Cooperative relationship prediction method and device based on sequence recommendation and electronic equipment

By using time embedding, author high-order embedding and content embedding in cooperative relationship prediction, combined with multiple tokens to represent authors, the problems of low accuracy and difficulty in capturing dynamic information in the existing technology are solved, and efficient and accurate Cooperation relationship forecast.

CN118966440AActive Publication Date: 2024-11-15INST OF MEDICAL INFORMATION CHINESE ACAD OF MEDICAL SCI

Patent Information

Application Number
CN202411041665.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-15
Estimated Expiration
2044-07-31

AI Technical Summary

Technical Problem

The existing cooperative relationship prediction methods are affected by subjective factors, have low accuracy, and are difficult to capture dynamic information and high-order cooperative relationships, especially when applied on large data sets, they are unrealistic.

Method used

Using a sequence recommendation-based method, we describe the dynamic changes of text through time embedding, combining the author's high-order embedding and content embedding, predict collaborators, and represent the author through multiple tokens to save computing resources, which is suitable for large-scale scenarios.

Benefits of technology

It improves the accuracy and comprehensiveness of cooperative relationship prediction, and can predict collaborators efficiently and accurately under large-scale data, with high accuracy and reliability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118966440A_ABST
    Figure CN118966440A_ABST
Patent Text Reader

Abstract

The invention provides a cooperation relation prediction method and device based on sequence recommendation and electronic equipment, and relates to the field of data prediction. When the cooperation relationship is predicted, the dynamic change condition of a text corresponding to an author along with time is described through time embedding, the historical cooperation relationship between the author and a partner is considered through author high-order embedding, the text core description condition is represented through content embedding, and then the cooperation relationship is predicted based on time embedding, content embedding, author embedding set and Token embedding. When the co-author of the target author is predicted, the cooperation relationship can be predicted from multiple angles, and the prediction accuracy is improved. Besides, a plurality of Tokens are configured for the authors to obtain the embedding of the plurality of Tokens corresponding to the authors, so that computing resources can be saved during model training, cooperation relationship prediction in a scene with a large number of authors can be realized, and accuracy, reliability and efficiency are high. And a k-best-crf algorithm is used to carry out multi-Token classification prediction, so that the prediction success rate is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data prediction, and more specifically, to a method, device and electronic device for predicting a cooperative relationship based on sequence recommendation. Background Art

[0002] As scientific research is widely carried out in the academic community, scholars have created a steady stream of scientific research results. As scientific research deepens, the cost of scientific research output becomes increasingly high, and a single scientific research collaborator must work together with other researchers to discover new results. If the cooperation of scholars can be predicted, it will be possible to reduce the time of finding collaborators, reduce the cost of scientific research, and improve the efficiency of scientific research when outputting scientific research results in the future.

[0003] Currently, when predicting cooperative relationships, suitable partners are selected manually based on experience. This method is easily affected by subjective factors, resulting in low accuracy of cooperative relationship prediction. Summary of the invention

[0004] In view of this, the present invention provides a method, device and electronic device for predicting a cooperative relationship based on sequence recommendation to solve the problem of low accuracy in predicting a cooperative relationship.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A cooperative relationship prediction method based on sequence recommendation, comprising:

[0007] Obtain target authors for cooperation relationship prediction;

[0008] Calculating the time embedding and content embedding of the text published by the target author;

[0009] Determine an author embedding set of the target author, wherein the author embedding set includes the scientific research embedding and high-order embedding of the target author; the high-order embedding is the scientific research embedding of the co-author who has a direct or indirect cooperative relationship with the target author;

[0010] Obtain multiple tokens configured for the target author, and query the token embedding corresponding to the token;

[0011] Based on the time embedding, the content embedding, the author embedding set and the token embedding, the collaborators of the target author are predicted; the collaborators are composed of multiple tokens.

[0012] Optionally, calculating the time embedding of the text published by the target author comprises:

[0013] Determine the time when the said text will be made public;

[0014] Performing a standardization operation on the disclosed time to obtain an intermediate time;

[0015] A time embedding value corresponding to the intermediate time is determined to obtain the time embedding of the text.

[0016] Optionally, calculating the content embedding of the text published by the target author includes:

[0017] Obtaining key information of the text;

[0018] An embedding vector of the key information is calculated as the content embedding of the text.

[0019] Optionally, determining an author embedding set of the target author includes:

[0020] Acquire co-authors who cooperate with the target author; the cooperation includes direct cooperation or indirect cooperation;

[0021] Calculating the research embedding of the target author based on the content embedding;

[0022] Calculate the research embedding of the co-authors;

[0023] Calculate the high-order embedding of the target author using the research embeddings of the co-authors and the author collaboration graph structure;

[0024] The scientific research embedding and high-order embedding of the target author are combined to obtain the author embedding set.

[0025] Optionally, calculating the scientific research embedding of the target author according to the content embedding includes:

[0026] The time-decayed weighted average of the content embedding is calculated to obtain the scientific research embedding of the target author.

[0027] Optionally, obtaining multiple tokens configured for the target author and querying the token embedding corresponding to the tokens includes:

[0028] Get the preset number of split tokens;

[0029] Performing a dimensionality reduction operation on a sparse matrix representing the relationship between the author and the text to obtain a token dictionary constructed based on quantization accuracy and a preset number of split tokens, and configuring multiple tokens for the target author based on the token dictionary;

[0030] Input each Token into the embedding layer to obtain the Token embedding corresponding to the Token.

[0031] Optionally, predicting the collaborators of the target author based on the time embedding, the content embedding, the author embedding set and the token embedding includes:

[0032] Combine the time embedding, the content embedding, the author embedding set and the token embedding to obtain input data;

[0033] Inputting the input data sequentially into a cooperative relationship prediction model based on sequence recommendation to obtain a cooperative relationship score of the target author;

[0034] A classifier is used to classify the cooperation relationship scores to obtain the collaborators.

[0035] Optionally, the generation process of the cooperative relationship prediction model based on sequence recommendation includes:

[0036] Obtain training samples, where the training samples include time embeddings, content embeddings, author embedding sets, and combinations of token embeddings corresponding to different authors;

[0037] Based on a preset training method, the training samples are used to perform a training operation on the cooperative relationship prediction model based on sequence recommendation until a training stop condition is met.

[0038] A cooperative relationship prediction device based on sequence recommendation, comprising:

[0039] The author acquisition module is used to obtain the target author for cooperation relationship prediction;

[0040] An embedding calculation module, used to calculate the time embedding and content embedding of the text published by the target author;

[0041] A set determination module, used to determine the author embedding set of the target author, wherein the author embedding set includes the scientific research embedding and high-order embedding of the target author; the high-order embedding is the scientific research embedding of the co-author who has a direct or indirect cooperative relationship with the target author;

[0042] An embedding acquisition module, used to acquire multiple tokens configured for the target author, and query the token embedding corresponding to the token;

[0043] The author prediction module is used to predict the collaborators of the target author based on the time embedding, the content embedding, the author embedding set and the token embedding; the collaborators are composed of multiple tokens.

[0044] An electronic device comprising: a memory and a processor;

[0045] Wherein, the memory is used to store programs;

[0046] The processor calls the program and is used to execute the above-mentioned cooperative relationship prediction method based on sequence recommendation.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention provides a method, device and electronic device for predicting cooperative relationships based on sequence recommendation. In the present invention, when predicting cooperative relationships, the dynamic changes of the text corresponding to the author over time are described by time embedding, the historical cooperative relationship between the author and the collaborator is considered by the author's high-order embedding, and the core description of the text is represented by content embedding. Then, based on the time embedding, the content embedding, the author embedding set and the token embedding, when predicting the collaborator of the target author, the cooperative relationship can be predicted from multiple angles to improve the accuracy of the prediction. In addition, the present invention does not directly use one author as a sequence token, but configures multiple tokens for the author to obtain multiple token embeddings corresponding to the author, which can save computing resources during model training, and can also realize cooperative relationship prediction in scenarios with a large number of authors, with high accuracy and reliability, and high efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0050] Figure 1 A method flow chart of a cooperative relationship prediction method based on sequence recommendation provided by an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of a document co-authoring network provided by an embodiment of the present invention;

[0052] Figure 3 A flowchart of a method for determining an author embedding set provided by an embodiment of the present invention;

[0053] Figure 4 A schematic diagram of input data provided by an embodiment of the present invention;

[0054] Figure 5 A flowchart of a method for determining a Token provided in an embodiment of the present invention;

[0055] Figure 6 A network structure diagram of a VAE provided in an embodiment of the present invention;

[0056] Figure 7 A schematic diagram of a model structure provided by an embodiment of the present invention;

[0057] Figure 8 A schematic diagram of a model training provided by an embodiment of the present invention;

[0058] Fig. 9 An operation logic diagram of a k-best-crf algorithm provided by an embodiment of the present invention;

[0059] Fig.10 A schematic diagram of the structure of a cooperative relationship prediction device based on sequence recommendation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Time embedding refers to: the vector corresponding to the publication time of the text published by the author. Specifically, the publication time of the text published by the author can be encoded into "time embedding" using a time encoding algorithm;

[0062] Content embedding refers to: the vector corresponding to the key information in the text published by the author;

[0063] Research embedding refers to: the weighted average of content embedding;

[0064] High-order embeddedness refers to the scientific research embeddedness of co-authors who have direct or indirect cooperative relationships with the target author.

[0065] Token embedding means: each author is represented by multiple tokens in the token dictionary, and the authors are appropriately densified by sharing tokens among different authors, encoded as multi-hot encoding, and the encoding is input into the embedding layer. The embedding layer is a trainable matrix, and the vector of the corresponding row in the matrix can be found by embedding the id (identity document). Using the id of the token to retrieve the corresponding vector from the embedding layer is token embedding.

[0066] Sequential recommendation means: arranging the above-mentioned time embedding, content embedding, author embedding set (including scientific research embedding and high-order embedding) and token embedding in ascending order according to the author's publication time to form a sequence, and inputting the embedding vectors sorted by time series into the cooperative relationship prediction model based on sequence recommendation to predict the author's collaborators.

[0067] Currently, when predicting cooperative relationships, suitable partners are selected manually based on experience. This method is easily affected by subjective factors, resulting in low accuracy of cooperative relationship prediction.

[0068] In order to avoid the problem of low accuracy caused by manual selection, intelligent recommendation methods can be used, such as the link prediction method of GCN (graph convolutional networks) for cooperation relationship prediction. However, when using this method for prediction, it is found that the model analyzes static graphs and cannot capture the dynamic information of collaborators, which makes the prediction results less accurate. In addition, when predicting, it is not possible to embed the time of cooperation, which loses the characteristics of the author's publication time. In addition, the scale of the model is limited, and it is difficult to capture more complex features, which further makes the prediction results less accurate.

[0069] If CoBERT (Collaboration Bidirectional Encoder Representationsfrom Transformers) is used to predict collaborative relationships, the model does not take into account the time when the text was posted, resulting in a low accuracy of the prediction results. In addition, the model only considers direct collaborators and does not consider higher-order author relationships other than direct collaboration, making the prediction results less comprehensive. In addition, for larger data sets, the VRAM (Video RAM) and computing requirements are too high, making it unrealistic to apply it to large data sets.

[0070] In order to solve the above problems, in the present invention, when predicting the cooperative relationship, the text time factor is considered through time embedding, thereby considering the dynamic information of the collaborator and improving the accuracy of the prediction result. In addition, in addition to considering direct collaborators, the present invention also considers high-level indirect collaborators to improve the comprehensiveness of the prediction results.

[0071] In addition, the present invention utilizes the powerful understanding and generation capabilities of large language models. Specifically, the transformer-based language model can understand and parse complex language structures and contexts. By treating the partner sequence as a language, the capabilities of large language models are fully utilized. At the same time, compared with graph neural network models, transformer-based language models are easy to scale up. When there is sufficient computing power and a large data set, large-scale transformer language models can capture more complex features, further improving the accuracy of prediction results.

[0072] On the basis of the above content, the present invention provides a method, device and electronic device for predicting cooperative relationships based on sequence recommendation. In the present invention, when predicting cooperative relationships, the dynamic changes of the text corresponding to the author over time are described by time embedding, the historical cooperative relationship between the author and the collaborator is considered by the author's high-order embedding, and the core description of the text is represented by content embedding. Then, based on the time embedding, the content embedding, the author embedding set and the token embedding, when predicting the collaborator of the target author, the cooperative relationship can be predicted from multiple angles to improve the accuracy of the prediction. In addition, the present invention does not directly use one author as a sequence token, but configures multiple tokens for the author to obtain multiple token embeddings corresponding to the author, which can save computing resources during model training, and can also realize cooperative relationship prediction in scenarios with a large number of authors, with high accuracy and reliability, and high efficiency.

[0073] Reference Figure 1 , the embodiment of the present invention provides a method for predicting cooperative relationships based on sequence recommendation, referring to Figure 1 , which may include:

[0074] S11. Obtain the target author for which cooperation relationship prediction is to be performed.

[0075] In this embodiment, the target author is the author for whom the cooperative relationship prediction based on sequence recommendation is required, and the target author may be one or more than one. If the target author is more than one, the prediction is performed independently for each target author.

[0076] After determining the target author, it is further necessary to obtain the texts published by the target author. The published texts can be published papers (scientific or non-scientific), applied patents, applied new drug content, published technical achievements or published comments, etc.

[0077] For a text, you can obtain its specific content, or only obtain key content, such as title, abstract, keywords and other information.

[0078] It should be noted that before predicting cooperative relationships based on sequence recommendation, it is necessary to solve the problem of multiple authors with the same name and authors with the same name from different institutions, that is, to normalize the authors. For example, if author A (such as Zhang San) publishes texts using the pen names Zhang Yier, Zhang Zhang, and Zhang San, then the authors of the texts published by Zhang San using all the pen names need to be unified as Zhang San. In addition, it is also necessary to distinguish whether Zhang Sans with the same name from multiple institutions are the same author Zhang San. This is to avoid missing some texts when counting the texts corresponding to the author, and to avoid counting the outputs of different authors as one author.

[0079] S12: Calculate the time embedding and content embedding of the text published by the target author.

[0080] In this embodiment, time embedding refers to the embedding vector corresponding to the time when the text is published, and content embedding refers to the embedding vector corresponding to the text content.

[0081] In practical applications, time embedding is determined based on the time when the text was published, and content embedding is determined based on the key content of the text.

[0082] Specifically, when using transformer to predict cooperative relationships based on sequence recommendation, the traditional position encoding of transformer is an absolute position encoding. The encoding of each position is only related to its position and the maximum length set by the model, which is given in advance and cannot be trained. In this embodiment, the position encoding is modified to time encoding.

[0083] In the text sequence corresponding to the target author, each text has a publication time, which can be directly queried, such as the publication time is March 2011, July 2023, etc. A text has a publication time. The publication time of text i can be recorded as P is text, such as papers, etc. The time is in years and is retained to months.

[0084] Since an author publishes multiple texts, such as papers, a text may have multiple authors, as shown in Table 1. Table 1 is an example of the corresponding text combination of the text publication time and the author. Multiple authors are located under one text, so first copy the time of each text multiple times so that the text time series is aligned with the author sequence. The first author of the author sequence in Table 1 is the author himself, and the time is set to zero. The padding time at the end of the sequence is set to a large value.

[0085] Table 1

[0086]

[0087] It should be noted that for several authors in the same text, their time codes are the same, and due to the characteristics of the transformer, their order cannot be distinguished. Therefore, multiple authors in the same text are order-independent.

[0088] Based on Table 1, the text disclosure time t k The subscript k in represents the subscript of the author sequence. For example, in the table above, t0 = 0, t4 = March 2011.

[0089] When calculating time embedding, the publication time of the text can be determined first, that is, the above t k , since the original t k It is a date, and the value is large. In order to facilitate subsequent calculations, it is necessary to standardize the public time to obtain the intermediate time. Specifically, its value range is reduced by translation and scaling. For example, if the beginning of 2000 is set as the time origin, and each year is represented by 0.3, then October 2021 can be represented as [(2021-2000)+10 / 12]×0.3=6.55, that is, t k In October 2021, the corresponding middle time is 6.55.

[0090] After determining the middle, the next step is to determine the time embedding value corresponding to the middle time to obtain the time embedding of the text.

[0091] Specifically, the "text disclosure time" is encoded as "time embedding". The dimension of time embedding is equal to the embedding dimension of the model, and the length is equal to the sequence length. Each dimension is a trigonometric function with different periods. In trigonometric functions, odd dimensions are sine functions and even dimensions are cosine functions. Let the embedding dimension be D, and use represents the kth element of the dth dimension of the time embedding, where d is the number of the dimension and k is the number of the sequence element, then:

[0092]

[0093] in,

[0094]

[0095] Then we can use the formula to calculate the kth element of the dth dimension Then by combining each Get the time embedding of each text

[0096] In this embodiment, by calculating time embedding, scientific research cooperation is regarded as a series of time-related sequences, which includes dynamic information that changes over time, such as changes in the author's scientific research interests, changes in cooperative relationships, etc. Compared with static graphs, it can consider the dynamic changes of text over time.

[0097] After calculating the time embedding, we also need to calculate the content embedding.

[0098] Specifically, key information of the text, such as title, abstract, keywords and the like, may be directly extracted from the text content, or directly obtained.

[0099] Then, the embedding vector of the key information is calculated as the content embedding of the text.

[0100] Specifically, the key information is constructed into a prompt sentence with a fixed pattern. The prompt sentence can be such as the title is ******, the summary is ******, and the keywords are *******.

[0101] Then, the text vector generation model is used to generate a fixed-size embedding vector corresponding to the prompt sentence. Specifically, the pre-trained model all-mpnet-base-v2 of SentenceTransformers (a neural network model and tool library for organizing natural language) can be used to calculate the embedding vector. The embedding vector is recorded as content embedding. The embedding of text i can be recorded as When the content embedded is too long, the fixed-size vector of the front part is truncated and the subsequent vectors are directly discarded, so as to ensure the consistency of the length of the content embedded.

[0102] It should be noted that the calculation of time embedding and content embedding in this embodiment has no particular order and can be performed simultaneously or sequentially. There is no requirement for the order of execution.

[0103] S13. Determine the author embedding set of the target author.

[0104] In this embodiment, the author embedding set includes the scientific research embedding and high-order embedding of the target author.

[0105] That is, in the present invention, when determining the author embedding set, not only the research embedding of the author itself is considered, but also the research embedding of the collaborator. The collaborator in this embodiment refers to the direct or indirect collaborator of the target author. That is, all authors who have a direct or indirect cooperative relationship with the target author are co-authors.

[0106] For example, refer to Figure 2 , Figure 2Schematic diagram of the document co-authorship network. The target author is A, and his direct collaborators are B and C, B's direct collaborators are D and E, and D's direct collaborator is F. In this embodiment, for the target author A, when performing collaboration prediction, the author embedding set used includes the scientific research embedding of the target author A, and also includes the scientific research embedding of B, C, D, and E, that is, only the direct collaborators and the direct collaborators of the direct collaborators are included.

[0107] In another implementation of the present invention, referring to Figure 3 , step S13 may include:

[0108] S21. Acquire co-authors who cooperate with the target author.

[0109] by Figure 2 For example, collaboration refers to direct or indirect collaboration. The direct collaborators of target author A are B and C, and the direct collaborators of the direct collaborator are D and E.

[0110] S22. Calculate the scientific research embedding of the target author based on the content embedding.

[0111] In this embodiment, the author embedding set can be pre-configured in a calculation method. In this embodiment, the text written by the author is directly related to the author. Therefore, the scientific research embedding can be determined based on the content embedding of the text.

[0112] Specifically, the weighted average value of the content embedding attenuated over time is calculated to obtain the scientific research embedding of the target author.

[0113] In this embodiment, scientific research embedding refers to the weighted average of content embedding, and the scientific research embedding of an author can be calculated based on all the author's texts. The scientific research embedding of author j is denoted as R stands for research.

[0114] In practical applications, the weighted average of the content embeddings of all the author’s texts is calculated to obtain the scientific research embedding.

[0115] In addition to the method for determining scientific research embedding in this embodiment, all the author's content embeddings can also be input into the model to obtain scientific research embedding. This method takes into account the changes in the text, and not only takes into account the author's past research direction but also reflects the author's latest research direction. It also takes into account the changes in the author's influence over time, so that scientific research embedding can be determined more quickly and accurately.

[0116] S23. Calculate the scientific research embedding of the co-authors.

[0117] The calculation process of the research embedding of the co-author is similar to that of the target author. Please refer to the corresponding instructions above.

[0118] S24. Calculate the high-order embedding of the target author using the scientific research embedding of the co-authors and the author collaboration graph structure.

[0119] Among them, the author cooperation relationship graph structure can be as follows Figure 2 As shown in Figure 2, high-order embeddedness refers to the scientific research embeddedness of co-authors who have direct or indirect cooperation relationships with the target author.

[0120] In this embodiment, for the convenience of calculation, when selecting collaborators, only direct collaborators and direct collaborators of direct collaborators are selected, that is, the target author is taken as the core and two layers of cooperative relationships are considered. In addition, three layers or higher layers of cooperative relationships can also be selected according to needs.

[0121] In this embodiment, a two-layer cooperative relationship is taken as an example. Figure 2 For the target author A, consider the direct cooperation relationship between A and B and C, and also consider the indirect cooperation relationship with D and E. This indirect cooperation relationship is a high-order cooperation.

[0122] The calculation process of the scientific research embedding of B, C, D and E is the same as that of the scientific research embedding of A. The scientific research embedding of B, C, D and E is the high-order embedding in this embodiment.

[0123] In this embodiment, not only the first-order neighbors of the author are considered, but also the high-order neighbors of the author are included in the sequence as features, which is equivalent to having a wider field of view from the perspective of the graph.

[0124] S25. Combining the scientific research embedding and high-order embedding of the target author to obtain an author embedding set.

[0125] After obtaining the scientific research embedding and high-order embedding of the target author, they are directly combined to obtain the author embedding set.

[0126] Taking the target author A as an example, the author embedding set of A is initialized as:

[0127] Scientific research embedding of A, B, C, D and E.

[0128] In this embodiment, when calculating the author embedding set, higher-order collaborators are considered, so that when performing subsequent cooperation relationship prediction based on sequence recommendation, not only authors with whom the author has cooperated in the past can be predicted, but also authors with whom the author has not cooperated in the past can be predicted.

[0129] S14. Obtain multiple tokens configured for the target author, and query the token embedding corresponding to the token.

[0130] Specifically, the above embodiment is a relatively simple method of one author corresponding to one Token, so there is no need for a Tokenizer (Tokenizer is a method of generating Tokens from an original sequence, also called a word segmenter). During model training, when the number of authors is not too large, that is, tens of thousands to hundreds of thousands, the method of one author corresponding to one Token is effective. If the number of authors is much larger than this number, for example, there are millions of authors, corresponding to millions of Tokens, then the token content in the formed Token dictionary will be more, and it will also lead to excessive GPU (Graphics Processing Unit) and video memory occupancy. The video memory of a single GPU is difficult to carry more Tokens, making the present invention unable to be applied to large data sets.

[0131] Therefore, in order to avoid the problem that when there are many authors, the model training cannot be realized by using too many tokens, this embodiment uses the Tokenizer for processing natural language models to split each author into multiple "author fragments", and let each author be represented by multiple tokens in the token dictionary. By letting different authors share tokens, the author's one-hot encoding is appropriately densified into multi-hot encoding. Since the author's characteristics are fully considered during the splitting, the tokens of similar authors are also similar, which will better capture the correlation between authors and improve the performance of the model.

[0132] When each author is represented by multiple tokens, Figure 4 For example, each author is represented by 3 tokens. Figure 4 Each element in the last three rows of the table in is split into three, and the sequence length becomes three times the original. The superscripts in rows e and h indicate whether it is the 0th, 1st, or 2nd of the three tokens representing an author.

[0133] Depend on Figure 4 As can be seen from the table, for a target author, there are content embedding, time embedding, author embedding set (including scientific research embedding and high-level embedding) and token embedding. Figure 4 Take the table in for example, which gives examples of five embeddings.

[0134] Combining the text content embedding, time embedding, author embedding set (including scientific research embedding and high-order embedding) and token embedding can get the collaborator sequence. In this sequence, the first position is the author himself, which is equivalent to a starting symbol. Then sort all the co-author lists in chronological order of the text.

[0135] It should be noted that is the target author’s token embedding, and the co-author’s token encoding is as follows Figure 4 As shown in the table.

[0136] The last line h is the data to be input into the model, that is, the obtained partner sequence, which is e A Line, e T Line, e R Line, e H The sum of the contents of row and row e:

[0137] h=e A +e T +e R +e H +e

[0138] in, And so on.

[0139] In order to embed the author's token, refer to Figure 5 , step S14 may include:

[0140] S31. Obtain the preset number of split tokens.

[0141] In this embodiment, the preset number of split Tokens is the number of Tokens into which an author is split, which is represented by t. Generally, it is set based on experience, such as 2, 3 or 4.

[0142] S32. Perform a dimensionality reduction operation on the sparse matrix representing the relationship between the author and the text to obtain a Token dictionary constructed based on quantization accuracy and a preset number of split Tokens, and configure multiple Tokens for the target author based on the Token dictionary.

[0143] In this embodiment, the dimension reduction operation can be performed in two ways, one is the truncated singular value decomposition (TSVD) method, and the other is the β-VAE (β-variational autoencoder) method. They are introduced below.

[0144] 1. Truncated singular value decomposition method

[0145] The calculation formula for truncated singular value decomposition is:

[0146] M≈PΣE T

[0147] Where M is a sparse matrix of text-author relationships, where 0 indicates no relationship and 1 indicates a relationship. P is an m×t text relationship embedding matrix, where m is the number of authors and t is the number of preset split tokens, i.e., the number of tokens to be decomposed for each author; E T is a t×n author relationship embedding matrix, where n is the number of texts; Σ is a t×t diagonal matrix with t largest singular values ​​on the diagonal.

[0148] The truncated singular value decomposition is based on the singular value decomposition (SVD) method. In the singular value decomposition, the square matrix M T M and MM T have the same eigenvalue λ i , whose standardized eigenvectors are p i and e i .Right now:

[0149] M T Mp i =λ i p i

[0150] MM T e i =λ i e i

[0151] All the eigenvectors p i and e i They form matrices P and E respectively T . Singular Values Construct a singular value diagonal matrix Σ. σ i The corresponding eigenvector is called a singular vector, indicating the direction in which the matrix stretching effect is strongest. T , which constitutes the singular value decomposition of M.

[0152] Compared with the singular value decomposition, the truncated singular value decomposition only takes the largest first t singular values ​​to form a new diagonal matrix Σ, and the corresponding eigenvectors to form new P and E T , so the corresponding PΣE T will be an approximation of M, that is, M≈PΣE T .

[0153] In the truncated singular value decomposition, the singular vectors need to be processed into token numbers. Specifically, the values ​​of the E matrix are first normalized by author so that each value is between 0 and 1.

[0154] Then add a mean of 0 and a variance of 10 to the E matrix -2Gaussian noise is added to the embeddings of any two authors, making the embeddings of any two authors different. This is to handle the situation where the Token sequence number is repeated. For example, when a text has two authors and they have no other texts, the two authors are indistinguishable only from the perspective of the text-author relationship, which will cause the embeddings calculated by the truncated singular value decomposition to be the same. Adding Gaussian noise can avoid the situation where the author's Token sequence number tuple is repeated.

[0155] In the truncated singular value decomposition, the E matrix needs to be quantized. Specifically, each vector is quantized. The 0-1 values ​​in the vector are divided into v steps and quantized into values ​​of 0 to v-1, where v is the quantization accuracy.

[0156] Offset the quantization values ​​of different dimensions so that different dimensions use quantization values ​​in different ranges. For example, the quantization value of the 0th dimension is 0~v-1, the quantization value of the 1st dimension is v~2v-1, and so on. We get v·t different Tokens as a dictionary. Each author is composed of t Tokens in the dictionary, so that we can use the Token dictionary to configure multiple Tokens for the target author. Specifically, after constructing the Token dictionary, for each author in the author set, based on their quantization values, select the quantized values ​​from v steps and quantize them to values ​​of 0~v-1 respectively.

[0157] The quantization values ​​of different dimensions are offset so that different dimensions use quantization values ​​of different ranges. For example, the quantization value of the 0th dimension is 0~v-1, the quantization value of the 1st dimension is v~2v-1, and so on, and t (t is the preset number of split tokens) tokens can be obtained.

[0158] In addition to the truncated singular value decomposition method, the principal component analysis method can also be used. The principal component analysis and truncated singular value decomposition can effectively decompose the characteristics of the source data into various dimensions to achieve dimensionality reduction, and they are independent of each other. These two linear methods are simple linear methods, and the processing results are poor when facing complex nonlinear relationships. Therefore, when facing complex nonlinearities, the β-VAE method can be used. The specific method can be selected by the technician according to the actual situation.

[0159] 2. β-VAE

[0160] In this example, β-VAE is used to generate a very low-dimensional embedding for the author, and the number of dimensions is used as the decomposition number. For example, if each author is to be decomposed into 8 tokens, an 8-dimensional author embedding is generated.

[0161] When decomposing, the following goals should be achieved:

[0162] Authors with similar author relationships should have more common tokens. Therefore, embedding should effectively decompose the features of the source data into various dimensions that are independent of each other.

[0163] In order to improve the ability to capture complex relationships, the β-VAE method is introduced in the graph node embedding calculation. The embedding generated by β-VAE has the characteristic of effectively decoupling independent features.

[0164] VAE is a generative model that approximately models the data generation process as a probability distribution process through variational inference. The model is encouraged to make the potential representation of each observed data as close to the standard Gaussian distribution as possible, that is, to make each dimension of the potential representation close to an independent Gaussian distribution.

[0165] β-VAE is a variation of the VAE model. Its parameter β can adjust the weight of the KL (Kullback-Leibler) divergence part, which means that the model pays more attention to the feature decoupling of the latent space.

[0166] When using β-VAE to decompose Token, in order to better decouple features and have lower requirements for reconstruction error, you can choose a value of β>1 to increase the weight of the KL divergence part to better decouple features. Figure 6 It is the network structure diagram of VAE. Where μ is the mean of the Gaussian distribution of the latent variable, and σ is the standard deviation of the Gaussian distribution of the latent variable.

[0167] For a set of observation data X = {x1, x2, ..., x n}, assume that a generative model p(x|z) is used to describe the relationship between X and a latent variable Z, which has a prior distribution p(z), which is assumed to be a standard Gaussian distribution.

[0168] The training goal of VAE is to maximize the marginal log-likelihood logp(x) of the observed data. Since direct calculation is not feasible, variational inference is used to approximate it, that is, the variational posterior distribution q(z|x) is introduced to approximate the true prior distribution p(z|x). q(z|x) will also be modeled as a Gaussian distribution and output by the encoder. Deform logp(x):

[0169]

[0170] In the formula, the first term is the expectation of the conditional log-likelihood of the data, denoted as The second term is to add the expectation of the prior, and the third term is to subtract the entropy of the variational posterior. Overall, it is to subtract the KL divergence D between the prior and the variational posterior. KL (q(z|x)||p(z)). Therefore, the above formula can be written as:

[0171]

[0172] The right side of the inequality is called the Evidence Lower Bound (ELBO). In VAE, logp(x) is approximately maximized by maximizing the evidence lower bound. In β-VAE, a parameter β is added to the KL divergence term of ELBO:

[0173]

[0174] The optimization objective of β-VAE is the above ELBO. A larger β means that the optimization objective places more emphasis on KL divergence.

[0175] The training steps are as follows:

[0176] 1) Set the potential layer dimension to the number of tokens to be decomposed t;

[0177] 2) Take the rows represented by each author in the sparse matrix M of the text-author relationship as the sparse embedding of the author, input it into the β-VAE model, and make the model output restore the input. You can also use graph convolution to obtain the input vector, but this method will complicate the model;

[0178] 3) Conduct model training;

[0179] 3.1) Take the mean output of the VAE latent layer as the data after dimensionality reduction;

[0180] 3.2) Add noise, normalize and quantize, and set a quantization accuracy v;

[0181] Through the above steps, v·t different tokens can be obtained as a dictionary, that is, a Token dictionary. All authors are composed of t tokens in the dictionary, so that multiple tokens can be configured for the target author using the Token dictionary.

[0182] S33. Input each Token into the embedding layer to obtain the Token embedding corresponding to the Token.

[0183] Specifically, the embedding layer is used to process the Token to obtain the Token embedding corresponding to the Token, and each Token corresponds to a Token embedding.

[0184] S15. Predicting collaborators of the target author based on the time embedding, the content embedding, the author embedding set, and the token embedding.

[0185] Wherein, the collaborator is composed of multiple Tokens.

[0186] In practical applications, after the content embedding, time embedding, author embedding set and the multiple Token embeddings (determined based on Token, Token can actually be Token id) are known, the cooperative relationship prediction model based on sequence recommendation can be used to predict the cooperative relationship.

[0187] In this embodiment, a pure decoder of a GPT (Generative Pre-Trained Transformer) model can be used as a cooperative relationship prediction model based on sequence recommendation. The structure of the cooperative relationship prediction model based on sequence recommendation is as follows: Figure 7 shown.

[0188] The generation process of the cooperative relationship prediction model based on sequence recommendation includes:

[0189] 1) Obtain training samples.

[0190] The training samples include time embedding, content embedding, author embedding set and token embedding combination corresponding to different authors. In this embodiment, the author name can be replaced with the id in the dictionary to achieve unification.

[0191] Specifically, we can obtain a large number of authors and the texts corresponding to them, and calculate the time embedding and content embedding for each text. Then, for each author, we use the truncated singular value decomposition method or β-VAE method to split the author and obtain multiple token embeddings. For details, please refer to Figure 4 The table in .

[0192] After knowing the time embedding, content embedding, author embedding set and token embedding, for each author, determine its input data h, that is, the collaborator sequence.

[0193] 2) Based on a preset training method, the training samples are used to perform a training operation on the cooperative relationship prediction model based on sequence recommendation until a training stop condition is met.

[0194] During specific training, it is necessary to divide the data into training sets and test sets. The version without the last element in the input data h is used as the training set, and the version containing the last element is used as the test set.

[0195] When training, let the model predict the next element for each element in the training set. When testing the model, test whether the output of the second-to-last element in the test set is the last element.

[0196] During training, refer to Figure 8, using the "teacher-forcing" training method. Specifically, since each author has multiple tokens, during training, the classification output of each author in the forced sequence is its next author. Specifically, the output of the first token of an author is the first token of the next author, the output of the second token is the second token of the next author, and so on. This is equivalent to an author at any position, without knowing the information behind it, using all the information of itself and the previous sequence to predict the collaborator behind it. Use cross entropy loss to reflect this goal. Figure 8 Shown is the relationship between the training set input data and labels during training. Figure 8 Each author has 4 tokens, represented by four squares filled with the same value; the upward arrow represents the inference of the output data above from the input data below. It can be seen that the labels used for training are formed by shifting the original input sequence by 4 positions.

[0197] The training stop condition during training can be configured according to the actual situation, such as when the cross entropy loss is less than a preset threshold, or when the number of training times reaches a threshold.

[0198] In addition, during training, we added a label smoothing regularization method and adopted the method of dropout (discarding) some data to prevent the model from overfitting and improve the generalization ability. It also prevents the model from being too confident in the output results and reducing the diversity of the recommendation results.

[0199] During training, a classifier is set after the cooperative relationship prediction model based on sequence recommendation, and the cooperative relationship prediction model based on sequence recommendation and the classifier can be trained at the same time.

[0200] Specifically, the classifier in the embodiment of the present invention adopts the k-best-crf algorithm (conditional random field algorithm of k optimal paths). The operation logic of the k-best-crf algorithm is shown in Fig. 9 Among them, k is the path length, which can be t, and γ is the number of classification labels.

[0201] In the k-best-crf algorithm, instead of calculating the optimal token for each j∈k separately, an optimal path is found from the k-dimensional tokens. Through this algorithm, the predicted k tokens can better represent a real author.

[0202] In actual calculation, in order to reduce the search space, the matrix a in the k-best-crf algorithm i,j (i∈γ,j∈k) stores the sum of all paths from j=1 to the current Token. After this planning algorithm, the search space is expanded from γ k Reduce the dimension to γ×k.

[0203] After training is completed, the training effect is tested using the test set. Each sequence in the test set has one more author (t tokens) than the training set. During testing, the classification output of the penultimate author in the test set is counted to see if it matches the last author to determine whether the training meets the requirements.

[0204] If the test set does not meet the requirements, retraining is performed. If it meets the requirements, the model is used to predict the cooperative relationship based on sequence recommendation. The author classification prediction output of the model is to predict the next t tokens in the current sequence to form the next collaborator. After classification calculation, several authors with the highest probability scores are regarded as potential partners.

[0205] After obtaining the model, the time embedding, the content embedding, the author embedding set and the token embedding can be combined to obtain input data, and the input data can be sequentially input into the cooperative relationship prediction model based on sequence recommendation to obtain the cooperative relationship score of the target author. The cooperative relationship score can be classified by using a classifier to obtain cooperative relationship data.

[0206] The output of the k-best-crf algorithm is any of the following:

[0207] 1) The top five authors by probability score;

[0208] 2) The probability scores of the authors are accumulated from high to low and reach the preset value. For example, the probability scores are accumulated in order from high to low. When the accumulated scores reach 80 points, all the accumulated authors are the predicted collaborators;

[0209] 3) Collaborators whose probability scores are greater than the preset score threshold;

[0210] 4) Select several of the historical collaborators and several of the ones with whom the historical collaborators have not collaborated, so as to diversify the recommendation methods.

[0211] 5) Recommend authors who rank high among authors from different ranges, such as dissimilar authors from the same department or the same city.

[0212] For example, the model output can be as follows:

[0213] Cooperate with Wang Wu in May 2023 and with Sun Liu in June 2023.

[0214] In practical applications, the input data of the model is the above h. After time embedding, content embedding, author embedding set and token embedding, for each author, its input data h is determined. h is composed of time embedding, content embedding, author embedding set and token embedding. The output collaborator is similar to h, and is also composed of time embedding, content embedding, author embedding set and token embedding. The predicted collaborator can be obtained through token embedding. That is, the collaborator prediction is realized through the present invention.

[0215] From the above, it can be seen that the present invention utilizes the co-authorship relationship formed by the authors of documents in the field to construct a document co-authorship network, and embeds the semantic information of the document (such as the key information mentioned above), time information and author information into the network nodes, and extracts the multi-head and high-dimensional information of the cooperation network through the Transformer decoder sequence model to predict the collaborators of a certain author, the time of cooperation and the subject of the cooperation text.

[0216] In this embodiment, when predicting cooperative relationships, time embedding is used to describe the dynamic changes of the text corresponding to the author over time, the historical cooperative relationship between the author and the collaborator is considered through the author's high-order embedding, and the core description of the text is represented through content embedding. Then, based on the time embedding, the content embedding, the author embedding set and the Token embedding, when predicting the collaborators of the target author, the cooperative relationship can be predicted from multiple angles, thereby improving the accuracy of the prediction.

[0217] In addition, the present invention does not directly use one author as a sequence token, but configures multiple tokens for the author, uses multiple tokens to represent one author, and obtains multiple token embeddings corresponding to the author. This can save computing resources during model training, and can process a large number of authors with a smaller dictionary size, thereby realizing cooperative relationship prediction in scenarios with a large number of authors, with higher accuracy and reliability and higher efficiency.

[0218] In addition, using the optimized Token decomposition method, compared with the one-hot representation method of a single Token, the multi-hot (multi-hot encoding) representation method of multiple Tokens is denser to a certain extent, and carries richer feature information, thereby making the prediction results more accurate.

[0219] In order to verify the prediction feasibility and prediction effect of the cooperative relationship prediction method based on sequence recommendation of the present invention, another embodiment of the present invention carried out the following verification experiment.

[0220] 1. Experimental Dataset

[0221] Literature in the field of immunology from 1970 to 2024 was retrieved from the Web of Science database, and the literature types were limited to papers, reviews, and related conference documents, including Article, Review, Proceeding Paper, Letter, and Meeting Abstract, with a total of 855,120 documents. The author information and the cooperation relationship between authors were extracted from them. Data preprocessing included deleting anonymous authors and normalizing the authors, that is, filtering out related authors named "[anonymous]". On this basis, the authors were normalized using the author's email address in the bibliographic information.

[0222] 2. Experimental Evaluation Indicators The present invention adopts the HR (Hits Rati) and NDCG (Normalized Discounted Cumulative Gain) indicators used in evaluating recommendation algorithms.

[0223]

[0224] Where N is the length of the test sample, hits(i) is whether the i-th recommendation result is in the test sample, if yes, it is 1, otherwise it is 0. HR emphasizes the accuracy of the prediction results. When N is the evaluation length K, it is concretely expressed as HR@K.

[0225]

[0226] Where N is the length of the test sample, p i is the position of the i-th recommendation result in the test sample. If this value does not exist in the test sample, then p i →∞. NDCG places more emphasis on the order accuracy of the prediction results. When N is the evaluation length K, it is concretely expressed as NDCG@K.

[0227] 3. Experimental benchmarking model

[0228] In order to compare the innovativeness of the present invention, the present invention is compared with the CoBERT model experiment, and the comparison results are shown in Table 4.

[0229] Table 4

[0230]

[0231]

[0232] 4. Experimental Results

[0233] Table 5

[0234]

[0235] The experimental results show that:

[0236] (1) In this experimental data set, the indicators of the present invention far exceed those of the Cobert model;

[0237] (2) In the Dropout comparison experiment, the model works best when Dropout=0.3, which shows that appropriately discarding some data can enhance the robustness of the model;

[0238] (3) In the feedforward width comparison experiment, the best effect is achieved when the width is 1024 in 4 layers. This shows that when the input vector dimension of this model is around 1000, the width of 1024 can avoid data loss (512 width) and can also avoid embedding too much padding.

[0239] (4) In the head number comparison experiment, the results of 8 and 16 heads in the transformer are better, which shows that the appropriate number of heads can capture more data without causing overfitting of the model;

[0240] (5) In the layer comparison experiment, the effect is better at 2 and 4 layers, which shows that the present invention can achieve better results when the number of layers is small, even if the parameters of the model are relatively small.

[0241] Based on the above embodiment of the method for predicting cooperative relationships based on sequence recommendation, another embodiment of the present invention provides a device for predicting cooperative relationships based on sequence recommendation, referring to Fig.10 , which may include:

[0242] The author acquisition module 11 is used to acquire the target author to be predicted for the cooperative relationship;

[0243] An embedding calculation module 12, used to calculate the time embedding and content embedding of the text published by the target author;

[0244] A set determination module 13 is used to determine the author embedding set of the target author, wherein the author embedding set includes the scientific research embedding and high-order embedding of the target author; the high-order embedding is the scientific research embedding of the co-author who has a direct or indirect cooperative relationship with the target author;

[0245] The embedding acquisition module 14 is used to acquire multiple tokens configured for the target author and query the token embedding corresponding to the token;

[0246] The author prediction module 15 is used to predict the collaborators of the target author based on the time embedding, the content embedding, the author embedding set and the token embedding; the collaborators are composed of multiple tokens.

[0247] In an optional implementation, when the embedding calculation module 12 is used to calculate the time embedding of the text published by the target author, it is specifically used to:

[0248] Determine the publication time of the text, perform a standardization operation on the publication time to obtain an intermediate time, determine a time embedding value corresponding to the intermediate time, and obtain the time embedding of the text.

[0249] In an optional implementation, when the embedding calculation module 12 is used to calculate the content embedding of the text published by the target author, it is specifically used to:

[0250] The key information of the text is obtained, and an embedding vector of the key information is calculated as the content embedding of the text.

[0251] In an optional implementation, the set determination module 13 includes:

[0252] An author acquisition submodule is used to acquire co-authors who have collaborated with the target author; the collaboration includes direct collaboration or indirect collaboration;

[0253] A first embedding calculation submodule, configured to calculate the scientific research embedding of the target author according to the content embedding;

[0254] A second embedding calculation submodule, used to calculate the scientific research embedding of the co-author;

[0255] A third embedding calculation submodule is used to calculate the high-order embedding of the target author by using the scientific research embedding of the co-author and the author cooperation relationship graph structure;

[0256] The set determination submodule is used to combine the scientific research embedding and the high-order embedding of the target author to obtain the author embedding set.

[0257] In an optional implementation, the first embedded computing submodule is specifically used for:

[0258] The time-decayed weighted average of the content embedding is calculated to obtain the scientific research embedding of the target author.

[0259] In an optional implementation, the embedding acquisition module 14 includes:

[0260] The data acquisition submodule is used to obtain the preset number of split tokens;

[0261] The Token configuration submodule is used to perform a dimensionality reduction operation on the sparse matrix representing the relationship between the author and the text, obtain a Token dictionary constructed based on quantization accuracy and a preset number of split Tokens, and configure multiple Tokens for the target author based on the Token dictionary;

[0262] The embedding determination submodule is used to input each Token into the embedding layer to obtain the Token embedding corresponding to the Token.

[0263] In an optional implementation, the author prediction module 15 includes:

[0264] A combination submodule, used for combining the time embedding, the content embedding, the author embedding set and the token embedding to obtain input data;

[0265] A score determination submodule, used to sequentially input the input data into a cooperative relationship prediction model based on sequence recommendation to obtain a cooperative relationship score of the target author;

[0266] The classification submodule is used to use a classifier to perform classification operations on the cooperation relationship scores to obtain collaborators.

[0267] In an optional implementation, the method further includes:

[0268] The model training module is used to obtain training samples, wherein the training samples include a combination of time embedding, content embedding, author embedding set and token embedding corresponding to different authors; based on a preset training method, the training samples are used to train the cooperative relationship prediction model based on sequence recommendation until the training stop condition is met.

[0269] In this embodiment, when predicting the cooperative relationship, the dynamic changes of the text corresponding to the author over time are described by time embedding, the historical cooperative relationship between the author and the collaborator is considered by the author's high-order embedding, and the core description of the text is represented by content embedding. Then, based on the time embedding, the content embedding, the author embedding set and the token embedding, when predicting the collaborator of the target author, the cooperative relationship can be predicted from multiple angles to improve the accuracy of the prediction. In addition, the present invention does not directly use one author as a sequence token, but configures multiple tokens for the author to obtain multiple token embeddings corresponding to the author, which can save computing resources during model training, and can also realize cooperative relationship prediction in scenarios with a large number of authors, with high accuracy and reliability, and high efficiency.

[0270] It should be noted that, for the working process of each module and sub-module in this embodiment, please refer to the corresponding description in the above embodiment, which will not be repeated here.

[0271] Another embodiment of the present invention provides an electronic device, including: a memory and a processor;

[0272] Wherein, the memory is used to store programs;

[0273] The processor calls the program and is used to execute the above-mentioned cooperative relationship prediction method based on sequence recommendation.

[0274] In this embodiment, when predicting the cooperative relationship, the dynamic changes of the text corresponding to the author over time are described by time embedding, the historical cooperative relationship between the author and the collaborator is considered by the author's high-order embedding, and the core description of the text is represented by content embedding. Then, based on the time embedding, the content embedding, the author embedding set and the token embedding, when predicting the collaborator of the target author, the cooperative relationship can be predicted from multiple angles to improve the accuracy of the prediction. In addition, the present invention does not directly use one author as a sequence token, but configures multiple tokens for the author to obtain multiple token embeddings corresponding to the author, which can save computing resources during model training, and can also realize cooperative relationship prediction in scenarios with a large number of authors, with high accuracy and reliability, and high efficiency.

[0275] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting cooperative relationships based on sequence recommendation, characterized in that: include: Obtain target authors for cooperation relationship prediction; Calculating the time embedding and content embedding of the text published by the target author; Determine an author embedding set of the target author, wherein the author embedding set includes the scientific research embedding and high-order embedding of the target author; the high-order embedding is the scientific research embedding of the co-author who has a direct or indirect cooperative relationship with the target author; Obtain multiple tokens configured for the target author, and query the token embedding corresponding to the token; Based on the time embedding, the content embedding, the author embedding set and the token embedding, the collaborators of the target author are predicted; the collaborators are composed of multiple tokens.

2. The method for predicting cooperative relationships based on sequence recommendation according to claim 1, characterized in that: Calculate the temporal embedding of the text published by the target author, including: Determine the time when the said text will be made public; Performing a standardization operation on the disclosed time to obtain an intermediate time; A time embedding value corresponding to the intermediate time is determined to obtain the time embedding of the text.

3. The method for predicting cooperative relationships based on sequence recommendation according to claim 1, characterized in that: Calculate the content embedding of the text published by the target author, including: Obtaining key information of the text; An embedding vector of the key information is calculated as the content embedding of the text.

4. The method for predicting cooperative relationships based on sequence recommendation according to claim 1, characterized in that: Determining an author embedding set of the target author, including: Acquire co-authors who have collaborated with the target author; the collaboration includes direct collaboration or indirect collaboration; Calculating the research embedding of the target author based on the content embedding; Calculate the research embedding of the co-authors; Calculate the high-order embedding of the target author using the research embeddings of the co-authors and the author collaboration graph structure; The scientific research embedding and high-order embedding of the target author are combined to obtain the author embedding set.

5. The method for predicting cooperative relationships based on sequence recommendation according to claim 4, characterized in that: Calculating the research embedding of the target author according to the content embedding includes: The time-decayed weighted average of the content embedding is calculated to obtain the scientific research embedding of the target author.

6. The method for predicting cooperative relationships based on sequence recommendation according to claim 1, characterized in that: Get multiple tokens configured for the target author, and query the token embedding corresponding to the token, including: Get the preset number of split tokens; Performing a dimensionality reduction operation on a sparse matrix representing the relationship between the author and the text to obtain a token dictionary constructed based on quantization accuracy and a preset number of split tokens, and configuring multiple tokens for the target author based on the token dictionary; Input each Token into the embedding layer to obtain the Token embedding corresponding to the Token.

7. The method for predicting cooperative relationships based on sequence recommendation according to claim 1, characterized in that: Predicting collaborators of the target author based on the time embedding, the content embedding, the author embedding set, and the token embedding includes: Combine the time embedding, the content embedding, the author embedding set and the token embedding to obtain input data; Inputting the input data sequentially into a cooperative relationship prediction model based on sequence recommendation to obtain a cooperative relationship score of the target author; A classifier is used to classify the cooperation relationship scores to obtain the collaborators.

8. The method for predicting cooperative relationships based on sequence recommendation according to claim 7, characterized in that: The generation process of the cooperative relationship prediction model based on sequence recommendation includes: Obtain training samples, where the training samples include time embeddings, content embeddings, author embedding sets, and combinations of token embeddings corresponding to different authors; Based on a preset training method, the training samples are used to perform a training operation on the cooperative relationship prediction model based on sequence recommendation until a training stop condition is met.

9. A cooperative relationship prediction device based on sequence recommendation, characterized in that: include: The author acquisition module is used to obtain the target author for cooperation relationship prediction; An embedding calculation module, used to calculate the time embedding and content embedding of the text published by the target author; A set determination module, used to determine the author embedding set of the target author, wherein the author embedding set includes the scientific research embedding and high-order embedding of the target author; the high-order embedding is the scientific research embedding of the co-author who has a direct or indirect cooperative relationship with the target author; An embedding acquisition module, used to acquire multiple tokens configured for the target author, and query the token embedding corresponding to the token; The author prediction module is used to predict the collaborators of the target author based on the time embedding, the content embedding, the author embedding set and the token embedding; the collaborators are composed of multiple tokens.

10. An electronic device, characterized in that: include: Memory and processor; Wherein, the memory is used to store programs; The processor calls the program and is used to execute the cooperative relationship prediction method based on sequence recommendation as described in any one of claims 1-8.

Citation Information

Patent Citations

  • A method of scholar recommendation and cooperative prediction based on representation learning and competition theory

    CN109145087A

  • Academic field data correlation prediction method based on deep learning and computer

    CN110209822A

  • Scientific achievement affiliation prediction method and system based on natural language processing and social network

    CN116644833A

  • Scientific research paper cooperation prediction method based on dynamic graph representation learning

    CN117033658A

  • Theme monitoring method and device, electronic equipment and readable storage medium

    CN118070786A

Cited By

  • Topic text generation method and related device

    CN119398037A

  • A method for generating a topic text and a related device

    CN119398037B