Enterprise-oriented Industrial Chain Intelligent Construction and Analysis System

Abstract

The present invention relates to the technical field of enterprise industrial chain analysis, and specifically discloses an intelligent construction and analysis system for an enterprise industrial chain, including: an industrial chain data collection module, used to collect all upstream, midstream and downstream products of the current enterprise's industrial chain and corresponding product-related company information; an industrial chain knowledge graph construction module, used to perform reference resolution and deduplication on all upstream, midstream and downstream products of the current enterprise's industrial chain and corresponding product-related company information, and then construct an industrial chain knowledge graph to obtain the current enterprise's industrial chain knowledge graph; an industrial chain intelligent analysis module, used to perform intelligent analysis on the current enterprise's industrial chain knowledge graph to obtain a variety of key performance indicators; so that the industrial chain can reflect the development status in real time and accurately, help enterprise managers understand the overall picture of industrial chain management and operation, and provide data-based decision support.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise industrial chain analysis, and in particular to an enterprise-oriented intelligent industrial chain construction and analysis system. Background Art

[0002] With the rapid development of enterprises, traditional industrial chain analysis methods often rely on manual data collection and empirical judgment, lack of systematicity and cannot meet the needs of complex and extensive markets. In modern enterprise management, industrial chain analysis is an important means for enterprises to gain competitive advantages. Enterprises are in urgent need of an intelligent and automated industrial chain construction and analysis system to improve decision-making efficiency, optimize resource allocation, and achieve sustainable development.

[0003] Therefore, the present invention proposes an intelligent construction and analysis system for the industrial chain of enterprises. Summary of the invention

[0004] The present invention provides an intelligent construction and analysis system for the industrial chain of enterprises, so that the generated industrial chain can cover all upstream, midstream and downstream products and corresponding product-related company information, can reflect the development status in real time and accurately, help enterprise managers understand the overall picture of industrial chain management and operation, and provide data-based decision support. Through intelligent technology, in-depth analysis and management of the enterprise industrial chain can be achieved, helping enterprises optimize resource allocation, improve operational efficiency and market competitiveness.

[0005] The present invention provides an enterprise-oriented intelligent industrial chain construction and analysis system, comprising:

[0006] The industrial chain data collection module is used to collect all upstream, midstream and downstream products of the current enterprise's industrial chain and the corresponding product-related company information;

[0007] The industrial chain knowledge graph construction module is used to construct the industrial chain knowledge graph after performing reference resolution and deduplication on all upstream, midstream and downstream products and corresponding product-related company information of the current enterprise's industrial chain, so as to obtain the industrial chain knowledge graph of the current enterprise;

[0008] The industrial chain intelligent analysis module is used to perform intelligent analysis on the current enterprise's industrial chain knowledge graph and obtain a variety of key performance indicators.

[0009] Preferably, the industrial chain data collection module includes:

[0010] The product information acquisition submodule is used to obtain all upstream, midstream and downstream products in the current enterprise's industrial chain;

[0011] The company information acquisition submodule is used to obtain product-related company information corresponding to all upstream, midstream and downstream products in the current enterprise's industrial chain based on a web crawler.

[0012] Preferably, the industrial chain knowledge graph construction module includes:

[0013] The first coreference resolution and deduplication execution sub-module is used to perform coreference resolution and deduplication on all upstream, midstream and downstream products of the industrial chain to obtain all upstream, midstream and downstream corrected products of the industrial chain;

[0014] The second coreference resolution and deduplication execution sub-module is used to perform coreference resolution and deduplication on the product-related company information corresponding to all upstream, midstream and downstream corrected products of the industrial chain of the current enterprise to obtain the corrected product-related company information corresponding to all upstream, midstream and downstream corrected products of the industrial chain;

[0015] The knowledge graph construction sub-module is used to construct the industrial chain knowledge graph of the current enterprise based on all upstream, midstream and downstream concise products of the industrial chain and the corresponding concise product-related company information.

[0016] Preferably, the first coreference resolution and deduplication execution sub-module includes:

[0017] The product definition attribute information acquisition unit is used to acquire the product definition attribute information set of each product among all upstream, midstream and downstream products of the industrial chain;

[0018] The ambiguous information item acquisition unit is used to determine the set of different meanings of each ambiguous information item in each product definition attribute information set;

[0019] The coreference resolution processing unit is used to perform coreference resolution processing on the set of different meanings of each ambiguous information item in each product definition attribute information set to obtain the clear product definition attribute information set of each product;

[0020] The similarity calculation unit is used to calculate the similarity between the clear product definition attribute information sets of every two products among all upstream, midstream and downstream products of the industrial chain;

[0021] The overlapping product identification unit is used to regard two products with a similarity between product definition attribute information sets not less than the similarity threshold among all upstream, midstream and downstream products of the industrial chain as an overlapping product group;

[0022] The product deduplication execution unit is used to perform merging processing on all overlapping product groups among all upstream, midstream and downstream products of the industrial chain to obtain all upstream, midstream and downstream corrected products of the industrial chain.

[0023] Preferably, the ambiguous information item acquisition unit includes:

[0024] The meaning similarity calculation sub-unit is used to retrieve the set of meanings of each definition attribute information in each product definition attribute information set in the meaning library and calculate the meaning similarity between every two meanings in the set of meanings of each definition attribute information;

[0025] The ambiguous meaning set identification subunit is used to screen out, from each product definition attribute information set, the definition attribute information of at least one group of meanings whose meaning similarity in the corresponding meaning set does not exceed the meaning similarity threshold as ambiguous information items, and regard each group of meanings whose corresponding meaning similarity does not exceed the meaning similarity threshold as the corresponding ambiguous meaning set of the ambiguous information items.

[0026] Preferably, the reference resolution processing unit includes:

[0027] The corpus acquisition subunit is used to acquire the original corpus of each ambiguous information item in each product definition attribute information set, and acquire the reference corpus of each meaning in the corresponding ambiguous meaning set of each ambiguous information item based on the web crawling method;

[0028] The valid word acquisition subunit is used to delete the stop words in the original corpus of each ambiguous information item to obtain all valid words of each ambiguous information item;

[0029] The word parameter statistics subunit is used to count the word frequency of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set. At the same time, it counts the total number of sentences containing each valid word of the corresponding ambiguous information item in the reference corpus of each meaning in each ambiguous meaning set;

[0030] The support rate calculation subunit is used to calculate the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set based on the total number of sentences contained in the reference corpus of each meaning in each ambiguous meaning set, the total number of sentences containing each valid word of the corresponding ambiguous information item in the corresponding reference corpus, and the word frequency of the corresponding valid word in the reference corpus of the corresponding meaning;

[0031] The reference resolution subunit is used to obtain the clear product definition attribute information set of each product based on the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set.

[0032] Preferably, the reference resolution subunit includes:

[0033] The comprehensive support rate calculation end is used to regard the sum of the support rates of all valid words of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set as the comprehensive support rate of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set;

[0034] The clear meaning determination end is used to regard the meaning corresponding to the maximum comprehensive support rate of each ambiguous information item in the corresponding ambiguous meaning set as the clear meaning of the corresponding ambiguous information item;

[0035] The ambiguous item annotation end is used to supplement and annotate the ambiguous information items in the product definition attribute information set of each product based on the clear meanings of all ambiguous information items, so as to obtain the clear product definition attribute information set of each product.

[0036] Preferably, the similarity calculation unit includes:

[0037] The first calculation subunit is used to calculate the meaning similarity between every two meanings in the meaning sets of every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream and downstream products belonging to the industrial chain;

[0038] The second calculation subunit is used to take the average value of the meaning similarities between all groups of meanings in the meaning sets of every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream and downstream products belonging to the industrial chain as the meaning similarity between every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream and downstream products belonging to the industrial chain;

[0039] The third calculation subunit is used to take the average value of the meaning similarities of all groups of definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream and downstream products belonging to the industrial chain as the similarity between the clear product definition attribute information sets of every two products among all the upstream, midstream and downstream products belonging to the industrial chain.

[0040] Preferably, the knowledge graph construction sub-module includes:

[0041] The product chain construction sub-module is used to concatenate all the simple products of the upstream, midstream and downstream of the industrial chain to obtain a product chain;

[0042] The knowledge graph construction sub-module is used to connect the simple product associated company information corresponding to each product in the product chain with the corresponding product in the product chain to construct the industrial chain knowledge graph of the current enterprise.

[0043] Preferably, the industrial chain intelligent analysis module includes:

[0044] The first analysis sub-module is used to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise based on the first preset analysis method to obtain the business performance evaluation value;

[0045] The second analysis sub-module is used to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise based on the second preset analysis method to obtain the industrial chain health degree;

[0046] Among them, the multiple key performance indicators include the business performance evaluation value and the industrial chain health degree.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows: the generated industrial chain can cover all upstream, midstream and downstream products and the corresponding product-related company information, can reflect the development status in real time and accurately, help enterprise managers understand the overall picture of industrial chain management and operation, and provide data-based decision support. Through intelligent technology, in-depth analysis and management of the enterprise industrial chain can be achieved, helping enterprises optimize resource allocation, improve operational efficiency and market competitiveness.

[0048] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0049] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0051] Figure 1 A schematic diagram of the internal functional modules of the enterprise-oriented industry chain intelligent construction and analysis system in an embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of the internal functional modules of the industrial chain data collection module in an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of the internal functional modules of the industrial chain knowledge graph construction module in an embodiment of the present invention;

[0054] Figure 4 Schematic diagram of the internal functional modules of the industrial chain intelligent analysis module in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0056] Example 1

[0057] The present invention provides an intelligent construction and analysis system for the industrial chain of enterprises, referring to Figure 1 ,include:

[0058] The industrial chain data collection module is used to collect all upstream, midstream and downstream products of the current enterprise's industrial chain and the corresponding product-related company information;

[0059] The industrial chain knowledge graph construction module is used to construct an industrial chain knowledge graph for the current enterprise by performing coreference resolution and deduplication on all upstream, midstream, and downstream products of the industrial chain of the current enterprise and the corresponding product-related company information;

[0060] The industrial chain intelligent analysis module is used to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise to obtain a variety of key performance indicators.

[0061] In this embodiment, the industrial chain of the current enterprise is the industrial chain with the products or services provided by the current enterprise to consumers as the links.

[0062] In this embodiment, the upstream, midstream, and downstream products are respectively:

[0063] Upstream products: The starting point of the entire industrial chain mainly includes basic industries and technology R & D links, which are responsible for delivering products or services to downstream links;

[0064] Midstream products: mainly refer to some intermediate industrial product or service links.

[0065] Downstream products: usually refer to products or services in the production and manufacturing links.

[0066] In this embodiment, the product-related company information is the information of the supplier company of the product or the information of the company associated with the product.

[0067] In this embodiment, coreference resolution is to determine the specific meaning of the information with unclear meaning in the information.

[0068] In this embodiment, the key performance indicators are the indicators that can be obtained from the industrial chain knowledge graph and are used to provide a basis for the decision-making of enterprise managers in the future.

[0069] In this embodiment, it further includes:

[0070] 1. Convenient industrial chain construction and editing tools: Provide industrial chain construction tools, including constructing the upstream, midstream, and downstream of the industrial chain, the enterprises, products, and their associated relationships in each industrial link.

[0071] 2. Quantitative assessment of industrial chain health: The system can quantitatively assess the health of the industrial chain, help enterprise managers understand the overall picture of industrial chain management and operation, and provide data-based decision support.

[0072] 3. Real-time and dynamic nature of industrial chain analysis: The system can collect and update industrial chain node data in a timely and dynamic manner, providing the ability for dynamic analysis of the industrial chain.

[0073] 4. Industrial Chain Comparison Tool: Intuitively compare the management capabilities, operating performance, product overlap, and supplier overlap of different industrial chains to optimize the industrial layout and identify the short - comings in the capabilities of potential partners.

[0074] The beneficial effects of the above - mentioned technology are as follows: The generated industrial chain can cover all upstream, mid - stream, and downstream products and the corresponding product - related company information, can reflect the current development status in real - time and accurately, help enterprise managers understand the overall picture of industrial chain management and operation, and provide data - based decision - making support. Through intelligent technology, in - depth analysis and management of the enterprise's industrial chain are realized, helping the enterprise optimize resource allocation, improve operation efficiency, and enhance market competitiveness.

[0075] Embodiment 2

[0076] Based on Embodiment 1, the industrial chain data acquisition module, with reference to Figure 2 , includes:

[0077] The product information acquisition sub - module is used to obtain all upstream, mid - stream, and downstream products of the current enterprise's industrial chain based on the data API interface;

[0078] The company information acquisition sub - module is used to obtain the product - related company information corresponding to all upstream, mid - stream, and downstream products of the current enterprise's industrial chain based on web crawling.

[0079] In this embodiment, obtaining the product - related company information corresponding to all upstream, mid - stream, and downstream products of the current enterprise's industrial chain based on web crawling means:

[0080] That is, using the existing web crawling methods (such as distributed crawling or parallel crawling) to obtain the product - related company information corresponding to all upstream, mid - stream, and downstream products of the current enterprise's industrial chain.

[0081] In this embodiment, if the relevant company is a listed company, automatically collect the relevant financial report data publicly disclosed by the company, including but not limited to operating income, gross profit, net profit, etc.

[0082] The beneficial effects of the above - mentioned technology are as follows: Complete the acquisition of all upstream, mid - stream, and downstream products of the current enterprise's industrial chain and the corresponding product - related company information.

[0083] Embodiment 3

[0084] Based on Embodiment 1, the industrial chain knowledge graph construction module, with reference to Figure 3 , includes:

[0085] The first coreference resolution and deduplication execution sub - module is used to perform coreference resolution and deduplication on all upstream, mid - stream, and downstream products of the industrial chain to obtain all upstream, mid - stream, and downstream corrected products of the industrial chain;

[0086] The second reference resolution and deduplication execution sub-module is used to perform reference resolution and deduplication on the product-related company information corresponding to all the upstream, midstream, and downstream calibration products of the industrial chain of the current enterprise, and obtain the calibrated product-related company information corresponding to all the upstream, midstream, and downstream calibration products of the industrial chain;

[0087] The knowledge graph construction sub-module is used to construct the industrial chain knowledge graph of the current enterprise based on all the upstream, midstream, and downstream concise products of the industrial chain and the corresponding concise product-related company information.

[0088] In this embodiment, a relevant product library is constructed. An industrial chain construction tool is provided, and the industrial chain in the form of a knowledge graph can be independently constructed through a human-computer interaction method, including the relationship between the industrial chain and products, the upstream, midstream, and downstream relationships between products, the relationship between products and companies, and so on. The constructed industrial chain is stored in the graph database in the data structure of an attribute graph.

[0089] The beneficial effects of the above technology are as follows: By first performing reference resolution and deduplication on all the upstream, midstream, and downstream products of the industrial chain, and then performing reference resolution and deduplication on the product-related company information corresponding to all the upstream, midstream, and downstream calibration products of the industrial chain of the current enterprise obtained after reference resolution and deduplication, the meaning of the product-related company information corresponding to all the upstream, midstream, and downstream products of the industrial chain of the current enterprise is made clear and simplified.

[0090] Embodiment 4

[0091] On the basis of Embodiment 3, the first reference resolution and deduplication execution sub-module includes:

[0092] The product definition attribute information acquisition unit is used to acquire the product definition attribute information set of each product among all the upstream, midstream, and downstream products of the industrial chain;

[0093] The ambiguity information item acquisition unit is used to determine the set of different meanings of each ambiguity information item in each product definition attribute information set;

[0094] The reference resolution processing unit is used to perform reference resolution processing on the set of different meanings of each ambiguity information item in each product definition attribute information set to obtain the clear product definition attribute information set of each product;

[0095] The similarity calculation unit is used to calculate the similarity between the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products of the industrial chain;

[0096] The overlapping product identification unit is used to regard two products with a similarity not less than the similarity threshold between the product definition attribute information sets among all the upstream, midstream, and downstream products of the industrial chain as an overlapping product group;

[0097] The product duplicate removal execution unit is used to perform a merging process on all overlapping product groups in all upstream, midstream, and downstream products of the industrial chain to obtain all upstream, midstream, and downstream corrected products of the industrial chain.

[0098] In this embodiment, the product definition attribute information set of a product is a set formed by summarizing multiple attribute information used to define the specific content of the product obtained through web crawling. It is obtained from the original corpus of the product obtained through web crawling. For example, the product definition attribute information set of a semiconductor chip includes: integrated circuits, silicon wafers, microcircuit chips, etc.

[0099] In this embodiment, an ambiguous information item is an item whose meaning is not clear in the product definition attribute information set to which it belongs. For example, "apple" has two ambiguous meanings: the fruit "apple" and the electronics brand "apple".

[0100] In this embodiment, the set of divergent meanings of an ambiguous information item is a set formed by all the ambiguous meanings of the ambiguous information item.

[0101] In this embodiment, the clear product definition attribute information set of a product is a product definition attribute information set that does not contain any ambiguous information items.

[0102] In this embodiment, the similarity between the clear product definition attribute information sets of two products represents the degree of similarity in meaning between the clear product definition attribute information sets of the two products.

[0103] In this embodiment, the similarity threshold is a preset value that the similarity between the product definition attribute information sets of two products included in an overlapping product group must not be less than.

[0104] In this embodiment, performing a merging process on all overlapping product groups in all upstream, midstream, and downstream products of the industrial chain to obtain all upstream, midstream, and downstream corrected products of the industrial chain means: randomly selecting one of the two products in the overlapping product group and retaining it, and randomly selecting one of the product definition attribute information with a relatively higher similarity in meaning in the clear product definition attribute information sets of the two products in the overlapping product group and retaining it, while retaining all the product definition attribute information with a relatively lower similarity in meaning in the overlapping product group, so as to obtain the retained product and its corresponding clear product definition attribute information set.

[0105] The beneficial effects of the above technology are: By identifying and resolving the set of divergent meanings of each ambiguous information item in each product definition attribute information set and performing product duplicate removal processing, not only is the product definition attribute information of all upstream, midstream, and downstream products of the industrial chain clarified, but also the simplification and correction of all upstream, midstream, and downstream products of the industrial chain are achieved.

[0106] Embodiment 5

[0107] Based on Embodiment 4, the ambiguity information item acquisition unit includes:

[0108] A semantic similarity calculation subunit, configured to retrieve the meaning sets of each defined attribute information in each product defined attribute information set in the meaning library, and calculate the semantic similarity between every two meanings in the meaning set of each defined attribute information based on methods in existing technologies such as dictionaries (which can be understood as the semantic similarity between two words corresponding to the two meanings);

[0109] A discrepant meaning set identification subunit, configured to screen out, in each product defined attribute information set, the defined attribute information with at least one group of meanings in the corresponding meaning set whose semantic similarity does not exceed the semantic similarity threshold as an ambiguity information item, and regard each group of meanings whose semantic similarity does not exceed the semantic similarity threshold as the discrepant meaning set corresponding to the ambiguity information item.

[0110] In this embodiment, the meaning library is a pre-prepared database containing meaning sets of a large number of defined attribute information.

[0111] In this embodiment, the defined attribute information is the attribute information used to define the actual meaning of the corresponding product, and is mostly words.

[0112] In this embodiment, the meaning set of the defined attribute information contains an information set that can explain the true meaning of the corresponding defined attribute information.

[0113] In this embodiment, the semantic similarity threshold is a preset value that the semantic similarity corresponding to a group of meanings that must exist in the meaning set corresponding to the ambiguity information item cannot exceed.

[0114] The beneficial effects of the above technologies are as follows: The semantic similarity between every two meanings in the meaning set of each defined attribute information is accurately calculated, and the ambiguity information items and their discrepant meaning sets are accurately screened out in each product defined attribute information set.

[0115] Embodiment 6

[0116] Based on Embodiment 5, the anaphora resolution processing unit includes:

[0117] A corpus acquisition subunit, configured to acquire the original corpus of each ambiguity information item in each product defined attribute information set, and acquire the reference corpus of each meaning in the discrepant meaning set of each ambiguity information item based on the web crawling method;

[0118] An effective word acquisition subunit, configured to delete the stop words in the original corpus of each ambiguity information item to obtain all effective words of each ambiguity information item;

[0119] The word parameter statistics subunit is used to count the word frequencies of each valid word of each ambiguous information item in each meaning of the corresponding reference corpus in the corresponding ambiguous meaning set. At the same time, it counts the total number of sentences in the reference corpus of each meaning in each ambiguous meaning set that contain each valid word of the corresponding ambiguous information item;

[0120] The support rate calculation subunit is used to calculate the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set based on the total number of sentences contained in the reference corpus of each meaning in each ambiguous meaning set, the total number of sentences in the corresponding reference corpus that contain each valid word of the corresponding ambiguous information item, and the word frequency of the corresponding valid word in the reference corpus of the corresponding meaning;

[0121] The anaphora resolution subunit is used to obtain the clear product definition attribute information set of each product based on the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set.

[0122] In this embodiment, the original corpus of the ambiguous information item is the context or corpus fragment obtained when obtaining the corresponding product based on the data API interface or web crawling technology.

[0123] In this embodiment, obtaining the reference corpus of each meaning in the ambiguous meaning set of each ambiguous information item based on the web crawling method means: using the existing web crawling method (such as distributed crawler or parallel crawler) to obtain the retrieval result of the preset length of each meaning in the ambiguous meaning set of each ambiguous information item from the preset search website as the corresponding reference corpus.

[0124] In this embodiment, the stop words are the auxiliary words and prepositions contained in the sentence.

[0125] In this embodiment, based on the total number of sentences L contained in the reference corpus of each meaning in each ambiguous meaning set, the total number of sentences l in the corresponding reference corpus that contain each valid word of the corresponding ambiguous information item, and the word frequency P of the corresponding valid word in the reference corpus of the corresponding meaning, calculating the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set is:

[0126] Taking the product of the ratio L / l of the total number of sentences L contained in the reference corpus of each meaning in each ambiguous meaning set to the total number of sentences l in the corresponding reference corpus that contain each valid word of the corresponding ambiguous information item and the word frequency P of the corresponding valid word in the reference corpus of the corresponding meaning as the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set.

[0127] In this embodiment, the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding set of ambiguous meanings represents the importance or weight of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding set of ambiguous meanings.

[0128] The beneficial effects of the above technology are as follows: The accuracy of the reference resolution results of the two meanings included in the set of ambiguous meanings of each ambiguous information item in each product definition attribute information set is improved.

[0129] Embodiment 7

[0130] Based on Embodiment 6, the reference resolution sub-unit includes:

[0131] The comprehensive support rate calculation terminal is used to regard the sum of the support rates of all valid words of each ambiguous information item in the reference corpus of each meaning in the corresponding set of ambiguous meanings as the comprehensive support rate of each ambiguous information item in the reference corpus of each meaning in the corresponding set of ambiguous meanings;

[0132] The clear meaning determination terminal is used to regard the meaning corresponding to the maximum comprehensive support rate of each ambiguous information item in the corresponding set of ambiguous meanings as the clear meaning of the corresponding ambiguous information item;

[0133] The ambiguous item annotation terminal is used to supplement and annotate the ambiguous information items in the product definition attribute information set of each product based on the clear meanings of all ambiguous information items, and obtain the clear product definition attribute information set of each product.

[0134] In this embodiment, the comprehensive support rate of an ambiguous information item in the reference corpus of each meaning in the corresponding set of ambiguous meanings represents the degree to which the actual meaning of the ambiguous information item conforms to each meaning in the corresponding set of ambiguous meanings.

[0135] The beneficial effects of the above technology are as follows: The accuracy of the reference resolution results of the two meanings included in the set of ambiguous meanings of each ambiguous information item in each product definition attribute information set is further improved.

[0136] Embodiment 8

[0137] Based on Embodiment 5, the proximity calculation unit includes:

[0138] The first calculation sub-unit is used to calculate the meaning proximity (which can be understood as the semantic proximity between the two words corresponding to the two meanings) between every two meanings in the meaning sets of every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products belonging to the industrial chain based on the dictionary and the methods in the prior art;

[0139] A second calculation subunit, configured to use the average of the semantic similarity degrees between all groups of semantic meanings of every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products belonging to the industrial chain as the semantic similarity degree between every two definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products of the industrial chain;

[0140] A third calculation subunit, configured to use the average of the semantic similarity degrees of all groups of definition attribute information in the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products belonging to the industrial chain as the similarity degree between the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products of the industrial chain.

[0141] The beneficial effects of the above technology are as follows: The similarity degree between the clear product definition attribute information sets of every two products among all the upstream, midstream, and downstream products of the industrial chain is calculated reasonably and accurately.

[0142] Embodiment 9

[0143] Based on Embodiment 3, the knowledge graph construction sub-module includes:

[0144] A product chain construction sub-module, configured to concatenate all the simple products of the upstream, midstream, and downstream of the industrial chain to obtain a product chain;

[0145] A knowledge graph construction sub-module, configured to connect the simple product associated company information corresponding to each product in the product chain with the corresponding product in the product chain to construct an industrial chain knowledge graph of the current enterprise.

[0146] The beneficial effects of the above technology are as follows: The construction of the industrial chain knowledge graph of the current enterprise is completed, so that the generated industrial chain can cover all upstream, midstream, and downstream products and the corresponding product associated company information.

[0147] Embodiment 10:

[0148] Based on Embodiment 1, the industrial chain intelligent analysis module, referring to Figure 4 , includes:

[0149] A first analysis sub-module, configured to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise based on a first preset analysis method to obtain an operation performance evaluation value;

[0150] A second analysis sub-module, configured to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise based on a second preset analysis method to obtain the industrial chain health degree;

[0151] Among them, multiple key performance indicators include the operation performance evaluation value and the industrial chain health degree.

[0152] In this embodiment, the first preset analysis method is implemented by, for example, an operating performance evaluation model trained based on an artificial intelligence algorithm. This model can use a large number of pre-prepared industrial chain knowledge graphs of enterprises (used as model input quantities during the training process) and the operating performance evaluation values calculated manually based on the industrial chain knowledge graphs of the corresponding enterprises (used as model output quantities during the training process) as training samples for training.

[0153] In this embodiment, the second preset analysis method can be implemented by, for example, an industrial chain health assessment model trained based on an artificial intelligence algorithm. This model can use a large number of pre-prepared industrial chain knowledge graphs of enterprises (used as model input quantities during the training process) and the industrial chain health calculated manually based on the industrial chain knowledge graphs of the corresponding enterprises (used as model output quantities during the training process) as training samples for training.

[0154] In this embodiment, the operating performance evaluation value is an evaluation value representing the performance of an enterprise in terms of operating performance (in terms of revenue quality).

[0155] In this embodiment, the industrial chain health represents a numerical value indicating the health degree of an enterprise's industrial chain (such as the rationality of the industrial chain layout, etc.).

[0156] In this embodiment, based on the financial data collection results of companies related to the industrial chain, an automated analysis of the industrial chain is performed from bottom to top, including data statistics and calculations for a certain product link, the overall upstream / midstream / downstream, and the overall industrial chain on the industrial chain. Finally, relevant health scores (industrial chain management entropy scores) and some key performance indicator results are obtained, which can reflect the management vitality and operating efficiency of the industrial chain and its various links. 4. Industrial chain analysis and display module: Provides visual data charts, enabling users to easily view various analysis data of the industrial chain, including analysis data such as the management vitality and operating performance of the overall industrial chain, specific upstream / midstream / downstream, and specific product links. And provides an industrial chain comparison tool to intuitively compare the management capabilities, operating performance, product overlap, and supplier overlap of different industrial chains, so as to optimize the industrial layout and identify the ability shortfalls of partners. Also provides an industrial chain editing tool to facilitate users to create and edit the industrial chain, enabling the industrial chain to more real-time and accurately reflect the development status.

[0157] The beneficial effects of the above technology are: By performing intelligent analysis on the industrial chain knowledge graph of the current enterprise, various key performance indicators are obtained, providing a deeper decision-making basis for enterprise managers.

[0158] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent construction and analysis system for the industrial chain of enterprises, characterized by: include: The industrial chain data collection module is used to collect all upstream, midstream and downstream products of the current enterprise's industrial chain and the corresponding product-related company information; The industrial chain knowledge graph construction module is used to construct the industrial chain knowledge graph after performing reference resolution and deduplication on all upstream, midstream and downstream products and corresponding product-related company information of the current enterprise's industrial chain, so as to obtain the industrial chain knowledge graph of the current enterprise; The industry chain intelligent analysis module is used to intelligently analyze the current enterprise's industry chain knowledge graph to obtain a variety of key performance indicators; Among them, the industry chain knowledge graph construction module includes: The first reference resolution and deduplication execution submodule is used to perform reference resolution and deduplication on all upstream, midstream and downstream products in the industrial chain to obtain all upstream, midstream and downstream corrected products in the industrial chain; The second reference resolution and deduplication execution submodule is used to perform reference resolution and deduplication on the product-related company information corresponding to all upstream, midstream and downstream correction products in the current enterprise's industrial chain, and obtain the correction product-related company information corresponding to all upstream, midstream and downstream correction products in the industrial chain; The knowledge graph construction submodule is used to construct the current enterprise's industry chain knowledge graph based on all upstream, midstream and downstream correction products in the industry chain and the corresponding correction product-related company information; Among them, the first reference resolution and deduplication execution submodule includes: A product definition attribute information acquisition unit, used to acquire a product definition attribute information set of each product in all upstream, midstream and downstream products in the industrial chain; An ambiguous information item acquisition unit, used to determine an ambiguous meaning set of each ambiguous information item in each product definition attribute information set; A reference resolution processing unit, used for performing reference resolution processing on the ambiguous meaning set of each ambiguous information item in each product definition attribute information set to obtain a clear product definition attribute information set for each product; A similarity calculation unit is used to calculate the similarity between the clear product definition attribute information sets of every two products in all upstream, midstream and downstream products of the industrial chain; The overlapping product identification unit is used to treat two products whose product definition attribute information sets have a similarity not less than a similarity threshold among all upstream, midstream and downstream products in the industrial chain as overlapping product groups; The product deduplication execution unit is used to merge all overlapping product groups in all upstream, midstream and downstream products of the industrial chain to obtain all upstream, midstream and downstream corrected products of the industrial chain; The ambiguous information item acquisition unit includes: The meaning similarity calculation subunit is used to retrieve the meaning set of each definition attribute information in each product definition attribute information set in the meaning library, and calculate the meaning similarity between each two meanings in the meaning set of each definition attribute information; The ambiguous meaning set identification subunit is used to screen out, in each product definition attribute information set, at least one set of definition attribute information of meanings in the corresponding meaning set whose meaning similarity does not exceed the meaning similarity threshold as the ambiguous information item, and regard each set of meanings whose corresponding meaning similarity does not exceed the meaning similarity threshold as the ambiguous meaning set of the corresponding ambiguous information item; Wherein, refers to the digestion processing unit, including: A corpus acquisition subunit is used to acquire the original corpus of each ambiguous information item in each product definition attribute information set, and acquire the reference corpus of each meaning in the ambiguous meaning set of each ambiguous information item based on a web crawling method; The valid word acquisition subunit is used to delete the stop words in the original corpus of each ambiguous information item and obtain all the valid words of each ambiguous information item; The word parameter counting subunit is used to count the word frequency of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set, and at the same time, count the total number of sentences containing each valid word of the corresponding ambiguous information item in the reference corpus of each meaning in each ambiguous meaning set; a support rate calculation subunit, configured to calculate the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in each ambiguous meaning set based on the total number of sentences contained in the reference corpus of each meaning in each ambiguous meaning set, the total number of sentences containing each valid word of the corresponding ambiguous information item in the corresponding reference corpus, and the word frequency of the corresponding valid word in the reference corpus of the corresponding meaning; A reference resolution subunit, for obtaining a clear product definition attribute information set for each product based on the support rate of each valid word of each ambiguous information item in the reference corpus corresponding to each meaning in the ambiguous meaning set; The support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in each ambiguous meaning set is calculated based on the total number of sentences contained in the reference corpus of each meaning in each ambiguous meaning set, the total number of sentences containing each valid word of the corresponding ambiguous information item in the corresponding reference corpus, and the word frequency of the corresponding valid word in the reference corpus of the corresponding meaning, including: The product of the ratio of the total number of sentences contained in the reference corpus of each meaning in each ambiguous meaning set to the total number of sentences containing each valid word of the corresponding ambiguous information item in the corresponding reference corpus and the word frequency of the corresponding valid word in the reference corpus of the corresponding meaning is taken as the support rate of each valid word of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set.

2. The enterprise-oriented industry chain intelligent construction and analysis system according to claim 1 is characterized in that: Industrial chain data collection module, including: The product information acquisition submodule is used to obtain all upstream, midstream and downstream products in the current enterprise's industrial chain; The company information acquisition submodule is used to obtain product-related company information corresponding to all upstream, midstream and downstream products in the current enterprise's industrial chain based on a web crawler.

3. The enterprise-oriented industry chain intelligent construction and analysis system according to claim 1 is characterized in that: Refers to the digestion subunit, including: A comprehensive support rate calculation end, used to take the sum of the support rates of all valid words of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set as the comprehensive support rate of each ambiguous information item in the reference corpus of each meaning in the corresponding ambiguous meaning set; A clear meaning determination terminal is used to regard the meaning corresponding to the maximum comprehensive support rate of each ambiguous information item in the corresponding ambiguous meaning set as the clear meaning of the corresponding ambiguous information item; The ambiguous item annotation end is used to supplement the annotations of the ambiguous information items in the product definition attribute information set of each product based on the clear meanings of all the ambiguous information items, so as to obtain a clear product definition attribute information set for each product.

4. The enterprise-oriented industry chain intelligent construction and analysis system according to claim 1 is characterized in that: The proximity calculation unit includes: The first calculation subunit is used to calculate the similarity between the meanings of each two meanings in the meaning set of each two definition attribute information in the clear product definition attribute information set of each two products in all upstream, midstream and downstream products belonging to the industrial chain; The second calculation subunit is used to take the average of all group meanings in the meaning set of each two definition attribute information in the clear product definition attribute information set of each two products of all upstream, midstream and downstream products belonging to the industrial chain as the meaning similarity of each two definition attribute information in the clear product definition attribute information set of each two products of all upstream, midstream and downstream products of the industrial chain; The third calculation subunit is used to take the average of the similarity in meaning of all group definition attribute information in the clear product definition attribute information sets of every two products among all upstream, midstream and downstream products of the industrial chain as the similarity between the clear product definition attribute information sets of every two products among all upstream, midstream and downstream products of the industrial chain.

5. The enterprise-oriented industry chain intelligent construction and analysis system according to claim 1 is characterized in that: The knowledge graph construction submodule includes: The product chain construction submodule is used to connect all upstream, midstream and downstream correction products in the industrial chain in series to obtain a product chain; The knowledge graph construction submodule is used to connect the correction product-related company information corresponding to each product in the product chain with the corresponding product in the product chain to construct the current enterprise's industrial chain knowledge graph.

6. The enterprise-oriented industry chain intelligent construction and analysis system according to claim 1 is characterized in that: The industry chain intelligent analysis module includes: The first analysis submodule is used to perform intelligent analysis on the current enterprise's industrial chain knowledge graph based on a first preset analysis method to obtain an operating performance evaluation value; The second analysis submodule is used to perform intelligent analysis on the industrial chain knowledge graph of the current enterprise based on a second preset analysis method to obtain the health of the industrial chain; Among them, various key performance indicators include operating performance evaluation values ​​and the health of the industrial chain.

Images