any target data tuple in the enhanced target data, a hard negative sample mining algorithm is used to construct a negative sample of the target data tuple to obtain a negative sample set corresponding to the target data tuple;

[0101] The first enhancement result and the second enhancement result of the target data tuple are taken as a positive sample set;

[0102] The target data tuple, the corresponding positive sample set and the negative sample set are used as training triplets, and all training triplets are used to perform self-supervised contrastive learning training on the pre-trained representation model.

[0103] By combining the results of data augmentation and LLM augmentation, a set of positive examples (P_t) for the target tuple is generated. For each augmented tuple, multiple augmented versions are generated, and a set of negative examples (N_t) is generated using a hard negative mining strategy. These positive and negative examples are input as training triplets (t, P_t, N_t) into self-supervised contrastive learning, thereby optimizing the representation model, making different representations of the same entity more similar and representations of different entities more distinct.

[0104] See also Figure 5 , is a flow chart of a cross-domain entity recognition method for entity linking provided in Example 5 of this application. Figure 5As shown, in the above step S204, pseudo labels are constructed for the enhanced target data to obtain corresponding pseudo labels, and the enhanced target data is associated with the corresponding pseudo labels to obtain a second supervised training sample, including the following steps:

[0105] Step S501 : performing a K-nearest neighbor search on each target data tuple in the enhanced target data to obtain the most similar K target data tuples corresponding to each target data tuple.

[0106] Step S502: For any target data tuple, the allocation probability of the K target data tuples is determined based on the similarity between the target data tuple and the K target data tuples. Based on the allocation probability of the K target data tuples, the K target data tuples are randomly sampled to obtain a pseudo-label data set corresponding to the target data tuple.

[0107] Step S503 : Using the pseudo-label dataset as the pseudo-label of the corresponding target data tuple and associating it with the target data tuple, traversing all target data tuples in the enhanced target data to obtain a second supervised training sample.

[0108] In this embodiment, unlike self-supervised contrastive learning, supervised contrastive learning relies on training data generated through pseudo-labels. Pseudo-labels are generated by predicting data in the target domain through the model to guide model training.

[0109] First, for each tuple t in the target domain, a K-nearest neighbor search is performed to find the K most similar tuples. These similar tuples are then used to generate a high-quality pseudo-label training set. The pseudo-label dataset in this embodiment is a generated set of pseudo-positive tuples, rather than directly predicting pseudo-labels for tuple pairs. This reduces noise interference during training.

[0110] Furthermore, to avoid over-reliance on the most similar tuples and mitigate noise interference in the training data, this embodiment calculates the similarity between each pair of tuples (using the inner product to calculate the similarity of the embedding vectors), assigns a probability p to each pair, and then randomly samples based on these probabilities to generate a high-quality pseudo-labeled dataset (C^T). This approach ensures data diversity while reducing the impact of noise and improving model performance.

[0111] In supervised contrastive learning training, a high-quality training dataset C^T generated with pseudo-labels is used together with labeled data D^S from the source domain to perform supervised contrastive learning on the representation model. By optimizing the representation model, the entity representations of the target and source domains are more consistent in the embedding space, thereby improving the performance of cross-domain entity matching.

[0112] In the embodiment of this application, a process of gradually optimizing the representation model M is designed by combining LLM enhancement, adversarial learning strategy and contrastive learning. The basic process is as follows:

[0113] (a) Data preprocessing and initial training: First, labeled and unlabeled data are obtained from the source domain (D^S) and target domain (D^T). The source domain data is converted into triplets (CL^S) and the model M is initially fine-tuned using contrastive learning to adapt it to the source domain data. During this process, the MLP discriminator D and an adversarial learning strategy are used to further optimize the model M so that the similarity representations of the source and target domains are aligned.

[0114] (b) Data augmentation and LLM enhancement: Tuples in the source and target domains are augmented using LLM to generate enhanced attribute sets (B^S_all and B^T_all). Subsequently, LLM is used to fill in the attributes of the tuples in the target domain to generate the final enhanced features.

[0115] (c) Iterative Training and Self-Supervised Learning: After data augmentation and LLM enhancement are complete, the iterative training phase begins. Through self-supervised contrastive learning, the model M is gradually optimized by combining the augmented tuples and pseudo-label generation mechanism. A K-nearest neighbor search is performed on each tuple in the target domain (R^T_l and R^T_r), selecting high-quality pseudo-label tuple pairs for training. The model is further optimized through supervised contrastive learning.

[0116] (d) Iterative Optimization: Model M is continuously trained using labeled data from the source domain and pseudo-labeled data from the target domain over multiple iterations until a predetermined maximum number of iterations is reached. Each iteration generates a new training dataset and updates the model parameters, gradually improving the model's performance in the target domain.

[0117] By comprehensively using technologies such as data augmentation, LLM enhancement, adversarial learning, contrastive learning and pseudo-label generation, the problem of data distribution differences in cross-domain entity linking is effectively solved, enabling the model to better match entities in the target domain, thereby improving the accuracy and robustness of the cross-domain entity linking task.

[0118] Corresponding to the cross-domain entity recognition method for entity linking in the above embodiment, Figure 6 The structure block diagram of the cross-domain entity recognition device for entity linking provided by the sixth embodiment of the present application is shown. The cross-domain entity recognition device is applied to Figure 1 For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0119] See also Figure 6 , the cross-domain entity recognition device includes:

[0120] An alignment training module 61 is configured to obtain source data in a source domain and target data in a target domain, embed the source data and target data using an initial representation model to obtain a source embedding representation and a target embedding representation, and perform adversarial learning training on the representation model with the goal of aligning the source embedding representation and the target embedding representation to obtain a pre-trained representation model.

[0121] A source data extension module 62 is configured to perform attribute extension on the source data in the source domain to obtain source extended data corresponding to the source data, and associate the source extended data with a label corresponding to the source data to obtain a first supervised training sample;

[0122] a target data expansion module 63 configured to perform attribute expansion on the target data in the target domain to obtain target expansion data corresponding to the target data, enhance the target expansion data to obtain enhanced target data, construct pseudo labels on the enhanced target data to obtain corresponding pseudo labels, and associate the enhanced target data with the corresponding pseudo labels to obtain a second supervised training sample;

[0123] a contrastive training module 64 for performing supervised contrastive learning training on the pre-trained representation model using the first supervised training sample and the second supervised training sample, and performing self-supervised contrastive learning training on the pre-trained representation model using the enhanced target data to obtain a trained representation model;

[0124] The entity linking module 65 is used to use the trained representation model to embed the data to be identified in the target domain to obtain the embedded expression to be identified, and perform entity linking on the data to be identified based on the embedded expression to be identified.

[0125] Optionally, the alignment training module 61 includes:

[0126] A training sample construction unit is used to construct a training sample set based on the source data and the target data, wherein each sample in the training sample set includes a source data tuple in the source data, a target data tuple matching the source data tuple, and a target data tuple not matching the source data tuple, and the target data tuple belongs to the target data;

[0127] A similarity representation generation unit is used to embed the source data tuple and any target data tuple of any sample using the initial representation model to obtain the corresponding source embedding representation and target embedding representation, and extract the source embedding representation and target embedding representation using the initial feature extractor to obtain the similarity embedding representation;

[0128] The target prediction unit is used to use the initial classifier to predict the matching relationship between the source embedding representation and the target embedding representation to obtain a matching prediction result, and use the initial discriminator to predict the source of the similarity embedding representation to obtain a domain prediction result;

[0129] The adversarial training unit is used to perform adversarial learning training on the initial representation model, the initial feature extractor and the initial discriminator with the matching prediction result as the true matching relationship between the source data tuple and any target data tuple and the domain prediction result not being the source domain as the goal, to obtain a pre-trained representation model.

[0130] Optionally, the source data expansion module 62 includes:

[0131] A first candidate unit, configured to input source data and a basic attribute set corresponding to the source data into a large language model, so that the large language model generates a first candidate attribute set according to a first preset instruction;

[0132] A first filtering unit is configured to obtain an attribute frequency of each attribute in the first candidate attribute set, and filter out attributes whose attribute frequencies are greater than a preset frequency or whose attribute frequencies are invalid from the first candidate attribute set to obtain a first filtered attribute set;

[0133] The source extension unit is configured to filter out a first extended attribute subset from the first filter attribute set based on a normalized mutual information evaluation index, and perform attribute extension on the source data using the first extended attribute subset to obtain source extended data corresponding to the source data.

[0134] Optionally, the target data expansion module 63 includes:

[0135] A second candidate unit, configured to input the target data and a basic attribute set corresponding to the target data into the large language model, so that the large language model generates a second candidate attribute set according to a second preset instruction;

[0136] A second filtering unit is configured to obtain an attribute frequency of each attribute in the second candidate attribute set, and filter out attributes with an attribute frequency greater than a preset frequency or an invalid attribute frequency from the second candidate attribute set to obtain a second filtered attribute set;

[0137] The target expansion unit is configured to filter the second filter attribute set using a maximum average difference optimization algorithm to obtain a second extended attribute subset, and perform attribute expansion on the target data using the second extended attribute subset to obtain target extended data corresponding to the target data.

[0138] Optionally, the target data expansion module 63 includes:

[0139] A first enhancement unit is configured to perform data enhancement on each target data tuple in the target extended data to obtain a first enhancement result corresponding to the target data tuple;

[0140] The second enhancement unit is configured to obtain a preset enhanced attribute set, and for any target data tuple in the target extended data, randomly select N enhanced attribute subsets from the enhanced attribute set as a second enhancement result for the target data tuple, where N is an integer greater than zero;

[0141] The enhanced data unit is used to combine each target data tuple with the corresponding first enhancement result and second enhancement result to obtain enhanced target data.

[0142] Optionally, the target data expansion module 63 includes:

[0143] A negative sample construction unit is used to construct a negative sample for any target data tuple in the enhanced target data using a hard negative sample mining algorithm to obtain a negative sample set corresponding to the target data tuple;

[0144] a positive sample construction unit, configured to use the first enhancement result and the second enhancement result of the target data tuple as a positive sample set;

[0145] The self-supervised training unit is used to take the target data tuple, the corresponding positive sample set and the negative sample set as training triplets, and use all the training triplets to perform self-supervised comparative learning training on the pre-trained representation model.

[0146] Optionally, the comparison training module 64 includes:

[0147] A similarity search unit is used to perform a K-nearest neighbor search on each target data tuple in the enhanced target data to obtain the most similar K target data tuples corresponding to each target data tuple;

[0148] A pseudo-label selection unit is used to determine the distribution probability of K target data tuples for any target data tuple based on the similarity between the target data tuple and the K target data tuples, and randomly sample the K target data tuples based on the distribution probability of the K target data tuples to obtain a pseudo-label data set corresponding to the target data tuple;

[0149] The supervised training sample construction unit is used to use the pseudo-label data set as the pseudo-label of the corresponding target data tuple and associate it with the target data tuple, traverse all target data tuples in the enhanced target data, and obtain a second supervised training sample.

[0150] It should be noted that the information interaction, execution process and other contents between the above modules are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0151] Figure 7This is a schematic diagram of the structure of a computer device provided in Example 7 of this application. Figure 7 As shown, the computer device of this embodiment includes: at least one processor ( Figure 7 Only one is shown), a memory, and a computer program stored in the memory and executable on at least one processor, which implements the steps of any of the above-mentioned embodiments of the cross-domain entity recognition method for entity linking when the processor executes the computer program.

[0152] The computer device may include, but is not limited to, a processor and a memory. It will be understood by those skilled in the art that Figure 7 The above is merely an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include a network interface, a display screen, and an input device.

[0153] The processor may be a CPU, other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0154] Memory includes readable storage media, internal memory, and the like. Internal memory can be the internal memory of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage medium. The readable storage medium can be the computer device's hard drive. In other embodiments, it can also be an external storage device, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, or a flash memory card. Furthermore, memory can include both the computer device's internal storage unit and external storage devices. Memory is used to store the operating system, application programs, boot loaders, data, and other programs, such as the program code of computer programs. Memory can also be used to temporarily store data that has been output or is about to be output.

[0155] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process steps in the above-described method embodiments by instructing the relevant hardware through a computer program. The computer program may be stored in a computer-readable storage medium. When executed by a processor, the computer program implements the steps of the above-described method embodiments. The computer program includes computer program code, which may be in source code form, object code form, executable file, or some intermediate form. Computer-readable media may include at least: any entity or device capable of carrying computer program code, recording media, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunications signals, and software distribution media. Examples include USB flash drives, removable hard drives, magnetic disks, or optical disks. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunications signals.

[0156] The present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed through a computer program product. When the computer program product runs on a computer device, the computer device can implement the steps in the above-mentioned method embodiment when executing it.

[0157] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0158] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0159] In the embodiments provided in this application, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which can be electrical, mechanical or other forms.

[0160] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0161] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A cross-domain entity recognition method for entity linking, characterized in that: The cross-domain entity recognition method includes: Obtain source data in a source domain and target data in a target domain, embed the source data and target data using an initial representation model to obtain a source embedding representation and a target embedding representation, perform adversarial learning training on the representation model with the goal of aligning the source embedding representation and the target embedding representation, and obtain a pre-trained representation model, wherein, in a text classification scenario, the source data is first language data and the target data is second language data; In the source domain, performing attribute extension on the source data to obtain source extended data corresponding to the source data, and associating the source extended data with a label corresponding to the source data to obtain a first supervised training sample; In the target domain, performing attribute extension on the target data to obtain target extended data corresponding to the target data, enhancing the target extended data to obtain enhanced target data, constructing pseudo labels on the enhanced target data to obtain corresponding pseudo labels, and associating the enhanced target data with the corresponding pseudo labels to obtain a second supervised training sample; Using the first supervised training sample and the second supervised training sample, the pre-trained representation model is trained by supervised contrastive learning, and using the enhanced target data, the pre-trained representation model is trained by self-supervised contrastive learning, to obtain a trained representation model; Using the trained representation model, embedding the data to be identified in the target domain to obtain an embedded expression to be identified, and performing entity linking on the data to be identified based on the embedded expression to be identified; In the source domain, the source data is attribute-extended to obtain source extended data corresponding to the source data, including: inputting the source data and a basic attribute set corresponding to the source data into a large language model, enabling the large language model to generate a first candidate attribute set according to a first preset instruction; Obtaining the attribute frequency of each attribute in the first candidate attribute set, and filtering out attributes whose attribute frequencies are greater than a preset frequency or whose attribute frequencies are invalid from the first candidate attribute set to obtain a first filtered attribute set; Based on a normalized mutual information evaluation index, a first extended attribute subset is screened out from the first filter attribute set, and the first extended attribute subset is used to perform attribute extension on the source data to obtain source extended data corresponding to the source data.

2. The cross-domain entity recognition method according to claim 1, characterized in that The initial representation model is used to embed the source data and the target data to obtain a source embedding representation and a target embedding representation, and the representation model is trained through adversarial learning with the goal of aligning the source embedding representation and the target embedding representation to obtain a pre-trained representation model, including: constructing a training sample set based on the source data and the target data, wherein each sample in the training sample set includes a source data tuple in the source data, a target data tuple matching the source data tuple, and a target data tuple not matching the source data tuple, wherein the target data tuple belongs to the target data; For any sample, use the initial representation model to embed the source data tuple and any target data tuple of the sample respectively to obtain corresponding source embedding representation and target embedding representation, and use the initial feature extractor to extract the source embedding representation and the target embedding representation to obtain a similarity embedding representation; Using the initial classifier, predict the matching relationship between the source embedding representation and the target embedding representation to obtain a matching prediction result; using the initial discriminator, predict the source of the similarity embedding representation to obtain a domain prediction result; With the matching prediction result being the true matching relationship between the source data tuple and any target data tuple and the domain prediction result being not the source domain as the goal, the initial representation model, the initial feature extractor and the initial discriminator are subjected to adversarial learning training to obtain a pre-trained representation model.

3. The cross-domain entity recognition method according to claim 1, characterized in that In the target domain, the target data is attribute-extended to obtain target extended data corresponding to the target data, including: inputting the target data and a basic attribute set corresponding to the target data into the large language model, enabling the large language model to generate a second candidate attribute set according to a second preset instruction; Obtaining the attribute frequency of each attribute in the second candidate attribute set, and filtering out attributes whose attribute frequencies are greater than a preset frequency or whose attribute frequencies are invalid from the second candidate attribute set to obtain a second filtered attribute set; The second filter attribute set is screened using a maximum average difference optimization algorithm to obtain a second extended attribute subset, and the second extended attribute subset is used to extend the attributes of the target data to obtain target extended data corresponding to the target data.

4. The cross-domain entity recognition method according to claim 1, characterized in that The step of enhancing the target extended data to obtain enhanced target data includes: Performing data enhancement on each target data tuple in the target extended data to obtain a first enhancement result corresponding to the target data tuple; Obtaining a preset enhanced attribute set, and for any target data tuple in the target extended data, randomly selecting N enhanced attribute subsets from the enhanced attribute set as a second enhancement result for the target data tuple, where N is an integer greater than zero; Each target data tuple is combined with the corresponding first enhancement result and second enhancement result to obtain enhanced target data.

5. The cross-domain entity recognition method according to claim 4, characterized in that: The performing self-supervised contrastive learning training on the pre-trained representation model using the enhanced target data includes: For any target data tuple in the enhanced target data, a hard negative sample mining algorithm is used to construct a negative sample for the target data tuple to obtain a negative sample set corresponding to the target data tuple; Taking the first enhancement result and the second enhancement result of the target data tuple as a positive sample set; The target data tuple, the corresponding positive sample set and the negative sample set are used as training triplets, and all the training triplets are used to perform self-supervised contrastive learning training on the pre-trained representation model.

6. The cross-domain entity recognition method according to any one of claims 1 to 5, characterized in that: The step of constructing a pseudo label for the enhanced target data to obtain a corresponding pseudo label, and associating the enhanced target data with the corresponding pseudo label to obtain a second supervised training sample includes: Performing a K-nearest neighbor search on each target data tuple in the enhanced target data to obtain the most similar K target data tuples corresponding to each target data tuple; For any target data tuple, determine the allocation probability of the K target data tuples based on the similarity between the target data tuple and the K target data tuples, and randomly sample the K target data tuples based on the allocation probability of the K target data tuples to obtain a pseudo-label dataset corresponding to the target data tuple; The pseudo-label data set is used as a pseudo-label corresponding to the target data tuple and is associated with the target data tuple, and all target data tuples in the enhanced target data are traversed to obtain a second supervised training sample.

7. A cross-domain entity recognition device for entity linking, characterized in that: The cross-domain entity recognition device includes: An alignment training module is configured to obtain source data in a source domain and target data in a target domain, embed the source data and target data using an initial representation model to obtain source embedded representations and target embedded representations, and perform adversarial learning training on the representation model with the goal of aligning the source embedded representations with the target embedded representations to obtain a pre-trained representation model. In a text classification scenario, the source data is data in a first language and the target data is data in a second language. a source data extension module, configured to perform attribute extension on the source data in the source domain to obtain source extended data corresponding to the source data, and associate the source extended data with a label corresponding to the source data to obtain a first supervised training sample; a target data expansion module, configured to perform attribute expansion on the target data in the target domain to obtain target expansion data corresponding to the target data, enhance the target expansion data to obtain enhanced target data, construct pseudo labels on the enhanced target data to obtain corresponding pseudo labels, and associate the enhanced target data with the corresponding pseudo labels to obtain a second supervised training sample; a contrastive training module, configured to perform supervised contrastive learning training on the pre-trained representation model using the first supervised training sample and the second supervised training sample, and to perform self-supervised contrastive learning training on the pre-trained representation model using the enhanced target data, to obtain a trained representation model; An entity linking module is configured to embed the data to be identified in the target domain using the trained representation model to obtain an embedded expression to be identified, and perform entity linking on the data to be identified based on the embedded expression to be identified; The source data expansion module includes: a first candidate unit, configured to input the source data and a basic attribute set corresponding to the source data into a large language model, so as to enable the large language model to generate a first candidate attribute set according to a first preset instruction; a first filtering unit, configured to obtain an attribute frequency of each attribute in the first candidate attribute set, and filter out attributes having an attribute frequency greater than a preset frequency or an invalid attribute frequency from the first candidate attribute set to obtain a first filtered attribute set; The source extension unit is configured to filter out a first extended attribute subset from the first filter attribute set based on a normalized mutual information evaluation index, and perform attribute extension on the source data using the first extended attribute subset to obtain source extended data corresponding to the source data.

8. A computer device, characterized in that: The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the cross-domain entity recognition method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the cross-domain entity recognition method according to any one of claims 1 to 6 is implemented.

Images