An Information Collection Method and System for the Additive Manufacturing Industry Based on Artificial Intelligence

Abstract

The present invention belongs to the field of data collection, and provides a method and system for collecting information in the additive manufacturing industry based on artificial intelligence, including obtaining original data related to the additive manufacturing industry from multi-source heterogeneous data channels; preprocessing the original data; performing representation learning on entities in the knowledge graph to obtain the analogy score relationship between entities, which is used to infer potential technology evolution directions; fusing the generated new terms and the inferred potential technology evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges; performing periodic updates and inference calculations on the dynamic semantic knowledge graph, and outputting structured industry intelligence information, where the industry intelligence information includes new terms, high-potential technology paths, cross-domain technology integration trends, and concept evolution trajectories. The present invention can adapt to the collection of information in the additive manufacturing industry where information changes rapidly.

Description

Technical Field

[0001] The present invention belongs to the field of data collection, and more particularly relates to a method and system for collecting information in the additive manufacturing industry based on artificial intelligence. Background Art

[0002] With the continuous progress of artificial intelligence technology and natural language processing capabilities, methods for automatically obtaining industry-related information based on intelligent systems have been widely used in fields such as policy research, industrial analysis, and technology forecasting. Existing industry information collection methods mainly use technical means such as keyword retrieval, rule extraction, and co-occurrence statistics, combined with traditional information scraping and tagging methods, to structurally organize text resources such as industry news, patent documents, and academic papers. However, the additive manufacturing industry is developing rapidly, with new technologies and new terms emerging continuously. The existing technologies have obvious deficiencies in the following aspects:

[0003] Existing methods rely on manually constructed keyword vocabularies or domain entity libraries, and it is difficult to identify newly emerged terms, expressions, or unnamed emerging concepts in the additive manufacturing industry, and they cannot adapt to the dynamic characteristics of the evolving term system. For example, with the emergence of new processes such as multi-material hybrid printing and laser energy adaptive control, a large number of professional terms have not been standardized, and traditional methods are difficult to identify or classify.

[0004] Existing methods mainly focus on the extraction and statistics of explicit knowledge, lacking the ability to understand the implicit semantic relationships between terms in the additive manufacturing industry, and unable to simulate the reasoning, induction, and concept evolution judgment in the cognitive process of domain experts, and unable to identify the functional analogy relationships, structural similarities, or technology migration paths between terms.

[0005] Existing systems are difficult to build an updatable, inferable, and evolvable knowledge structure. Most of the results obtained are static knowledge tables or collections of entries, lacking a time-series update mechanism and semantic map support, and unable to model and predict the evolution trajectory of concepts, trend intersection points, or cross-domain integration behaviors.

[0006] Therefore, there is an urgent need for an artificial intelligence-based method for collecting information in the additive manufacturing industry that can integrate deep language modeling and analogical reasoning capabilities, has the functions of new term discovery and dynamic trend analysis, and supports knowledge structuring and map evolution and update, so as to break through the semantic bottleneck and structural limitations of existing technologies and improve the intelligent perception and deduction capabilities of the system in the complex industry background of the rapid development of emerging technologies. Summary of the Invention

[0007] To solve the problems in the existing technology, the present invention provides an artificial intelligence-based method for collecting information in the additive manufacturing industry, including the following steps:

[0008] Obtain the original data related to the additive manufacturing industry from multi-source heterogeneous data channels according to the preset industry theme;

[0009] Preprocess the original data, and use a deep language representation model to encode the original data to generate a high-dimensional semantic vector representation;

[0010] Construct a term space based on the high-dimensional semantic vectors, and through a concept combination function, combine and / or transform and / or expand the terms or technical descriptions in the semantic neighborhood to generate new terms that do not directly appear in the original data but are semantically derivable;

[0011] Based on the identified industry entities, construct an analogical relationship graph with similarities in structure, function, or evolutionary path, and perform representation learning on the entities in the graph through a graph neural network model to obtain the analogical score relationship between entities, which is used to infer potential technological evolution directions;

[0012] Integrate the generated new terms and the inferred potential technological evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges;

[0013] Periodically update and perform inference calculations on the dynamic semantic knowledge graph, and output structured industry intelligence information, which includes new terms, high-potential technology paths, cross-domain technology integration trends, and concept evolution trajectories.

[0014] Further, the steps of generating new terms that do not directly appear in the original data through a concept combination function include:

[0015] Based on the term space, define multiple types of concept combination functions, including: term combination function, term transformation function, and term expansion function;

[0016] Among them, the term combination function is used to perform structural-level combination of two or more semantically related or complementary terms to form new terms expressing more complex or refined concepts; the term transformation function is used to perform semantic replacement, expression transformation, or syntactic rewriting on existing terms to form semantically equivalent or evolved variant terms; the term expansion function is used to add context modification components, application conditions, or technical constraints to terms to form semantically enhanced terms.

[0017] Further, the steps of constructing an analogical relationship graph with similarities in structure, function, or evolutionary path and performing representation learning on the entities in the graph through a graph neural network model include:

[0018] Construct a set of industry entities and entity attribute tensors, and construct structural analogy edges, functional analogy edges, and evolutionary analogy edges based on the structural similarity, functional relevance, or evolutionary chain among entities in terms of control logic, module composition, and process constraints;

[0019] Based on the constructed graph structure, use a graph neural network to update the node vectors in multiple rounds, and aggregate the adjacent node information based on the edge type weights in each round of iteration;

[0020] Calculate the semantic similarity scores between nodes, which are used to determine the analogy strength and potential evolutionary direction between technical entities.

[0021] Furthermore, the steps of constructing a dynamic semantic knowledge graph including explicit entity relationships and new concept inference edges include:

[0022] Connect the new terms generated in step S30 to the original industry knowledge graph, construct new term nodes, and attach term source type, generation path, first appearance time, semantic cluster label attribute information;

[0023] According to the term generation path and semantic similarity, establish semantic relationship edges such as term extension edges, combination derivation edges, and function enhancement edges;

[0024] Connect the potential technical evolution paths inferred in step S40 to the graph in the form of analogy edges or trend edges. For terms that have not been verified, they can be set as placeholder nodes and the "to be confirmed" status can be set;

[0025] Record the graph snapshot difference and mark the node update time and update frequency for subsequent graph evolution path tracking and structure maintenance.

[0026] Furthermore, the steps of outputting structured industry intelligence information include:

[0027] Set up a graph update scheduling mechanism to perform incremental updates on the dynamic semantic knowledge graph regularly, including node attribute updates, edge weight adjustments, node active status evaluations, and version snapshot generations;

[0028] Perform path reasoning, graph representation learning, hot spot clustering analysis, and multi-hop logic chain tracking on the updated graph to obtain term semantic change trends, fusion nodes, and technical paths;

[0029] Organize the analysis results into structured intelligence for output. The output content includes: new term intelligence reports, high-potential technical path lists, cross-domain fusion trend graphs, and concept evolution trajectory graphs, and supports export in the form of structured data files or combination of text and graphics.

[0030] On the other hand, the present invention also provides an information collection system for the additive manufacturing industry based on artificial intelligence, including the following modules:

[0031] A data acquisition module, configured to obtain raw data related to the additive manufacturing industry from multi-source heterogeneous data channels according to a preset industry theme;

[0032] A semantic encoding module, configured to preprocess the raw data, and encode the raw data by using a deep language representation model to generate a high-dimensional semantic vector representation;

[0033] A term generation module, configured to construct a term space based on the high-dimensional semantic vector, and combine and / or transform and / or expand terms or technical descriptions in the semantic neighborhood through a concept combination function to generate new terms that do not directly appear in the raw data but are semantically derivable;

[0034] An analogical reasoning module, configured to construct an analogical relationship graph with similarities in structure, function, or evolutionary path based on the identified industry entities, and perform representation learning on the entities in the graph through a graph neural network model to obtain the analogical score relationship between the entities for inferring potential technology evolution directions;

[0035] A graph construction module, configured to fuse the generated new terms and the inferred potential technology evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges;

[0036] An intelligence output module, configured to periodically update and perform inference calculations on the dynamic semantic knowledge graph, and output structured industry intelligence information, where the industry intelligence information includes new terms, high-potential technology paths, cross-domain technology integration trends, and concept evolution trajectories.

[0037] Further, the term generation module includes:

[0038] A concept combination function library, configured to define a term combination function, a term transformation function, and a term expansion function;

[0039] A combination control unit, configured to select term pairs in the semantic neighborhood in the term space, call the concept combination function to generate term semantic vectors, and input the generated term semantic vectors into a term decoder to generate a language-readable new term text;

[0040] A term scoring unit, configured to perform semantic consistency scoring and knowledge matching verification on the generated new terms, and mark the terms with a scoring result higher than a preset threshold as semantic emergence terms.

[0041] Further, the analogical reasoning module includes:

[0042] An entity attribute construction unit, configured to construct an industry entity set and its attribute tensor;

[0043] The analogical relationship construction unit is used to construct structural analogical edges, functional analogical edges, and evolutionary analogical edges based on characteristics such as control logic, module composition, and process constraints;

[0044] The graph neural network learning unit is used to update the vectors of entity nodes through multi-round aggregation of adjacent node information in the graph structure;

[0045] The analogical scoring unit is used to calculate the semantic similarity based on node vectors and output the analogical scores between entities for subsequent evolutionary path reasoning.

[0046] Furthermore, the graph construction module further includes:

[0047] The node access unit is used to access new terms into the original industry graph in the form of nodes, attaching term sources, generation paths, first appearance times, and semantic cluster labels;

[0048] The relationship generation unit is used to establish term extension edges, combined derivative edges, and functional enhancement edges in the graph based on term generation paths and semantic similarity;

[0049] The inference edge access unit is used to add potential technology evolution paths to the graph in the form of analogical edges or trend edges, setting unvalidated term nodes as placeholder nodes and setting the "to be confirmed" status;

[0050] The evolution management unit is used to record the graph snapshot differences and maintain node update times, active statuses, and evolution tracking information.

[0051] Furthermore, the intelligence output module includes:

[0052] The update scheduling unit is used to periodically trigger graph incremental update operations to update node attributes, edge weights, and graph version snapshots;

[0053] The inference analysis unit is used to perform path reasoning, graph representation learning, hot term clustering, and multi-hop logic chain analysis on the updated knowledge graph;

[0054] The intelligence organization unit is used to extract new terms, high-potential technology paths, cross-domain trend nodes, and evolution trajectory information and organize them into a structured intelligence report;

[0055] The output interface unit is used to export the industry intelligence information in the form of structured data or graphic reports.

[0056] Compared with existing industry information collection technologies, the present invention significantly improves the system's capabilities in term recognition, trend insight, and knowledge structure organization by introducing a number of artificial intelligence technologies such as term space construction, concept combination functions, semantic emergence reasoning mechanisms, analogical relationship graphs, and graph neural network evolution modeling, and has the following beneficial effects:

[0057] Instead of relying on a fixed glossary of terms or manual rules, the present invention constructs a semantic vector space for terms and uses combination and / or transformation and / or extension mechanisms to automatically generate new terms that are not explicitly present in the original data but are semantically reasonable, breaking through the limitation of traditional methods that can only extract existing terms and significantly enhancing the ability to identify emerging concepts.

[0058] By constructing a multi-dimensional relationship graph of structural analogy, functional analogy, and evolutionary analogy, and using a graph neural network to deeply model the semantic commonalities between entities, potential migration paths and integration directions between different technologies can be identified, with the ability to simulate "heterogeneous analogy" in expert cognition and improving the system's cross-domain association and trend judgment levels.

[0059] The present invention can integrate explicit terms, newly generated terms, and analogy paths into a unified dynamic semantic knowledge graph, supporting periodic incremental updates of nodes and edges, semantic activity evaluation, and tracking of time evolution trajectories, providing structural support for trend evolution analysis.

[0060] By performing periodic reasoning calculations on the knowledge graph, the system can automatically output structured industry intelligence information, including a list of new terms, high-potential technology paths, cross-domain integration trends, and concept evolution chains, providing an intelligent auxiliary decision-making basis for policy formulation, enterprise R & D, technology investment, etc.

[0061] In summary, the present invention breaks through many limitations of traditional industry information collection technologies in terms of term recognition scope, semantic association depth, trend evolution modeling ability, and structural organization flexibility, and can be widely applied in industries with rapid term evolution and high technology cross-over such as additive manufacturing, electronic manufacturing, synthetic biology, and intelligent transportation, having significant technical value and practical prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0063] Figure 1 is a flowchart of the method of the present invention;

[0064] Figure 2 is a system block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] The following will give a preferred description of the invention in combination with the drawings and specific embodiments.

[0066] This embodiment solves the above problems through the following steps:

[0067] In one embodiment, referring to Figure 1 , the present invention provides an information collection method for the additive manufacturing industry based on artificial intelligence. By simulating the interpretation and deduction behaviors of domain experts on additive manufacturing industry data, new concepts, technical contexts, and industrial trends related to the target field are automatically mined, which is applicable to multiple scenarios such as policy research, market prediction, and technology reconnaissance.

[0068] Specifically, the method includes the following steps:

[0069] Step S10, obtain the original data related to the additive manufacturing industry from multi-source heterogeneous data channels according to a preset industry theme.

[0070] Specifically, the "industry theme" can be input by the user, preset by the task system, or automatically recommended according to industry hotspots. The theme is used to define the content scope and semantic boundary of information collection, such as "energy control technology of metal additive manufacturing equipment", "application of powder bed fusion (PBF) process in aerospace components", "market trend of new printing material Ti6Al4V", etc.

[0071] The multi-source heterogeneous data channels refer to an information collection containing various data formats, structures, and sources, including but not limited to:

[0072] Structured data sources: such as industry standard databases, technology supervision websites, manufacturing cost structures and equipment sales quantities in enterprise financial reports, etc.

[0073] Semi-structured data sources: such as patent documents, academic papers, technical white papers, etc., which contain structured field information and free text.

[0074] Unstructured data sources: such as news reports, enterprise website information, industry interviews, social media technology topics, forum comments, etc.

[0075] Multimodal data sources: such as physical pictures of printing equipment, videos of additive printing processes uploaded by users, spectral images of powder materials, conference recordings, etc.

[0076] Taking the additive manufacturing industry as an example, after the system receives the industry theme of "development trend of metal additive manufacturing in the aerospace field", the following data collection process is automatically triggered:

[0077] Retrieve the full text and metadata of papers containing keywords such as "laser powder bed fusion", "additive manufacturing in the aerospace field", "L-PBF Ti6Al4V", etc. from academic websites.

[0078] Retrieve the published patent texts and claim contents containing keywords such as "aeronautical parts", "additive manufacturing", and "print path optimization" from the public patent database.

[0079] Collect industry news on new equipment, new processes, and new materials in the past six months from 3D printing industry portals.

[0080] Extract the technical specifications, video demonstration links, and user feedback documents of the new model printers released on the company's official website.

[0081] Download the conference schedule, abstract collection, and expert speech PPT documents from the industry conference website.

[0082] Retrieve the explanatory videos containing "additive manufacturing of aeroengine blades processing" from platforms such as video platforms, and automatically identify the voice content therein and transcribe it into text.

[0083] After the data is collected, meta - attributes such as its source, time, language, and modality will be marked, and it will be uniformly transmitted to the subsequent pre - processing and semantic modeling steps.

[0084] Step S20: Pre - process the original data, and use a deep language representation model to encode the original data to generate a high - dimensional semantic vector representation.

[0085] The original data may come from various channels such as web texts, social media comments, news reports, etc., which may contain a large amount of noise information and non - standard expressions. The purpose of pre - processing is to improve the data quality to make it more suitable for subsequent model processing. Pre - processing mainly includes the following sub - steps:

[0086] Data cleaning: The original data may contain noise information such as HTML tags, special symbols, redundant spaces, and line breaks, which will interfere with the subsequent processing process. Therefore, it is necessary to remove them. For example, if the original data is news content crawled from a web page, there may be such as 、 HTML tags and special characters such as spaces need to be filtered out using regular expressions.

[0087] Stop word removal: Stop words, such as "的," "是," "在," and "和," appear frequently in text but contribute little to the text's semantics. Removing stop words can reduce data dimensionality and improve processing efficiency. You can use a predefined stop word list to filter out stop words from text.

[0088] Word segmentation: For Chinese text, continuous text needs to be segmented into individual words. Word segmentation is a fundamental step in Chinese natural language processing. Common word segmentation tools include Jieba Word Segmenter and HanLP. For example, the sentence "I love natural language processing" can be segmented into "I," "love," and "natural language processing."

[0089] Part-of-speech tagging and stemming (optional): Part-of-speech tagging assigns each word its part of speech, such as noun, verb, or adjective. Stemming reduces words to their stem form, such as "学" (learning) and "学者" (learner) to "学" (learning). These two steps can further improve data quality and feature consistency, but are not required for all tasks.

[0090] After completing data preprocessing, you need to use a deep language representation model to convert the text data into a high-dimensional semantic vector representation. The deep language representation model can learn the semantic information of the text and map it to a high-dimensional vector space, so that semantically similar texts are closer in the vector space. Here we use the BERT model as an example to illustrate:

[0091] Loading a pre-trained model: BERT is a pre-trained language model based on the Transformer architecture. It performs unsupervised learning on large amounts of text data and acquires rich linguistic knowledge. You can use Hugging Face's Transformers library to load the pre-trained BERT model and its corresponding tokenizer.

[0092] Input processing: Preprocessed text data is fed into the tokenizer, which converts it into an input format acceptable to the model. This typically requires adding special markers, such as [CLS] and [SEP], to indicate the start and end of a sentence, respectively. Furthermore, the words are converted to their corresponding word indices, and padding and truncation are performed to ensure that all input sequences have the same length.

[0093] Model inference: The processed input sequence is fed into the BERT model for inference. The model outputs the contextual representation of each word. The output vector corresponding to the [CLS] tag is usually used as the semantic representation of the entire sentence.

[0094] After being processed by the BERT model, the output vector corresponding to the [CLS] token is the high-dimensional semantic vector representation of the sentence. This vector usually has a relatively high dimension (such as 768 dimensions or 1024 dimensions) and contains the semantic information of the sentence. This vector can be used for subsequent tasks such as text classification, sentiment analysis, information retrieval, etc.

[0095] Step S30: Based on the high-dimensional semantic vector, construct a term space, and through a concept combination function, combine and / or transform and / or expand the terms or technical descriptions in the semantic neighborhood to generate new terms that do not directly appear in the original data but are semantically derivable.

[0096] This step is to further implement a term generation mechanism based on semantic space logical reasoning on the basis of completing the acquisition and semantic modeling of the original industry data. Through this step, the system can not only identify existing terms but also predictively generate new terms that are semantically related but not explicitly expressed, expand the industry semantic boundary, and enhance the trend prediction ability.

[0097] Industry terms often have the characteristic of periodic update. With the continuous emergence of new technologies, new materials, and new processes, their term systems are also evolving rapidly. For example, in the field of additive manufacturing in recent years, a series of sub-concepts such as "metal additive manufacturing", "multi-material composite printing", and "selective laser sintering" have emerged from "3D printing".

[0098] Traditional methods can only extract terms based on existing word lists or entity recognition models and are difficult to handle the following scenarios:

[0099] A new process that has not been named begins to be described in papers but has not been uniformly expressed;

[0100] The combination of terms from multiple different fields forms a new technology path;

[0101] Some forms of expression are in a fuzzy state but have a clear semantic reference.

[0102] This step is precisely to solve the above problems and proposes an overall mechanism of term space construction + semantic combination function + new term generation + multi-dimensional verification.

[0103] Specifically, step S30 includes the following sub-steps:

[0104] Step S301: Semantic neighborhood modeling and mining of term aggregation relationships

[0105] Based on the set of term semantic vectors obtained in the previous step , use a vector clustering method based on semantic distance (such as K-Means, hierarchical clustering, or spectral clustering) to construct a neighbor network between terms and extract a set of semantic neighborhoods , each term is mapped into the cluster structure of its semantic context.

[0106] According to the term occurrence frequency, context relevance, and neighborhood semantic density, the association strength between terms is adjusted, and finally a term relationship network with the following structure is constructed:

[0107] Strong semantic adjacency edges: such as "laser sintering" and "laser focusing control";

[0108] Potential replacement edges: such as "superalloy" and "high-entropy alloy";

[0109] Heterogeneous analogy edges: such as "metal powder bed melting" and "ceramic wire synchronous deposition".

[0110] Term space graph The edge set 𝐸 in records different types of combination possibilities.

[0111] Step S302: Combination function design and term generation vector construction

[0112] Based on the term space, the system defines multiple types of concept combination functions for different types of semantic reasoning tasks. The combination functions are divided according to different combination methods:

[0113] 1. Term combination

[0114] Two or more semantically related or complementary terms are combined at the structural level to form a new term expressing a more complex or refined concept. The combination relationship can be language structures such as coordination, modification, subject-predicate, noun-adjective, etc.

[0115] Design template structures, such as "[material] + [process]", "[functional word] + [equipment name]", and embed the term text for combination.

[0116] Exemplarily:

[0117] Term pairs:

[0118] "Multi-laser scanning"

[0119] "Titanium alloy powder"

[0120] After combination, it generates:

[0121] "Titanium alloy powder multi-laser scanning printing"

[0122] "Multi-laser titanium-based powder deposition process"

[0123] 2. Term transformation

[0124] Semantically replace existing terms, transform expressions, and rewrite grammar to form variant terms with different expressions but semantically equivalent or evolved, applicable to industry backgrounds with inconsistent term standards and coexisting multiple description methods.

[0125] The implementation methods include:

[0126] Semantic replacement: Replace a certain component in the term, such as replacing "laser" with "electron beam".

[0127] Structure inversion: Change "control energy density" to "energy density control".

[0128] Stylistic style transformation: For example, change "selective sintering technology" to "selective sintering method".

[0129] Vector difference operation: Simulate "laser printing": "metal" = X : "ceramic" ⇒ X = "ceramic printing"

[0130] Exemplarily:

[0131] Original term: "Optimization of high-entropy alloy printing path"

[0132] Transformed term:

[0133] "Adjustment of high-entropy printing path"

[0134] "Control of complex alloy printing parameters"

[0135] "Adaptive adjustment of multi-component alloy path"

[0136] 3. Term expansion

[0137] Add context modification components, specific context limitations, applicable object descriptions, or technical condition supplements to the original term to expand new term expressions and strengthen its semantic dimension and description clarity.

[0138] The implementation methods include:

[0139] Upper and lower limit expansion: Add specific limiting components before / after the original term, such as "space-grade", "microstructure", "sub-millimeter scale".

[0140] Process environment / constraint condition expansion: For example, add "under variable air pressure environment" to "fused deposition".

[0141] Semantic reference expansion: Expand "laser sintering" to "laser sintering technology applicable to multi-material hybrid molding".

[0142] Nested context vector fusion: Weightedly fuse the term vector and the context paragraph vector to generate a semantically enhanced term.

[0143] Exemplarily:

[0144] Original term: "Powder bed melting"

[0145] Expanded term candidates:

[0146] "Powder bed melting process under low - pressure environment"

[0147] "Powder bed melting technology applicable to refractory metals"

[0148] "Powder bed forming method dedicated to aerospace engine components"

[0149] The above - mentioned three types of operations can not only be used independently, but the system also allows them to be used in combination. For example:

[0150] First combine the terms and then perform transformation;

[0151] Add an expanded description to the combined terms;

[0152] Perform a restrictive expansion on the basis of the transformation.

[0153] Exemplarily:

[0154] Term 1: "Laser energy modulation"

[0155] Term 2: "Ceramic matrix composite"

[0156] Combination: → "Ceramic composite laser modulation"

[0157] Transformation: → "Composite ceramic laser multi - zone power regulation"

[0158] Expansion: → "Multi - zone power laser composite ceramic printing strategy applicable to heterogeneous interface control"

[0159] Step S303: Term candidate generation and multi - candidate hierarchical decoding

[0160] For each newly constructed vector , adopt the following strategy to generate term candidates:

[0161] Top - k Beam Search: Generate multiple term candidates with different syntactic structures but close semantics;

[0162] Semantic segmentation decoding: First generate the basic stem and then generate the modification structure;

[0163] Language structure evaluator scoring: Eliminate expressions that are verbose or logically unclear.

[0164] Exemplarily:

[0165] Combined input:

[0166] Term A: "High-Energy Density Scanning"

[0167] Term B: "Composite Ceramic Material"

[0168] Generate term candidates:

[0169] "High-Energy Density Composite Ceramic Melting"

[0170] "Laser Scanning of Composite Ceramic Materials"

[0171] "High-Energy Sintering Technology for Ceramic Composites"

[0172] Each candidate term retains its semantic generation path: starting term pair, combination method, vector distance, language model probability.

[0173] Step S304: Term Confidence Learning and Semantic Consensus Verification

[0174] To determine whether the term candidates are acceptable in the industry and semantically valid, the system introduces a confidence scoring model , and the scoring metrics include:

[0175] Semantic Consistency Score : The average semantic proximity to the input term;

[0176] Linguistic Naturalness Score : The language model's judgment on the smoothness of the generated expression;

[0177] Industry Consensus Score : The matching frequency of the candidate term in the known term graph or external database;

[0178] Substitutability Score : Whether it can naturally replace part of the existing expression.

[0179] Total Score:

[0180] ,

[0181] Among them, are configurable weight coefficients. Terms with scores higher than the threshold (e.g., 0.75) are marked as "semantically emerging terms" and enter the term library.

[0182] Step S305: Construction of Term Tracing Path and Preparation for Updating the Evolution Map

[0183] The system constructs a "term generation path" for all newly verified terms, including:

[0184] Source term list;

[0185] Combination function number used;

[0186] Semantic neighborhood

[0187] Decoding model version

[0188] Timestamp

[0189] Whether the external corpus matches

[0190] Build all candidates marked as semantic emergence terms into the nodes of the term evolution map to support trend modeling and knowledge fusion in the next step

[0191] Step S40: Based on the identified industry entities, construct an analogical relationship map with similarities in structure, function, or evolution path, and use a graph neural network model to perform representation learning on the entities in the map to obtain the analogical score relationship between entities for inferring potential technological evolution directions

[0192] The purpose of this step is to construct an analogical relationship map based on entity content such as industry terms, technology names, material types, and equipment parameters extracted in the previous steps, and use a graph neural network model to learn the representations of the entities in the map, so as to obtain the analogical score relationship between entities for inferring potential technological evolution paths, cross-domain migration paths, or process fusion trends that have not been clearly proposed but are semantically possible

[0193] Different from traditional co-occurrence maps, the focus of this step is to mine the structural analogy, functional similarity, and evolutionary relevance between entities, emphasizing the ability of analogical reasoning rather than explicit co-occurrence relationships

[0194] Specific implementation steps

[0195] Step S401: Construct an industry entity set and entity attribute tensors

[0196] The system first structures all industry terms, material names, equipment names, process methods, control strategies, etc. extracted in the previous steps to form an entity set

[0197] ,

[0198] Each entity is equipped with an attribute tensor , including but not limited to

[0199] Technical attributes (such as whether it is laser-related, whether it involves melting, whether it is powder / wire input);

[0200] Functional attributes (such as whether it controls energy, whether it supports multi-channel output);

[0201] Application scenarios (such as whether it is used in aviation, medical, micro-structure control);

[0202] Time attributes (first appearance time, active period, trend curve, etc.);

[0203] Module structure (such as print head structure, composition of control modules, etc.).

[0204] These attributes will be used to judge and score the structure - function analogy edges in the subsequent graph construction process.

[0205] Step S402: Establish analogy relation edges and construct graph spectrum structure

[0206] Taking the entity set E as nodes, the system constructs three types of analogy edges based on the structural similarity, functional relevance, and historical evolution path between entity attribute tensors:

[0207] Structural analogy edge

[0208] For example: Although the entities "laser cladding" and "electron beam deposition" use different energy sources, their control logics and module architectures are similar.

[0209] Functional analogy edge

[0210] For example: "Laser focusing control" and "current density control" can form an equivalent analogy in the energy field regulation logic.

[0211] Evolutionary analogy edge

[0212] For example: "Titanium alloy powder bed printing" evolves into "Titanium aluminum alloy multi - laser printing" → mapping out that "High - entropy alloy powder bed melting" may become a trend evolution node.

[0213] The graph spectrum is denoted as:

[0214] ,

[0215] where includes the above three types of edge relations and their edge attributes (similarity score, edge type, edge direction, etc.).

[0216] Step S403: Analogy graph spectrum embedding and graph neural network modeling

[0217] The system encodes the analogy graph spectrum using a graph neural network. The representation vector of each entity node is updated through the following iterative formula:

[0218] ,

[0219] where:

[0220] represents the set of adjacent entities of entity ;

[0221] Represents the relationship weight corresponding to the edge type 𝑟;

[0222] Represents the activation function;

[0223] Represents the normalization factor;

[0224] Represents the initial vector obtained by projecting the entity attribute tensor.

[0225] After the graph embedding learning is completed, the system obtains the context representation of each entity in the analogy graph, which can be used to calculate the analogy score.

[0226] Step S404: Analogy score calculation and evolution trend reasoning

[0227] For any two entities 、 Its analogy score function is defined as:

[0228] ,

[0229] The commonly used similarity calculation method is cosine similarity:

[0230] ,

[0231] The system sorts all potential relationships according to the analogy score and filters out entity pairs with scores higher than a set threshold (such as 0.85) as candidate paths with potential technological evolution possibilities.

[0232] Exemplarily:

[0233] Structural analogy reasoning

[0234] Entity 1: Laser cladding printing system (including optical module, scanning path control, material injection system)

[0235] Entity 2: Plasma spraying deposition (including high-temperature energy source, path control, powder feeding)

[0236] The system extracts and matches the structural modules of the two, determines it as a structural analogy edge, and assigns an analogy score of 0.82.

[0237] → Reasoning conclusion: If the plasma dual-energy field regulation technology is introduced in laser cladding, it is possible to form a new "hybrid energy field deposition path control technology".

[0238] Functional analogy reasoning

[0239] Entity 1: Multi-region scanning path adjustment

[0240] Entity 2: Multi-point pressure sensing feedback mechanism

[0241] Although the two technologies are different, one is printing path control and the other is control closed loop, they share the same functional semantics of "regional-level adaptive regulation".

[0242] → Score 0.89, system prompt: The concept of "multi-area sensing-path integrated linkage printing system" may appear.

[0243] Evolutionary analogical reasoning

[0244] The entity evolution path is known:

[0245] "Single laser titanium alloy printing" → "Multi-laser titanium alloy printing" → "Synchronous multi-laser printing of high entropy alloys"

[0246] → Systematic analogy: "Multi-material micro-powder composite printing" has a multi-component structure in terms of materials, and may refer to the evolution path of "high entropy alloys".

[0247] The system generates a future candidate term: "high-component particle multi-laser partitioning forming".

[0248] This step constructs an analogy relationship graph that includes similarities in structure, function, and evolutionary path, and introduces graph neural networks for entity representation learning and analogy score calculation, thereby achieving modeling of deep semantic associations between industry entities and reasoning about evolutionary trends. It can effectively identify potential technology migration paths and integration directions, breaking through the limitations of traditional co-occurrence statistics or rule matching methods, and significantly improving the information collection system's forward-looking judgment and cross-domain association capabilities in complex technical fields, providing more intelligent and semantic support for new terminology discovery, technology forecasting, and strategic analysis.

[0249] In step S50, the generated new terms and the inferred potential technology evolution direction are integrated with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept reasoning edges.

[0250] This step aims to unify and integrate the semantically emergent terms generated by step S30 and the potential technology evolution direction inferred by step S40 with the industry terms, technical entities, material entities, relationship pairs and other information explicitly extracted from multi-source data in the previous step, to build a dynamic semantic knowledge graph with incremental update capabilities, clear structural semantics, and support for evolutionary reasoning, which is used to support subsequent industry analysis, trend monitoring and strategic decision-making.

[0251] The specific implementation process includes:

[0252] Step S501: Initial structure preparation

[0253] The system first loads or constructs the original industrial semantic knowledge graph , where:

[0254] represents the set of original entities, including technical terms, material types, equipment structures, control strategies, application fields, etc.;

[0255] represents the set of original relationships, including common domain semantic relationships such as "applicable to", "contains", "based on", "belongs to", "improved from", "replaces", etc.

[0256] The original graph usually comes from the previous information extraction module (step S20) or existing standard knowledge bases (such as CMeKG, patent co-occurrence graph, etc.).

[0257] Step S502: Access of new term nodes and edge construction

[0258] For the new terms generated and verified in step S30 , the system accesses them one by one into the graph and constructs the following information:

[0259] Create a new node for each term , and attach the following attributes:

[0260] Term source type (such as "combination", "transformation", "expansion");

[0261] Original term combination path;

[0262] First generation timestamp;

[0263] Semantic cluster label to which it belongs;

[0264] External verification result (whether it has appeared in the literature or database);

[0265] The system automatically matches the semantic proximity relationship between the new term and the existing nodes, constructs relationship edges such as "evolved from", "belongs to", "function enhanced", "term extended from", etc., and writes them into .

[0266] Exemplarily:

[0267] New term: "Multi-region energy density adaptive melting strategy"

[0268] The system determines that it is derived from the combination of "energy density control" and "multi-region scanning" → establish the following edge:

[0269] "Multi-region energy density adaptive melting strategy" → "energy density control": relationship = "combination derivative"

[0270] "Multi-region energy density adaptive melting strategy" → "Multi-region scanning": Relationship = "Semantic fusion"

[0271] Step S503: Evolution path edge fusion

[0272] For the potential evolution paths inferred in step S40 , the system processes them according to the following rules:

[0273] If the target node is a new term, directly connect the "evolution trend edge";

[0274] If does not exist yet, the system can temporarily set a "potential term placeholder node" with a "to be confirmed" flag for future verification;

[0275] If the confidence level of the inference relationship is relatively high (e.g., score > 0.9), it is labeled as "strong evolution connection"; if it is between 0.7 and 0.9, it is "possible trend connection".

[0276] The system embeds the evolution inference results and the term generation path into the knowledge graph through this mechanism to form a complete semantic link from concept generation → analogical reasoning → trend evolution → structural organization.

[0277] Step S504: Semantic graph structure update and evolution management

[0278] After the fusion is completed, the system constructs the latest state of the graph:

[0279] ,

[0280] At the same time, record the graph snapshot difference , for supporting subsequent evolution path backtracking and trend evaluation. Each node or edge is added with time attributes, update frequency marks, and verification status (such as "original term", "graph model prediction", "externally confirmed").

[0281] To control the complexity of the graph structure, the system can set the following mechanisms:

[0282] Edge relationship sparsification threshold: Delete low-correlation weak edges;

[0283] Node activity assessment: If a term has not appeared in subsequent data for a long time, set a "cold node" mark;

[0284] Term merging mechanism: For new and old terms with highly similar semantics, the system avoids graph redundancy by merging nodes or setting an "alias" relationship.

[0285] Exemplarily:

[0286] New term:

[0287] "Ceramic Composite Microwave Synchronous Sintering"

[0288] "High-Frequency Multimode Energy Adaptive Printing"

[0289] Inference Edge:

[0290] "Laser Sintering" → "Multimode Energy Deposition": Evolutionary Edge (Score 0.91)

[0291] "Electron Beam Printing" → "Microwave Printing": Structural Analogy Edge (Score 0.88)

[0292] Create a new node: "Ceramic Composite Microwave Synchronous Sintering"

[0293] Attributes: "Combination + Extended Generation"; Semantic Source Terms: "Ceramic Composite Powder" + "Microwave Energy Synchronous Control"

[0294] Edge Update:

[0295] "Ceramic Composite Microwave Synchronous Sintering" → "Microwave Energy Synchronous Control": Edge Type = "Term Extended From"

[0296] "Electron Beam Printing" → "Microwave Printing": Edge Type = "Analogy Migration Path"

[0297] "Microwave Printing" → "Ceramic Composite Microwave Synchronous Sintering": Edge Type = "Potential Evolutionary Edge"; Score = 0.84; Status = "To Be Confirmed"

[0298] In step S50, the system realizes the expansion of the industrial knowledge graph system from explicit term knowledge to emerging concepts, from co-occurrence graphs to analogy graphs, and from static structures to dynamic evolution, with the following technical advantages: new terms and trend edges can be naturally embedded in the original knowledge graph; it is convenient to display the path of "term birth - development - integration"; it can accommodate different source information from data, language models, and reasoning models; it provides structural support for trend prediction, hot spot discovery, and technology evolution modeling.

[0299] In step S60, the dynamic semantic knowledge graph is periodically updated and reasoning calculations are performed to output structured industrial intelligence information, which includes new terms, high-potential technology paths, cross-domain technology integration trends, and concept evolution trajectories.

[0300] The purpose of this step is to, based on the dynamic semantic knowledge graph constructed in step S50, extract and output structured industrial intelligence information with strategic value, forward-lookingness, and interpretability by setting a periodic update mechanism and a reasoning calculation model, thereby realizing the closed-loop logic chain of "information → knowledge → trend → decision".

[0301] The industry intelligence information not only includes the statistical results of explicit data, but also includes the potential knowledge deduced by the artificial intelligence model based on the graph structure, such as the propagation path of new terms in different fields, the possible direction of technology evolution, the phased law of the evolution of industry concepts, etc.

[0302] The specific implementation steps include:

[0303] Step S601: Graph update scheduling mechanism

[0304] The system sets a regular trigger mechanism, such as once a day, once a week or once a month, to update and iterate the knowledge graph. The updated content includes:

[0305] Integrate new term nodes and relationship edges;

[0306] Delete, downgrade or freeze low-active nodes and edges (such as no citation for 6 months);

[0307] Update node attributes (such as the first appearance time, current activity, appearance frequency, etc.);

[0308] Generate a snapshot of the graph version to form a time-series graph structure 。

[0309] The graph update is based on the graph incremental calculation strategy, and preferentially processes the newly added / modified content to reduce the calculation complexity.

[0310] Step S602: Inference calculation module

[0311] For the knowledge graph that has been updated and completed , a variety of graph inference algorithms are used for deductive analysis, including:

[0312] Path inference

[0313] Identify the high-confidence paths between terms - technologies - materials - products, such as:

[0314] "High-entropy alloy" → "Multi-laser sintering" → "Aero-engine nozzle"

[0315] → Output the possible technology implementation routes;

[0316] Graph representation learning

[0317] Use graph neural networks (such as GAT, RGCN) to learn the vector representations of entities in the graph, and then calculate:

[0318] The semantic change trend between terms;

[0319] The evolution trajectory of entity vectors over time;

[0320] Hot spot clustering and semantic diffusion analysis

[0321] Judge the diffusion degree of new terms in the graph, whether they can quickly form multilateral connections and participate in multiple trend paths, as an indicator for judging "concept popularity".

[0322] Multi-hop logical chain tracking

[0323] Identify the multi-hop logical chain of concept A → technology B → material C → application D for complex trend modeling.

[0324] Step S603: Industry intelligence information output module

[0325] After the inference calculation is completed, the system automatically organizes the results into a structured intelligence report, and the output information includes but is not limited to:

[0326] New term intelligence report

[0327] New term name

[0328] Source path (combined term / analogical reasoning)

[0329] Current graph activity index

[0330] Whether it has been verified in the external database

[0331] List of high-potential technology paths

[0332] Path starting point (usually basic technology or material)

[0333] Intermediate control strategy node

[0334] Potential output structure or application scenario

[0335] Path confidence score and recommendation level

[0336] Cross-domain technology fusion trend graph

[0337] In the last two rounds of graphs, the significant growth path of the connection relationship between terms in different fields

[0338] Form cross-border fusion nodes (such as "microwave-assisted metal sintering" connecting "electronic communication modulation technology" and "metallurgical laser forming")

[0339] Concept evolution trajectory graph

[0340] Describe the node expansion trajectory, relationship density change and evolution path of typical concepts (such as "energy density control") in multiple time graphs

[0341] Visualize as a time series flow graph or a vein tree graph structure

[0342] The output format can be PDF data, or generate "Intelligent Industry Intelligence Weekly Report / Monthly Report" in the form of charts + text summaries.

[0343] Exemplarily, as shown in Table 1:

[0344] Table 1 New Term Information Output

[0345] New term Source path Activity Whether to verify Multi-channel laser scanning path control "Multi-laser system" + "Path adjustment strategy" High No Microwave-assisted powder bed sintering "Powder bed melting" + Analogy to "Microwave heating" Medium Yes (appeared in arXiv)

[0346] High-potential Technology Path

[0347] Starting point: "High-entropy alloy powder"

[0348] Path: "High-entropy alloy powder" → "Zone temperature-controlled sintering" → "Interlayer microstructure regulation" → "Directional printing of high-temperature fatigue parts"

[0349] Path confidence: 0.89

[0350] Application potential: Aero-engine support frame

[0351] Cross-field integration trend

[0352] "Laser galvanometer synchronous scanning" ↔ "Optical communication beam splitting regulation"

[0353] → The system identifies it as an analog fusion point of "Multi-channel optical modulation scheme", which may give birth to a new field of "Frequency-tunable laser sintering".

[0354] Through this step, the system transforms the static term map into a dynamic industry knowledge system that can be updated, deduced, and output. Its core technical advantages include:

[0355] The periodic update mechanism ensures the real-time nature and trend perception ability of the industry map; the structured intelligence output makes the system not only a data collector, but also a reasoning analyzer with "expert-like thinking"; it supports multiple types of analysis requirements, such as new term warning, technology path discovery, cross-field integration identification, and concept cycle management; it significantly improves the semantic understanding and strategic judgment ability of the artificial intelligence system for complex technology industries

[0356] See Figure 2 , in another embodiment, the present invention also provides an additive manufacturing industry information collection system based on artificial intelligence, including:

[0357] A data acquisition module for obtaining original data related to the additive manufacturing industry from multi-source heterogeneous data channels according to a preset industry theme;

[0358] A semantic encoding module for preprocessing the original data and encoding the original data using a deep language representation model to generate a high-dimensional semantic vector representation;

[0359] A term generation module, which is used to construct a term space based on the high-dimensional semantic vectors, and combine and / or transform and / or expand the terms or technical descriptions in the semantic neighborhood through a concept combination function to generate new terms that do not directly appear in the original data but can be semantically deduced;

[0360] An analogical reasoning module, which is used to construct an analogical relationship graph with similarities in structure, function or evolutionary path based on the identified industry entities, and perform representation learning on the entities in the graph through a graph neural network model to obtain the analogical score relationship between entities for inferring potential technological evolution directions;

[0361] A graph construction module, which is used to fuse the generated new terms and the inferred potential technological evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges;

[0362] An intelligence output module, which is used to perform periodic updates and inference calculations on the dynamic semantic knowledge graph and output structured industry intelligence information, where the industry intelligence information includes new terms, high-potential technology paths, cross-domain technology fusion trends, and concept evolution trajectories.

[0363] In a further implementation manner, the term generation module includes:

[0364] A concept combination function library, which is used to define a term combination function, a term transformation function, and a term expansion function;

[0365] A combination control unit, which is used to select term pairs in the semantic neighborhood in the term space, call the concept combination function to generate term semantic vectors, and input the generated term semantic vectors into a term decoder to generate a new term text that can be read in language;

[0366] A term scoring unit, which is used to perform semantic consistency scoring and knowledge matching verification on the generated new terms, and mark the terms with a scoring result higher than a preset threshold as semantically emergent terms.

[0367] In a further implementation manner, the analogical reasoning module includes:

[0368] An entity attribute construction unit, which is used to construct an industry entity set and its attribute tensor;

[0369] An analogical relationship construction unit, which is used to construct structural analogical edges, functional analogical edges, and evolutionary analogical edges according to features such as control logic, module composition, and process constraints;

[0370] A graph neural network learning unit, which is used to update the vectors of entity nodes through multi-round aggregation of adjacent node information in the graph structure;

[0371] An analog scoring unit, which is used to calculate the semantic similarity based on node vectors, output the analog scores between entities, and is used for subsequent evolutionary path reasoning.

[0372] In a further implementation manner, the graph construction module further includes:

[0373] A node access unit, which is used to access new terms into the original industry graph in the form of nodes, attach term sources, generation paths, first appearance times, and semantic cluster labels;

[0374] A relationship generation unit, which is used to establish term extension edges, combined derivative edges, and function enhancement edges in the graph based on term generation paths and semantic similarity;

[0375] An inference edge access unit, which is used to add potential technology evolutionary paths to the graph in the form of analog edges or trend edges, set unvalidated term nodes as placeholder nodes and set the "to be confirmed" status;

[0376] An evolution management unit, which is used to record the graph snapshot difference and maintain node update times, active statuses, and evolution tracking information.

[0377] In a further implementation manner, the intelligence output module includes:

[0378] An update scheduling unit, which is used to periodically trigger graph incremental update operations, update node attributes, edge weights, and graph version snapshots;

[0379] An inference analysis unit, which is used to perform path reasoning, graph representation learning, hot term clustering, and multi-hop logical chain analysis on the updated knowledge graph;

[0380] An intelligence organization unit, which is used to extract new terms, high-potential technology paths, cross-domain trend nodes, and evolution trajectory information, and organize them into a structured intelligence report;

[0381] An output interface unit, which is used to export the industry intelligence information in the form of structured data or a graphic report.

[0382] It should be noted that the above explanation of the embodiments of the additive manufacturing industry information collection method based on artificial intelligence also applies to the devices in the embodiments of the present application, and will not be repeated here.

[0383] Those of ordinary skill in the art can realize that the units and algorithm steps described in the embodiments disclosed herein can be implemented by a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0384] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0385] In several embodiments provided in the present application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (hereinafter referred to as ROM), random access memory (hereinafter referred to as RAM), magnetic disks, or optical discs that can store program codes.

[0386] The above is only the specific implementation manner of the present application. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application. For the part of the module structure not specifically defined in the present invention, the content recorded in the prior art shall prevail. The prior art mentioned in the foregoing background art part and specific embodiment part of the present invention can be used as a part of the present invention to understand the meaning of some technical features or parameters.

Claims

1. An information collection method for the additive manufacturing industry based on artificial intelligence, characterized in that, The method includes the following steps: Obtain the original data related to the additive manufacturing industry from multi-source heterogeneous data channels according to a preset industry theme; Preprocess the original data, and encode the original data using a deep language representation model to generate a high-dimensional semantic vector representation; Construct a term space based on the high-dimensional semantic vector, and combine and / or transform and / or expand the terms or technical descriptions in the semantic neighborhood through a concept combination function to generate new terms that do not directly appear in the original data but are semantically derivable; Based on the identified industry entities, construct an analogy relationship graph with structural, functional, or evolutionary path similarities, and perform representation learning on the entities in the graph through a graph neural network model to obtain the analogy score relationship between entities for inferring potential technology evolution directions; Fuse the generated new terms and the inferred potential technology evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges; Periodically update and perform inference calculations on the dynamic semantic knowledge graph, and output structured industry intelligence information, where the industry intelligence information includes new terms, technology paths, cross-domain technology integration trends, and concept evolution trajectories; 2. The method for collecting information in the additive manufacturing industry based on artificial intelligence according to claim 1, wherein The step of generating new terms that do not directly appear in the original data through a concept combination function includes: Based on the term space, define multiple types of concept combination functions, and the combination functions include: a term combination function, a term transformation function, and a term expansion function; Among them, the term combination function is used to perform structural-level combination of two or more semantically related or complementary terms to form new terms expressing more complex or more refined concepts; the term transformation function is used to perform semantic substitution, expression transformation, or grammatical rewriting on existing terms to form variant terms that are semantically connected or evolved; the term expansion function is used to add context modification components, application conditions, or technical constraints to terms to form semantically enhanced terms; 3. The method for collecting information in the additive manufacturing industry based on artificial intelligence according to claim 1, wherein The step of constructing an analogy relationship graph with structural, functional, or evolutionary path similarities and performing representation learning on the entities in the graph through a graph neural network model includes: Construct an industry entity set and an entity attribute tensor, and construct structural analogy edges, functional analogy edges, and evolutionary analogy edges based on the structural similarities, functional relevance, or evolutionary chains among entities in terms of control logic, module composition, and process constraints; Based on the constructed graph structure, use a graph neural network to update the node vectors in multiple rounds, and aggregate the adjacent node information based on the edge type weights in each round of iteration; Calculate the semantic similarity score between nodes for determining the analogy strength and potential evolution direction between technical entities; 4. The information collection method for the additive manufacturing industry based on artificial intelligence according to claim 1, characterized in that The step of constructing a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges includes: [[ID=១៥]]Connect the generated new terms to the original industry knowledge graph, construct new term nodes and attach term source type, generation path, first appearance time, semantic cluster label attribute information; Establish semantic relationship edges such as term expansion edges, combination derivation edges, and function enhancement edges according to the term generation path and semantic similarity; Integrate the inferred potential technology evolution paths into the knowledge graph in the form of analogy edges or trend edges. For terms that have not been verified, placeholder nodes can be set and the "to be confirmed" status can be set. Record the snapshot differences of the knowledge graph and mark the node update time and update frequency for subsequent tracking of the knowledge graph evolution path and structure maintenance.

5. The information collection method for the additive manufacturing industry based on artificial intelligence according to claim 1, characterized in that The steps of outputting structured industry intelligence information include: Set up a knowledge graph update scheduling mechanism to perform incremental updates on the dynamic semantic knowledge graph regularly, including node attribute updates, edge weight adjustments, node active status evaluation, and version snapshot generation. Perform path reasoning, graph representation learning, hot spot clustering analysis, and multi-hop logical chain tracking on the updated knowledge graph to obtain the semantic change trends of terms, integrated nodes, and technology paths. Organize the analysis results into structured intelligence output. The output content includes: new term intelligence reports, technology path lists, cross-domain integration trend graphs, and concept evolution trajectory graphs, and supports export in the form of structured data files or a combination of text and graphics.

6. An information collection system for the additive manufacturing industry based on artificial intelligence, characterized in that, The system includes the following modules: A data collection module for obtaining raw data related to the additive manufacturing industry from multi-source heterogeneous data channels according to a preset industry theme. A semantic encoding module for preprocessing the raw data and encoding the raw data using a deep language representation model to generate a high-dimensional semantic vector representation. A term generation module for constructing a term space based on the high-dimensional semantic vectors, and generating new terms that do not directly appear in the raw data but can be semantically deduced by combining and / or transforming and / or expanding terms or technical descriptions in the semantic neighborhood through a concept combination function. An analogical reasoning module for constructing an analogical relationship graph with structural, functional, or evolutionary path similarities based on the identified industry entities, and performing representation learning on the entities in the graph through a graph neural network model to obtain the analogical score relationship between entities for inferring potential technology evolution directions. A knowledge graph construction module for integrating the generated new terms and the inferred potential technology evolution directions with the originally extracted industry knowledge entities and relationships to construct a dynamic semantic knowledge graph containing explicit entity relationships and new concept inference edges. An intelligence output module for performing periodic updates and inference calculations on the dynamic semantic knowledge graph and outputting structured industry intelligence information, where the industry intelligence information includes new terms, technology paths, cross-domain technology integration trends, and concept evolution trajectories.

7. The information collection system for the additive manufacturing industry based on artificial intelligence according to claim 6, wherein The term generation module includes: A concept combination function library for defining term combination functions, term transformation functions, and term extension functions. A combination control unit for selecting term pairs in the semantic neighborhood in the term space, calling the concept combination function to generate term semantic vectors, and inputting the generated term semantic vectors into a term decoder to generate a new term text that can be read in language. A term scoring unit for performing semantic consistency scoring and knowledge matching verification on the generated new terms, and marking the terms with a scoring result higher than a preset threshold as semantic emergence terms.

8. The information collection system for the additive manufacturing industry based on artificial intelligence according to claim 6, wherein The analogical reasoning module includes: An entity attribute construction unit for constructing a set of industry entities and their attribute tensors. The analogical relationship construction unit is used to construct structural analogical edges, functional analogical edges, and evolutionary analogical edges based on control logic, module composition, and process constraint features; The graph neural network learning unit is used to update the vectors of entity nodes through multi-round aggregation of adjacent node information in the graph structure; The analogical scoring unit is used to calculate the semantic similarity based on node vectors and output the analogical scores between entities for subsequent evolutionary path reasoning.

9. The information collection system for the additive manufacturing industry based on artificial intelligence according to claim 6, wherein The graph construction module further includes: The node access unit is used to access new terms into the original industry graph in the form of nodes, attaching term sources, generation paths, first appearance times, and semantic cluster labels; The relationship generation unit is used to establish term extension edges, combination derivative edges, and functional enhancement edges in the graph based on term generation paths and semantic similarity; The inference edge access unit is used to add potential technology evolution paths to the graph in the form of analogical edges or trend edges, setting unvalidated term nodes as placeholder nodes and setting the "to be confirmed" status; The evolution management unit is used to record the graph snapshot differences and maintain node update times, active statuses, and evolution tracking information.

10. The information collection system for the additive manufacturing industry based on artificial intelligence according to claim 6, characterized in that, The intelligence output module includes: The update scheduling unit is used to periodically trigger graph incremental update operations to update node attributes, edge weights, and graph version snapshots; The inference analysis unit is used to perform path reasoning, graph representation learning, hot term clustering, and multi-hop logic chain analysis on the updated knowledge graph; The intelligence organization unit is used to extract new terms, technology paths, cross-domain trend nodes, and evolution trajectory information and organize them into a structured intelligence report; The output interface unit is used to export the industry intelligence information in the form of structured data or graphic reports.

Images