A method and system for expressing scientific and technological achievements based on software system description

Abstract

The present invention relates to a method and system for expressing scientific and technological achievements based on software system descriptions. The method comprises: treating specific interfaces in the software system as nodes in a graph structure, extracting the associations between the various interfaces of the software system as edges, and organizing the hierarchical structure of the software system; and simultaneously embedding multi-dimensional feature information into each node in the hierarchical structure of the software system, processing multiple software systems, and constructing a graph structure of the software system based on the embedded labels. The method provided by the present invention can construct a software system graph structure based on scientific and technological achievements. The graph structure can not only effectively organize scientific and technological achievements and accurately describe each system, but also search and compare scientific and technological achievements and integrate them into new scientific and technological achievement solutions.

Description

Technical Field

[0001] The present invention relates to the field of knowledge graph technology, and specifically to a method and system for expressing scientific and technological achievements based on software system description. Background Art

[0002] Scientific and technological achievements can be expressed in a variety of forms, such as through textualized intellectual property rights, representing various inventions and methods. In the IT field, many scientific and technological achievements can be displayed through corresponding practical software systems. For example, the Beijing Smart Government Online Service Platform can be considered a software system corresponding to scientific and technological achievements related to smart government. A reasonable description of this software system not only facilitates content analysis of the system but also allows for the characterization of the corresponding scientific and technological achievements.

[0003] Generally speaking, such software systems contain multiple interfaces, each of which covers multiple functional points, resulting in a complex system architecture. Natural language descriptions of these scientific and technological achievement software systems are difficult to describe and characterize. For example, system description documents and system function specifications can only describe the functional points within a single system, which limits their ability to describe these functional points and makes it difficult to accurately describe the software system structure.

[0004] Scientific and technological achievements can be displayed through corresponding software systems. Therefore, a reasonable description of the software system can effectively depict the corresponding scientific and technological achievements. The description of a software system includes a description of its structure and functions. Because the specific functional interfaces of software systems have complex technical relationships and diverse and complex relationships, they can be expressed using graph structures. Therefore, how to implement the description of one or more scientific and technological achievement software systems using graph structures has become an urgent problem. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention provides a method and system for expressing scientific and technological achievements based on software system description.

[0006] The technical solution of the present invention is: a method for expressing scientific and technological achievements based on software system description, including:

[0007] Step S1: obtaining at least one scientific and technological achievement software system, traversing each interface from the homepage of the software system, extracting the hierarchical structure of the software system, and obtaining a directed acyclic graph of the hierarchical architecture of at least one of the software systems;

[0008] Step S2: labeling each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standard natural language text containing specific semantics and is divided into multiple dimensions; and embedding the semantic label as a node in the directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, fusing the multiple software system graph structures at the label nodes with a similarity threshold above the similarity threshold to obtain a graph structure of scientific and technological achievements;

[0009] Step S3: using the search terms, searching for nodes in the graph structure of the scientific and technological achievements, thereby realizing the retrieval of scientific and technological achievements in the plurality of software systems;

[0010] Step S4: For two given scientific and technological achievements in the same or different fields, the scientific and technological achievements can be compared in the graph structure of the scientific and technological achievements, and the two scientific and technological achievements can be integrated.

[0011] Compared with the prior art, the present invention has the following advantages:

[0012] This invention discloses a method for expressing scientific and technological achievements based on software system descriptions. This method maps the description of the software system to the description of scientific and technological achievements in sub-fields, thereby organizing and associating the scientific and technological achievements. This method can facilitate the retrieval of scientific and technological achievements within a field, and by comparing scientific and technological achievements, it can link the corresponding business functions and technical points to generate solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a flowchart of a method for expressing scientific and technological achievements based on software system description in an embodiment of the present invention;

[0014] Figure 2 A schematic diagram of a software system diagram structure in an embodiment of the present invention;

[0015] Figure 3 Schematic diagram of the comparison process of two scientific and technological achievements in an embodiment of the present invention;

[0016] Figure 4 Schematic diagram of the integration process of two scientific and technological achievements in an embodiment of the present invention;

[0017] Figure 5 This is a structural block diagram of a scientific and technological achievement expression system based on software system description in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The present invention provides a method for expressing scientific and technological achievements based on software system description, which can realize the retrieval and comparison of scientific and technological achievement points, connect the business functions and technical points corresponding to the scientific and technological achievements, and generate solutions.

[0019] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below through specific implementation and in conjunction with the accompanying drawings.

[0020] Example 1

[0021] like Figure 1 As shown, an embodiment of the present invention provides a method for expressing scientific and technological achievements based on software system description, including the following steps:

[0022] Step S1: obtaining at least one scientific and technological achievement software system, traversing each interface from the homepage of the software system, extracting the hierarchical structure of the software system, and obtaining a directed acyclic graph of the hierarchical architecture of at least one software system;

[0023] Step S2: Label each interface in the software system with at least one semantic label in each dimension, where the semantic label is a standardized natural language text containing specific semantics and is divided into multiple dimensions; and embed the semantic labels as nodes into a directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, merge the multiple software system graph structures at the label nodes above the similarity threshold to obtain a graph structure of scientific and technological achievements;

[0024] Step S3: Using the search terms, search for nodes in the graph structure of scientific and technological achievements to achieve retrieval of scientific and technological achievements in multiple software systems;

[0025] Step S4: For two given scientific and technological achievements in the same or different fields, the scientific and technological achievements can be compared in the graph structure of the scientific and technological achievements, and the two scientific and technological achievements can be integrated.

[0026] In one embodiment, the above step S1: obtaining at least one scientific and technological achievement software system, traversing each interface from the homepage of the software system, extracting the hierarchical structure of the software system, and obtaining a directed acyclic graph of the hierarchical architecture of the at least one software system, specifically includes:

[0027] Step S11: Taking the homepage or top-level interface of the software system as the root node, traverse all interfaces in a depth-first traversal manner; at the same time, record the link relationship from one interface to another interface and add it to the interface link relationship set;

[0028] A scientific and technological achievement software system usually refers to a practical software system with application value, such as a complete web page system or application software. For example, the Beijing Smart Government Online Service Platform is a typical smart government system with high practical value. Such a software system consists of multiple interfaces, and each interface consists of multiple functional points, such as displaying data, layout display, click interaction, operation control and other functions for user operation. For each software system, the system is decomposed based on the interface as the basic unit, and each interface of the system is obtained as the atomic unit of interface organization. There are various complex relationships between the interfaces of the software system, such as the hierarchical relationship in the navigation bar, the jump link relationship of the interface itself, etc. The link relationship from one interface to another is added to the interface link relationship set.

[0029] Step S12: In the interface link relationship set, according to the business attribute specification of the software system, the association relationship of an interface containing multiple function points connected to the interface of a specific function point is defined as a downward association relationship; otherwise, it is defined as an upward association relationship; association relationships that have no direct connection in business attributes are defined as other association relationships; the downward association relationships are filtered out from the interface link relationship set as edges, and the interfaces connected by the edges are used as nodes to construct a directed acyclic graph.

[0030] Generally speaking, the functional interfaces of software systems described above have a hierarchical structure. For example, in an e-government network, a navigation page for personal services can lead to multiple specific service windows. Therefore, the navigation page can be considered the upper-level interface for specific services, while each service window is at the same level. To capture this hierarchical structure, these relationships need to be standardized based on the software system's service attributes, categorizing them into upward, downward, and other relationships. Downward relationships are defined as links from an interface containing multiple functional points to a specific interface within that function. For example, in the example above, if the navigation page for personal services leads to a specific service window, then the navigation interface and the specific service window have a downward relationship. Conversely, upward relationships, such as a specific service window and the navigation page, are considered upward relationships. Relationships without direct connections based on service attributes are considered other relationships, such as jumps to friendly links or jumps to the next page. To minimize the complexity of the software system's graph representation and remove redundant edges, only downward relationships are retained when extracting the system's hierarchical structure, while upward and other relationships are not retained.

[0031] The association relationship between interfaces can be recorded by traversing the software system interface. Traverse the system by manual clicks or machine design rule simulation. When entering another interface according to business logic from one interface, record the downward association between the two interfaces. Organize the software system in a hierarchical manner based on these association relationships. The specific method of obtaining the hierarchical structure of the software system is: take the homepage or top-level interface of the application software system as the root node, that is, the top layer in the hierarchy. For each interface, obtain the interface containing the downward association relationship, as the relationship from the upper interface to the lower interface as a downward edge in the hierarchy. In this way, all the connection edges in the software system are obtained to form the hierarchical structure of the software system;

[0032] The software system hierarchy is a tree or directed acyclic graph with the software system homepage as the root node, which is represented by a set of triples. Its formal definition is:

[0033] System = {P, PR}

[0034] Where P is the set of all interfaces, defined as:

[0035] P = {p|p is an interface}

[0036] PR is a set representing the hierarchical relationship between interfaces, defined as:

[0037] PR={(p i ,'sub',p j )|p i ,p j ∈P}

[0038] Among them, 'sub' means from p i to p j Downward association relationship;

[0039] The root node p at the top of the hierarchy root is defined as:

[0040]

[0041] In one embodiment, the above step S2: annotating each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standard natural language text containing specific semantics and is divided into multiple dimensions; and embedding the semantic label as a node in a directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, fusing the multiple software system graph structures at the label nodes above the similarity threshold to obtain a graph structure of scientific and technological achievements, specifically including:

[0042] Step S21: Through crowd-labeling, label each interface in the software system with at least one semantic tag in each dimension. Semantic tags are used to describe the functional semantics, application scenarios, and specific business of the interface. In the field of computer technology, "AI," "Internet of Things," and "Cloud Computing" can all be used as semantic tags. The semantic tags of each dimension themselves have a hierarchical relationship, such as "deep learning" and "machine learning" are both subtags of "AI." The acquisition of labels and their architecture can be constructed by experts, extracted manually from authoritative literature in the field, or extracted through machine-defined rule algorithms.

[0043] The semantic labels of the embodiment of the present invention are also represented by a set of triples. The formal definition is similar to the system hierarchy. The semantic labels of each dimension are represented by the following definition:

[0044] Tag={T,TR}

[0045] T = {t|t is a label}

[0046] TR={(t i ,'sub',t j )|t i ,t j ∈T}

[0047] The type of label is called dimension. The set of all dimensions is represented by Q = {q0,q1,q2,…}, which satisfies the finiteness A dimension q contains all the labels that describe the properties of an aspect of the interface:

[0048]

[0049] And any label belongs to and only belongs to one dimension:

[0050]

[0051] The label T can be divided into different subsets by different dimensions

[0052] T q ={t|t∈q},q∈Q

[0053] Step S22: embed the semantic labels as nodes and the corresponding relationships between interfaces and semantic labels as edges into a directed acyclic graph to construct a software system graph structure. This process is called interface description of the software system.

[0054] Specifically, for each interface, a set of labels that conform to the interface is used to describe the functional semantics, application scenarios, specific business, etc. of the interface. Interface description can be completed through the "crowd-labeling" method. For each interface, multiple people select appropriate labels to describe the interface; or domain experts can select appropriate labels for description; through the interface description of the software system, the software system can be mapped to a graph structure. The information that uses semantic labels to describe the interface can be embedded in the graph structure as edges in the graph structure; and the labels and system interfaces are abstracted as nodes in the graph. Through such a mapping method, the scientific and technological software system is organized in the form of a graph, and the system interface description can be formally expressed as:

[0055] Description = {P, T, DR}

[0056] DR={(p,dim,t)|p∈P,t∈T}

[0057] For labels of different dimensions, the dim used is also different; dim refers to the dimension of label t, and the triplet means: under a certain dimension dim, label t can be used to describe the interface p.

[0058] Whether it is an interface q or a label t, its attribute is defined as: a description of the nature of the software system interface or label. Its value can be continuous or discrete, and the value range is determined by the real-world meaning corresponding to different attributes. It can be formalized as follows:

[0059]

[0060] The set of property names of all attributes of all objects is represented by K, n i Represents any interface or label.

[0061] In summary, the following formal definition can be made for the software system diagram structure:

[0062] Graph = {P, T, PR, TR, DR}

[0063] For example, in the e-government network, from the personal service navigation page to the household registration window page, this edge can be represented by a triple ("personal service", "sub", "household registration"), where sub represents the association between the "personal service" interface and the lower-level interface "household registration".

[0064] For the Beijing Smart Government Network, the triple ("Beijing Smart Government Network," "function," "One-stop service") represents the connection between the network and the label "One-stop service." The label "One-stop service" describes the function of the Beijing Smart Government Network, where the predicate "function" indicates a description of the function. The triple ("Beijing Smart Government Network," "application," "smart government") represents the application domain of the network, which is smart government. The predicate "application" indicates a description of the application domain.

[0065] The hierarchical relationship between tags is captured by natural language relations, describing the relationship between tags. The starting and ending nodes of an edge are both tag nodes, and the predicate is the downward hierarchical relationship between tags. For example, the edge ("artificial intelligence", "sub", "machine learning") indicates that machine learning is a subtag of artificial intelligence, and the predicate also uses "sub" to indicate the hierarchical structure.

[0066] like Figure 2 As shown in the figure, the left half of the subgraph, rooted at screen 1, represents the hierarchical structure of the software system. The right half shows the hierarchical structure of labels within a certain dimension. Screen 1 is connected to labels 1 and 2, indicating that labels 1 and 2 can be used to describe screen 1. Figure 2 A part of the software system diagram structure that combines the hierarchical structure of the software system and its labels is shown. The complete diagram structure contains labels of multiple dimensions.

[0067] Step S23: When there are multiple software systems, repeat steps S21 to S22 to obtain multiple software system graph structures, perform similarity calculation on the semantic labels of the multiple software system graph structures, merge the semantic label nodes that are greater than a threshold, and obtain a complete graph structure of scientific and technological achievements;

[0068] Specifically, for multiple Figure 2 The software system graph structure shown is organized into an overall graph structure, which is a complete graph structure of scientific and technological achievements, by fusing semantic label nodes through semantic label similarity calculation.

[0069] In one embodiment, the above step S3: using the search term to search for nodes in the graph structure of scientific and technological achievements to achieve the retrieval of scientific and technological achievements in multiple software systems, specifically includes:

[0070] Step S31: According to the given search terms in several dimensions, matching the label nodes that are identical or partially identical to the search terms based on the node names in the graph structure of scientific and technological achievements;

[0071] Step S32: Filtering out the system interface nodes connected to all matching label nodes from the node connection relationship in the graph structure of the scientific and technological achievements as the returned interface result set;

[0072] Step S33: For each interface node in the interface result set, continuously recursively traverse its parent interface in turn until the root node, and restore the organizational structure of the interface node, thereby obtaining the hierarchical information of the interface node corresponding to the search term in the software system.

[0073] In one embodiment, the above step S4: for two given scientific and technological achievements in the same or different fields, the scientific and technological achievements can be compared in the graph structure of the scientific and technological achievements, and the two scientific and technological achievements can be integrated, specifically including:

[0074] Step S41: For two given scientific and technological achievements in the same or different fields, construct a feature vector for the interface in the graph structure of the scientific and technological achievements based on the semantic labels embedded in the two, as the representation vector of the specific scientific and technological achievement point;

[0075] Step S42: Starting from the leaf nodes of the scientific and technological achievements, pairing the leaf nodes is performed based on the eigenvectors to obtain the achievement point similarity, and combinations with similarity less than or equal to the hyperparameter ψ are filtered out; when a leaf node matches more than one leaf node in other systems, the Hungarian algorithm is used to obtain the optimal match of the combined similarity in the similarity matrix of the two, and the corresponding relationship between the leaf nodes of the two different scientific and technological achievements is established;

[0076] Step S43: Calculate the similarity of each non-leaf node in each layer from the bottom up, from the matched leaf node to the root node of each scientific and technological achievement; if the similarity value of two non-leaf nodes is greater than a certain set hyperparameter They are regarded as functionally similar non-leaf nodes, and their corresponding relationships are recorded. For non-leaf nodes, the Hungarian algorithm is also used to obtain the optimal matching of non-leaf nodes. Thus, all similar and non-similar matching relationships of the two scientific and technological achievements are obtained, thereby realizing the comparison of scientific and technological achievements.

[0077] Step S44: In the similar result matching relationship, starting from the leaf nodes, delete the unmatched result points from the bottom up, and reconstruct the common result tree of the matching result points: for a pair of matching leaf nodes, recursively search upward for the ancestor nodes of the two until the closest ancestor node that can match is found and used as the parent node of the two matching leaf nodes, thereby forming a common result tree;

[0078] Step S45: From the graph structure of scientific and technological achievements, add non-similar achievement points to the common achievement tree from top to bottom, and finally merge the two scientific and technological achievements into a new scientific and technological achievement tree. The new scientific and technological achievement tree integrates the similar achievement points and different achievement points in the two scientific and technological achievements.

[0079] like Figure 3 This diagram illustrates the comparison process for two scientific achievements. For two given scientific achievements, A and B, starting from their lowest-level achievement nodes, the achievement nodes are paired based on their feature vectors. Similarity is determined based on the degree of name character overlap, word vector cosine values, or label overlap. Pairings with similarities less than or equal to the hyperparameter ψ are filtered out, while pairs with similarities greater than ψ are retained. A single leaf node may correspond to multiple leaf nodes, but scientific achievements should only match at most one achievement node. Therefore, the Hungarian algorithm is used to find the optimal match based on the combined similarity in the similarity matrix between the two. This establishes the corresponding leaf nodes for the two different scientific achievements: (A1211, B211), (A122, B212), (A221, B112), and (A21, B12).

[0080] Assume that two systems A and B match a total of K leaf node pairs, and assign a K-dimensional one-hot feature vector to each matching node pair (the position corresponding to the node pair is set to 1, and the rest are set to 0). For unmatched leaf nodes, their feature vectors are K-dimensional all-zero vectors.

[0081] Starting from the leaf nodes, calculate the eigenvector F of each non-leaf node from bottom to top by summing up until the root node. Then calculate the pairwise cosine similarity based on the eigenvector F of each non-leaf node. If the similarity value of two achievement points is greater than a certain set hyperparameter ψ, they are regarded as functionally similar achievement points, and their corresponding relationship is recorded. For non-leaf nodes, the Hungarian algorithm is also used to obtain the optimal match of non-leaf nodes. Establish the corresponding relationship between the non-leaf nodes in two different scientific and technological achievements: (A21, B12), (A1, B2), and (A, B). Through the above correspondence, all similar achievement points and different achievement points between the two scientific and technological achievements can be identified, thereby realizing the comparison of scientific and technological achievements.

[0082] like Figure 4 This diagram shows the fusion process of two scientific and technological achievements. Based on the corresponding relationship between the nodes in the scientific and technological achievements in the above example, a common achievement tree of matching achievement points is reconstructed from the leaf nodes upwards. The steps are as follows: the matching nodes at the bottom layer are used as common leaf nodes; for two matching leaf nodes, the ancestor nodes are searched upwards recursively until the closest matching ancestor node is found, which becomes the parent node of the two matching leaf nodes.

[0083] From the original scientific and technological achievements A and B, the unmatched non-similar achievement points are added to the public achievement tree from top to bottom, and finally the two scientific and technological achievements are merged into a new scientific and technological achievement, which covers the similar achievement points and different achievement points in the two scientific and technological achievements.

[0084] For example, there are two existing scientific and technological achievements, "City Brain" and "City Cloud Brain". By constructing a part of the software system graph structure that includes these two software systems, the relationship between the interface of the software system of the scientific and technological achievements, scientific and technological achievement points, and tags is described. The graph is very complex. When using the graph structure, it is necessary to use tags to search for achievement points and systems that meet the search conditions in the graph. The search tags contain multiple dimensions, such as business dimensions, technical dimensions, regional dimensions, etc. The system returns achievement points that meet all the search conditions at the same time, returns the corresponding achievement points and the system they are in, and returns multi-dimensional information such as connection tags.

[0085] Furthermore, the graph structure allows for the comparison of scientific and technological achievements, providing guidance for the generation of domain solutions. For example, through the description of the graph structure, it is possible to describe the modules, functions, and business areas covered by the "City Brain" scientific and technological achievement. The functional points required for each module and the technologies required to implement each functional point can be identified. By comparing the scientific and technological achievements of the "City Brain" and "City Cloud Brain," it is possible to view the common functional points and the different functional points of each, and to identify the technologies required to implement each functional point. Therefore, by connecting the functional points as needed, a new scientific and technological achievement solution can be formed.

[0086] This invention discloses a method for expressing scientific and technological achievements based on software system descriptions. This method maps the description of the software system to the description of scientific and technological achievements in sub-fields, thereby organizing and associating the scientific and technological achievements. This method can facilitate the retrieval of scientific and technological achievements within a field, and by comparing scientific and technological achievements, it can link the corresponding business functions and technical points to generate solutions.

[0087] Example 2

[0088] like Figure 5 As shown, the embodiment of the present invention provides a scientific and technological achievement expression system based on software system description, including the following modules:

[0089] The software system directed acyclic graph construction module 51 is used to obtain at least one scientific and technological achievement software system, traverse each interface from the software system's homepage, extract the hierarchical structure of the software system, and obtain a directed acyclic graph of the hierarchical architecture of the at least one software system;

[0090] The graph structure construction module 52 of scientific and technological achievements is used to label each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standardized natural language text containing specific semantics and is divided into multiple dimensions; and the semantic labels are embedded as nodes in a directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, the multiple software system graph structures are merged at the label nodes with a similarity threshold above to obtain the graph structure of the scientific and technological achievements;

[0091] The scientific and technological achievements retrieval module 53 is used to use the search terms to search for nodes in the graph structure of scientific and technological achievements, thereby realizing the retrieval of scientific and technological achievements in multiple software systems;

[0092] The scientific and technological achievement comparison and fusion module 54 is used to compare two given scientific and technological achievements in the same or different fields in the graph structure of the scientific and technological achievements, and to fuse the two scientific and technological achievements.

[0093] The above embodiments are provided for the purpose of describing the present invention only and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims. Various equivalent substitutions and modifications made without departing from the spirit and principles of the present invention are intended to be within the scope of the present invention.

Claims

1. A method for expressing scientific and technological achievements based on software system description, characterized in that: include: Step S1: obtaining at least one scientific and technological achievement software system, traversing each interface from the homepage of the software system, extracting the hierarchical structure of the software system, and obtaining a directed acyclic graph of the hierarchical architecture of at least one of the software systems; Step S2: Label each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standard natural language text containing specific semantics and is divided into multiple dimensions; and embed the semantic label as a node in the directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, fuse the multiple software system graph structures at the label nodes above the similarity threshold to obtain a graph structure of scientific and technological achievements; wherein the semantic label is represented by the following definition: The type of label is called dimension; the collection of all dimensions is called Indicates that it satisfies finiteness ; One dimension Contains all tags that describe an aspect of the interface: And any label belongs to and only belongs to one dimension: The labels can be Divided into different subsets : ； Step S3: using the search terms, searching for nodes in the graph structure of the scientific and technological achievements, thereby realizing the retrieval of scientific and technological achievements in the plurality of software systems; Step S4: For two given scientific and technological achievements in the same or different fields, the scientific and technological achievements can be compared in the graph structure of the scientific and technological achievements, and the two scientific and technological achievements can be integrated.

2. The method for expressing scientific and technological achievements based on software system description according to claim 1, characterized in that: The step S1: obtaining at least one scientific and technological achievement software system, traversing each interface from the homepage of the software system, extracting the hierarchical structure of the software system, and obtaining a directed acyclic graph of the hierarchical architecture of at least one of the software systems, specifically includes: Step S11: Taking the homepage or top-level interface of the software system as the root node, traverse all interfaces in a depth-first traversal manner; at the same time, record the link relationship from one interface to another interface and add it to the interface link relationship set; Step S12: In the interface link relationship set, according to the business attribute specification of the software system, the association relationship of an interface containing multiple function points connected to an interface of a specific function point is defined as a downward association relationship; otherwise, it is defined as an upward association relationship; association relationships that have no direct connection in business attributes are defined as other association relationships; the downward association relationships are filtered out from the interface link relationship set as edges, and the interfaces connected by the edges are used as nodes to construct a directed acyclic graph.

3. The method for expressing scientific and technological achievements based on software system description according to claim 1, characterized in that: Step S2: annotating each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standard natural language text containing specific semantics and is divided into multiple dimensions; and embedding the semantic label as a node in the directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, fusing the multiple software system graph structures at the label nodes above the similarity threshold to obtain a graph structure of scientific and technological achievements, specifically including: Step S21: labeling each interface in the software system with at least one semantic tag in each dimension through crowd-labeling, wherein the semantic tag is used to describe the functional semantics, application scenario, and specific business of the interface; Step S22: embedding the semantic labels as nodes and the corresponding relationships between the interfaces and the semantic labels as edges into the directed acyclic graph to construct a software system graph structure; Step S23: When there are multiple software systems, repeat steps S21 to S22 to obtain multiple software system graph structures, calculate the similarity of the semantic labels of the multiple software system graph structures, and fuse the semantic label nodes that are greater than the threshold to obtain a complete graph structure of scientific and technological achievements.

4. The method for expressing scientific and technological achievements based on software system description according to claim 1 is characterized in that: The step S3: using the search terms to search for nodes in the graph structure of the scientific and technological achievements, thereby realizing the search for scientific and technological achievements in multiple software systems, specifically includes: Step S31: According to the given search terms in several dimensions, matching the label nodes that are identical or partially identical to the search terms based on the node names in the graph structure of the scientific and technological achievements; Step S32: Filtering out the system interface nodes connected to all matching label nodes from the node connection relationship in the graph structure of the scientific and technological achievements as the returned interface result set; Step S33: For each interface node in the interface result set, continuously recursively traverse its upper interface in turn until the root node, and restore the organizational structure of the interface node, thereby obtaining the hierarchical information of the interface node corresponding to the search term in the software system.

5. The method for expressing scientific and technological achievements based on software system description according to claim 1 is characterized in that: Step S4: for two given scientific and technological achievements in the same or different fields, a comparison of the scientific and technological achievements can be performed in the graph structure of the scientific and technological achievements, and the two scientific and technological achievements can be integrated, specifically including: Step S41: for two given scientific and technological achievements in the same or different fields, construct a feature vector for the interface in the graph structure of the scientific and technological achievements based on the semantic labels embedded in the two, as a representation vector of the specific scientific and technological achievement point; Step S42: Starting from the leaf nodes of the scientific and technological achievements, pairing the leaf nodes is performed based on the eigenvectors to obtain the achievement point similarity, and combinations with similarity less than or equal to the hyperparameter ψ are filtered out; when a leaf node matches more than one leaf node in other systems, the Hungarian algorithm is used to obtain the optimal match of the combined similarity in the similarity matrix of the two, and the corresponding relationship between the leaf nodes in the two different scientific and technological achievements is established; Step S43: Calculate the similarity of each non-leaf node in each layer from the matched leaf nodes, from bottom to top, until the root node of each scientific and technological achievement; if the similarity value of two non-leaf nodes is greater than a certain set hyperparameter φ, they are regarded as functionally similar non-leaf nodes, and their corresponding relationship is recorded; for non-leaf nodes, the Hungarian algorithm is also used to obtain the optimal match of the non-leaf nodes; thereby obtaining all similar achievement matching relationships and non-similar achievement matching relationships in the two scientific and technological achievements, thereby achieving comparison of scientific and technological achievements; Step S44: In the similar result matching relationship, starting from the leaf nodes, unmatched result points are deleted from the bottom up, and a common result tree of matching result points is reconstructed: for a pair of matching leaf nodes, the ancestor nodes of the two are searched upward recursively until the closest matching ancestor node is found as the parent node of the two matching leaf nodes, thereby forming a common result tree; Step S45: From the graph structure of the scientific and technological achievements, non-similar achievement points are added to the common achievement tree from top to bottom, and finally the two scientific and technological achievements are merged into a new scientific and technological achievement tree, which integrates the similar achievement points and different achievement points in the two scientific and technological achievements.

6. A scientific and technological achievement expression system based on software system description, characterized in that: Includes the following modules: A software system directed acyclic graph construction module is used to obtain at least one scientific and technological achievement software system, traverse each interface from the homepage of the software system, extract the hierarchical structure of the software system, and obtain a directed acyclic graph of the hierarchical architecture of at least one of the software systems; The graph structure construction module of scientific and technological achievements is used to label each interface in the software system with at least one semantic label in each dimension, wherein the semantic label is a standard natural language text containing specific semantics and is divided into multiple dimensions; and the semantic label is embedded as a node in the directed acyclic graph to obtain a software system graph structure; when there are multiple software system graph structures, the multiple software system graph structures are merged at the label nodes with a similarity threshold above the similarity threshold to obtain the graph structure of scientific and technological achievements; wherein the semantic label is represented by the following definition: The type of label is called dimension; the collection of all dimensions is called Indicates that it satisfies finiteness ; One dimension Contains all tags that describe an aspect of the interface: And any label belongs to and only belongs to one dimension: The labels can be Divided into different subsets : ； A scientific and technological achievement retrieval module is used to use search terms to search for nodes in the graph structure of the scientific and technological achievements, thereby realizing the retrieval of scientific and technological achievements in multiple software systems; The scientific and technological achievements comparison and fusion module is used to compare two given scientific and technological achievements in the same or different fields in the graph structure of the scientific and technological achievements, and to fuse the two scientific and technological achievements.

Images