Intelligent content navigation method and system based on power grid standard chapters

Abstract

The invention discloses an intelligent content navigation method and system based on power grid standard chapters, and relates to the technical field of data processing. Comprising the steps of performing content classification and tagging on power grid standard chapters; constructing an intelligent search engine, and quickly positioning related chapters according to user input; semantic understanding is realized by utilizing a natural language processing technology, and the query intention of the user is understood; a personalized recommendation system is established, and related chapters are recommended according to historical records of the user; and an interactive interface is designed, and a user is supported to use on different devices and platforms. According to the method, the user experience can be enhanced through personalized recommendation, so that the user can more easily find interested contents, the user satisfaction and the use efficiency are improved, and the user can easily browse, search and navigate the system through the design of the interactive interface.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent content navigation method and system based on power grid standard chapters. Background Art

[0002] In recent years, the standard system of power grid equipment has been continuously improved, and a large number of technical standards and technical documents have been issued in various stages such as planning and design, procurement and manufacturing, engineering construction, operation and maintenance, and decommissioning. The release and implementation of these standards have played an important supporting role in supporting the development of the power grid. For example, when maintaining power grid equipment, it is often necessary to refer to these power grid equipment standards, and when the power grid equipment is in use, it is necessary to search and navigate the content of the power grid chapter, so there is an urgent need for a fast navigation method that is easy for people to use. The following problems exist in the existing technology:

[0003] 1. The existing intelligent content navigation method based on power grid standard chapters cannot enable users to find the content they are interested in more quickly and easily, cannot meet user needs, and cannot provide users with a more intuitive and friendly user experience during use. It has poor practicality. At the same time, the accessibility and coverage of the system are relatively small and too limited. Summary of the invention

[0004] In view of the problems existing in the existing intelligent content navigation method based on standard sections of power grid, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is how to find interesting content more easily, thereby improving user satisfaction and usage efficiency.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides an intelligent content navigation method based on power grid standard chapters, which includes classifying and labeling the content of power grid standard chapters; building an intelligent search engine to quickly locate relevant chapters based on user input; using natural language processing technology to achieve semantic understanding and understand the user's query intent; establishing a personalized recommendation system to recommend relevant chapters based on user historical records; designing an interactive interface and supporting users to use it on different devices and platforms.

[0008] As a preferred solution of the intelligent content navigation method based on power grid standard chapters described in the present invention, the use of natural language processing technology to achieve semantic understanding includes the following steps: identifying the relationship between entities from the text; applying dependency syntactic analysis to identify the phrase structure in the sentence; constructing a topic-keyword mapping table to associate keywords with predefined topics.

[0009] As a preferred solution of the intelligent content navigation method based on standard power grid chapters described in the present invention, the construction of a topic-keyword mapping table and associating keywords with predefined topics includes the following steps: retrieving chapters in the index based on the identified topics and keywords; using a vector space model to calculate the similarity between the query and the chapter; and applying semantic retrieval technology to calculate semantic similarity.

[0010] As a preferred solution of the intelligent content navigation method based on power grid standard chapters of the present invention, wherein: the searching for chapters in the index according to the identified topics and keywords includes: extracting the topics of the document set using the LDA algorithm, wherein the probability of the LDA generation process is:

[0011]

[0012] Where t is a word, z is a topic, d is a document, N0 is the number of keywords selected for each topic; K is the number of topics; the TF-IDF algorithm is used to extract keywords for each topic, where the TF-IDF calculation formula is:

[0013] TF-IDF(t,d,D)=TF(t,d)×IDF(t,D)

[0014] TF(t,d)=c t / c0

[0015] IDF(t,D)=log(z d / z0)

[0016] Among them, c t is the number of occurrences of word t in document d, c0 is the total number of words in document d; z d is the total number of documents, z0 is the number of documents containing word t; for each topic, select the first h0 words with the highest TF-IDF value and build a mapping table.

[0017] As a preferred solution of the intelligent content navigation method based on power grid standard chapters described in the present invention, wherein: the use of the vector space model to calculate the similarity between the query and the chapter includes: using an inverted index structure to segment each chapter and construct a word-chapter mapping; according to the input keywords, find the relevant topics from the mapping table, and expand the query q to: q'=q∪{all keywords related to q}; find all chapters containing any word in q', and return a candidate chapter set C; use the TF-IDF weighted bag of words model to represent documents and queries, document vector: d=[w1,w2,...,wn], where wi is the TF-IDF weight of word i; for each candidate chapter c, calculate the cosine similarity with the query q', where the cosine similarity formula is:

[0018]

[0019] Among them, n0 is the number of elements contained in q' and c; the candidate chapters are sorted in descending order according to the cosine similarity.

[0020] As a preferred solution of the intelligent content navigation method based on standard power grid chapters described in the present invention, the establishment of a personalized recommendation system and the recommendation of relevant chapters based on user history records include the following steps: analyzing the user's historical search records and browsing behavior to understand preferences and interests; and recommending other chapters and topics related to the user's interests based on the content of the chapters that the user has viewed.

[0021] As a preferred solution of the intelligent content navigation method based on the power grid standard chapter of the present invention, the analysis of the user's historical search records and browsing behaviors to understand preferences and interests includes: collecting data on search records and browsing behaviors, and calculating the user's interest in a certain topic:

[0022] I(u,z)=ω1×S(u,z)+ω2×B(u,z)

[0023] Where u is the user, S(u,z) is the search frequency, B(u,z) is the browsing time, ω1 and ω2 are weight parameters; a linear decay function is applied to reduce the impact of old behavior:

[0024] D(z)=1-(cu time -ac time ) / max time

[0025] Where D(z) is the time decay factor, cu time is the current time, ac time is the time when the behavior occurs, max time is the maximum time span; the recommendation of other chapters and topics related to the user's interests includes: recommending similar chapters based on the chapters the user has recently browsed:

[0026] R u ={c|max(J(c,v)),v∈V(u)}

[0027] Where V(u) is the set of chapters recently browsed by user u; R u is the set of chapters recommended for user u; c is the candidate chapter; J(c,v) is the Jaccard similarity between the candidate chapter c and the chapter v most recently browsed by the user.

[0028] In the second aspect, an embodiment of the present invention provides an intelligent content navigation system based on power grid standard chapters, which includes a classification module for classifying power grid standard chapters by topic; a labeling module for adding appropriate labels to each chapter; an intelligent search engine for quickly locating relevant chapters based on user input; a semantic understanding module for understanding the user's query intent; a personalized recommendation module for recommending relevant chapters based on user history records; and an interactive interface module for providing an intuitive and friendly user interface.

[0029] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the intelligent content navigation method based on power grid standard chapters as described in the first aspect of the present invention are implemented.

[0030] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the intelligent content navigation method based on power grid standard chapters as described in the first aspect of the present invention are implemented.

[0031] The beneficial effects of the present invention are as follows: the present invention can enhance the user experience through personalized recommendations, making it easier to find content of interest, thereby improving user satisfaction and usage efficiency; the interactive interface design can enable users to easily browse, search and navigate the system, providing an intuitive and friendly user experience; the interactive design aims to make the user's interaction with the content of the power grid chapter natural, efficient and enjoyable, and can reduce the user's cognitive burden and operational difficulties, so that users can easily complete tasks, improve user satisfaction with the power grid and desire to use it, and by analyzing user needs and behavior patterns, design interfaces and functions that meet user expectations to help users better achieve their goals. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0033] Figure 1 The flowchart of the intelligent content navigation method based on power grid standard chapters. DETAILED DESCRIPTION

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0035] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0036] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0037] Example 1

[0038] Reference Figure 1 , which is the first embodiment of the present invention, and provides an intelligent content navigation method based on power grid standard chapters, including:

[0039] S1: Categorize and label the content of power grid standard chapters.

[0040] Specifically, identify the chapter structure of the document, such as titles, subheadings, paragraphs, etc., and split the document into independent chapter units;

[0041] Define the main categories of power grid standards, such as power system planning, transmission technology, distribution system, etc. Use the relational classification model support vector machine and train the classifier based on the classification algorithm to automatically classify each chapter into predefined categories.

[0042] Extract keywords and phrases from each chapter, filter and select the most representative keywords, generate tags based on the extracted keywords and chapter classification results, associate the tags with the corresponding chapters, establish a tag index, support fast retrieval and filtering, and realize visual display of tags to facilitate user browsing and selection.

[0043] S2: Build an intelligent search engine to quickly locate relevant chapters based on user input.

[0044] Smart search is to implement an intelligent search engine that can quickly locate relevant chapters based on user input and provide relevant suggestions to help users find the information they need more accurately, including:

[0045] Choose a suitable search engine framework, such as Elasticsearch or Solr, design the index structure, including fields such as chapter content, title, and tags, use a tokenizer to tokenize the text, support Chinese tokenization, and establish an inverted index to improve search efficiency.

[0046] S3: Use natural language processing technology to achieve semantic understanding and understand the user's query intention.

[0047] Semantic understanding and natural language processing use natural language processing technology to enable the system to understand the user's query intent and provide corresponding chapter links and content summaries. This is achieved by implementing semantic analysis and keyword extraction technology.

[0048] Furthermore, semantic understanding and natural language processing include the following steps:

[0049] S3.1: Use natural language processing technology to understand the semantics of user queries, not just simple keyword matching. This includes identifying entities, relationships, and actions in queries to better understand user intent. Entity relationship recognition is the process of identifying the relationship between entities from text. The specific recognition process is as follows:

[0050] The input text is preprocessed, including word segmentation, part-of-speech tagging and named entity recognition, and the identified entity set is denoted by E={E1,E2,...Ei,...,En}, where Ei represents the i-th identified entity.

[0051] Define a relationship type set R = {R1, R2, ...Ri, ..., Rm}, where Ri represents the i-th predefined relationship type; for any two entities Ei and Ej (i≠j), determine whether there is a relationship between them, and record the possible relationship as Lx(Ei→Ej), where Lx represents the x-th relationship type, and Lx∈R.

[0052] For each pair of entities (Ei, Ej), extract the context information between them, including the word sequence and syntactic structure between the entities, and record the context feature set of Ei as CEi, and the context feature set of Ej as CEj. Based on CEi and CEj, generate the relation feature set of the entity pair (Ei, Ej), denoted as Fij, and Fij=CEi∪CEj.

[0053] Use the pre-trained relationship classification model to classify the feature set Fij and obtain the relationship type probability distribution P(Lx|Fij); if max(P(Lx|Fij))>θ, where θ is the preset probability threshold, it is considered that there is a relationship Lx between entities Ei and Ej; otherwise, it is considered that there is no predefined relationship between Ei and Ej.

[0054] All identified entity relationships are counted to generate a relationship set W, all entity pairs that are identified to have relationships are generated as a set Q, and all entity pairs that are determined to have no relationship are generated as a set N; the coverage rate C of relationship identification is calculated, and the specific calculation formula is as follows:

[0055]

[0056] Among them, NUM(Q) and NUM(N) respectively represent the number of entity pairs in set Q and set N.

[0057] S3.2: By analyzing the keywords and phrases in the user query, the system can quickly identify the topics of interest to the user and provide relevant chapter links and content summaries accordingly.

[0058] Preferably, the user input is segmented, stop words are removed, such as meaningless words like "de", "le", "shi", etc., and "power grids" is restored to "power grid".

[0059] The analysis process of phrases is as follows: Use part-of-speech tagging to identify noun phrases, verb phrases, etc., and apply dependency syntax analysis to identify the phrase structure in the sentence. The specific operations are as follows:

[0060] Use a dependency syntax analysis tool, such as Stanford Parser, to determine the dependency relationships between words in the sentence. Each word has a dependent word and is connected to it through the dependency relationship; construct a dependency tree based on the dependency relationships, where each node represents a word and the edges represent the dependency relationships.

[0061] In the dependency tree, the phrase structure can be identified through the following steps:

[0062] The core verb of the sentence is usually the root node of the dependency tree. Look for subtrees centered on nouns. The noun and its modifiers, such as adjectives, definite articles, etc., form a noun phrase; look for subtrees centered on verbs. The verb and its object, adverbial, etc., form a verb phrase. Take the sentence "The quick brown fox jumps over the lazy dog." as an example. The root node is the verb "jumps". The noun phrases include "The quick brown fox" and "the lazy dog", and the verb phrase is "jumps over the lazy dog".

[0063] S3.3: Construct a topic-keyword mapping table to associate keywords with predefined topics, including the following steps:

[0064] S3.3.1: Retrieve relevant chapters in the index according to the identified topics and keywords.

[0065] Preferably, use the LDA algorithm to extract the topics of the document set. Among them, the probability of the LDA generation process is:

[0066]

[0067] Among them, t is a word, z is a topic, d is a document, N0 is the number of keywords selected for each topic; K is the number of topics;

[0068] The TF-IDF algorithm is used to extract keywords for each topic, where the TF-IDF calculation formula is:

[0069] TF-IDF(t,d,D)=TF(t,d)×IDF(t,D)

[0070] TF(t,d)=c t / c0

[0071] IDF(t,D)=log(z d / z0)

[0072] Among them, c t is the number of occurrences of word t in document d, c0 is the total number of words in document d; z d is the total number of documents, z0 is the number of documents containing word t; for each topic, select the first few words with the highest TF-IDF value, such as 20 words as keywords, and build a mapping table: {topic i: [keyword 1, keyword 2, ..., keyword N]}, etc.

[0073] S3.3.2: Use the vector space model to calculate the similarity between the query and the chapter.

[0074] Preferably, an inverted index structure is used to segment each chapter and construct a word-chapter mapping; based on the input keywords, relevant topics are found from the mapping table, and the query q is expanded to: q'=q∪{all keywords of topics related to q}; all chapters containing any word in q' are found, and a candidate chapter set C is returned.

[0075] The TF-IDF weighted bag-of-words model is used to represent documents and queries. The document vector is: d = [w1, w2, ..., wn], where wi is the TF-IDF weight of word i; for each candidate chapter c, the cosine similarity with the query q' is calculated, where the cosine similarity formula is:

[0076]

[0077] Among them, n0 is the number of elements contained in q' and c; the candidate chapters are sorted in descending order according to the cosine similarity.

[0078] S3.3.3: Apply semantic retrieval techniques, such as using BERT to calculate semantic similarity.

[0079] Specifically, the pre-trained BERT model is used to encode the query q' and each candidate chapter c, where the BERT parameters include: model: BERT-base (12 layers, 768-dimensional hidden layer, 12 attention heads), maximum sequence length: 512 tokens.

[0080] Using the output of the [CLS] tag as the representation of the entire sequence, the query representation is: E q =BERT(q′); Section representation: E c =BERT(c).

[0081] The semantic similarity between the query and the chapter is calculated using cosine similarity:

[0082] sim(q',c)=cos(E q ,E c )

[0083] Combining the results of the vector space model and BERT semantic similarity, the final similarity calculation formula is:

[0084] S(q',c)=γ×cos(q',c)+(1-γ)×sim(q',c)

[0085] Wherein, γ is a balancing factor, and the optimal value (such as λ=0.3) can be determined by cross-validation.

[0086] The candidate chapters are finally sorted according to S(q',c), and the first few, such as the 10 most relevant chapters, are returned as retrieval results.

[0087] S4: Establish a personalized recommendation system to recommend relevant chapters based on user history records.

[0088] Personalized recommendations are based on users' historical search records and browsing behaviors, recommending relevant chapters or topics to them to improve user experience and increase content discoverability.

[0089] Furthermore, personalized recommendation includes the following steps:

[0090] S4.1: Analyze users’ historical search records and browsing behaviors to understand their preferences and interests so as to recommend relevant grid standards-based chapters and topics.

[0091] Collect data, including search records, such as keywords, timestamps, click results, etc., and browsing behaviors including section ID, dwell time, scroll depth, etc.

[0092] Calculate the user's interest in a topic:

[0093] I(u,z)=ω1×S(u,z)+ω2×B(u,z)

[0094] Among them, u is the user, S(u,z) is the search frequency, B(u,z) is the browsing time, and ω1 and ω2 are weight parameters.

[0095] Apply a linear decay function to reduce the impact of the old behavior:

[0096] D(z)=1-(cu time -ac time ) / max time

[0097] Where D(z) is the time decay factor, cu time is the current time, ac time is the time when the behavior occurs, max time is the maximum time span used to normalize time differences.

[0098] S4.2: Recommend other chapters and topics related to the user's interests based on the content of the chapters the user has viewed.

[0099] Recommend similar chapters based on the chapters the user has recently viewed:

[0100] R u ={c|max(J(c,v)),v∈V(u)}

[0101] Where V(u) is the set of chapters recently browsed by user u; R u is the set of chapters recommended for user u; c is the candidate chapter; J(c,v) is the Jaccard similarity between the candidate chapter c and the chapter v most recently browsed by the user.

[0102] Personalized recommendations can enhance user experience and make it easier to find content of interest, thereby improving user satisfaction and usage efficiency.

[0103] S5: Design interactive interfaces to provide intuitive and friendly user experience.

[0104] Interactive interface design is to design an intuitive and friendly user interface that enables users to easily browse, search and navigate the power grid standard chapters, including concise and clear navigation menus, search boxes and filter functions.

[0105] Among them, the interactive interface design can enable users to browse, search and navigate the system easily, providing an intuitive and friendly user experience. The interactive design aims to make the user's interaction with the content of the power grid chapter natural, efficient and pleasant, and can reduce the user's cognitive burden and operational difficulties, so that users can easily complete tasks, improve user satisfaction with the power grid and desire to use it, and by analyzing user needs and behavior patterns, design interfaces and functions that meet user expectations to help users better achieve their goals.

[0106] S6: Support users to use the system on different devices and platforms.

[0107] Multi-platform support ensures that the system can run on different devices and platforms, including desktops, mobile devices, and web browsers, to meet the needs of users in different scenarios. Multi-platform support allows users to use the system on different devices and platforms, increasing the accessibility and coverage of the system, thereby expanding the user base.

[0108] Furthermore, the present embodiment also provides an intelligent content navigation system for power grid standard chapters, including a content classification module for classifying power grid standard chapters by topic; a labeling module for adding appropriate labels to each chapter; an intelligent search engine for quickly locating relevant chapters based on user input; a semantic understanding module for understanding the user's query intent; a personalized recommendation module for recommending relevant chapters based on user history records; and an interactive interface module for providing an intuitive and friendly user interface.

[0109] This embodiment also provides a computer device, which is suitable for the case of an intelligent content navigation method for power grid standard chapters, and includes a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the intelligent content navigation method based on power grid standard chapters as proposed in the above embodiment.

[0110] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0111] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing intelligent content navigation of power grid standard chapters as proposed in the above embodiment is implemented.

[0112] In summary, the present invention can enhance the user experience through personalized recommendations, making it easier to find content of interest, thereby improving user satisfaction and usage efficiency. The interactive interface design can enable users to easily browse, search and navigate the system, providing an intuitive and friendly user experience. The interactive design aims to make the user's interaction with the content of the power grid chapter natural, efficient and enjoyable, and can reduce the user's cognitive burden and operational difficulties, so that users can easily complete tasks, improve user satisfaction with the power grid and desire to use it, and by analyzing user needs and behavior patterns, design interfaces and functions that meet user expectations to help users better achieve their goals.

[0113] Example 2

[0114] Referring to Table 1, which is a second embodiment of the present invention, this embodiment provides an intelligent content navigation method based on power grid standard chapters. In order to verify the beneficial effects of the present invention, a scientific demonstration is carried out through simulation experiments.

[0115] In order to verify the effectiveness of the intelligent content navigation method based on power grid standard chapters, the research team selected the power grid standard document library of a provincial power company as the research object. The document library contains about 10,000 power grid-related standard documents, covering multiple fields such as power system planning, transmission technology, and distribution system.

[0116] First, the research team used natural language processing technology to pre-process the documents, including text cleaning, word segmentation, and removal of stop words. Then, a relational classification model was trained using the support vector machine (SVM) algorithm to automatically classify the documents. The model's training set contained 5,000 manually annotated documents, and the test set contained 1,000 documents.

[0117] Next, the research team used the TF-IDF algorithm and the TextRank algorithm to extract keywords and phrases for each document.

[0118] To optimize the keyword extraction effect, the researchers set different parameter combinations and selected the best parameter settings through manual evaluation. The extracted keywords and phrases were combined with the document classification results to generate a tag index.

[0119] In terms of building an intelligent search engine, the research team chose Elasticsearch as the search engine framework and customized the configuration based on the characteristics of power grid standard documents. In order to improve the accuracy of Chinese searches, the team used the jieba word segmenter for Chinese word segmentation and built a dictionary for specific fields.

[0120] The research team used the BERT model for semantic encoding. In order to adapt to the particularity of the power grid field, the researchers used 100,000 power grid-related documents based on the pre-trained BERT-base model to perform domain adaptability fine-tuning and obtained the ElectricBERT model.

[0121] In order to build a personalized recommendation system, the team collected search records and browsing behavior data from 100 power engineers during their three-month system use. Based on these data, the team developed a hybrid recommendation algorithm based on collaborative filtering and content recommendation.

[0122] Responsive design principles were adopted to ensure that the system can provide a good user experience on different devices. The team conducted multiple rounds of user testing and iterative optimization, and finally determined an intuitive and friendly interface design.

[0123] In order to support multi-platform use, the research team adopted a front-end and back-end separation architecture design. The back-end used the Java Spring Boot framework to develop RESTful APIs, and the front-end used the Vue.js framework to develop web applications, as well as Android and iOS native applications.

[0124] After completing the system development, the research team invited 50 power engineers to conduct a two-week system trial and collected their usage data and feedback. The following is the detailed data of the trial results:

[0125] Table 1 Test results

[0126]

[0127] By analyzing the above table data, it can be clearly seen that in terms of search accuracy, the intelligent content navigation system has reached 92.5%, which is much higher than the 76.8% of the traditional keyword search system and the 81.3% of the rule-based navigation system. This significant improvement is mainly due to the semantic understanding and natural language processing technology used in the present invention. In particular, the ElectricBERT model fine-tuned for the power grid field was used, and its accuracy reached 94.2%, which is a significant improvement over the 89.6% of the general BERT model. This shows that the domain adaptive fine-tuning strategy proposed in the present invention can effectively improve the system's understanding ability in professional fields.

[0128] The intelligent content navigation system can complete the search in just 156 milliseconds, which is twice as fast as the traditional keyword search system and nearly 10 times faster than manual assisted search. This is due to the efficient index structure and intelligent search engine optimization adopted in the present invention.

[0129] It can be seen from the above table that the intelligent content navigation method based on power grid standard chapters proposed in the present invention has obvious advantages in many aspects such as search accuracy, retrieval speed, user satisfaction, knowledge acquisition efficiency, system response time and cross-platform compatibility.

[0130] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent content navigation method based on power grid standard chapters, Features: include, Categorize and label the content of power grid standard chapters; Build an intelligent search engine to quickly locate chapters based on user input; Use natural language processing technology to achieve semantic understanding and understand the user's query intent; Establish a personalized recommendation system to recommend chapters based on user history records; Design interactive interfaces and support users on different devices and platforms.

2. The intelligent content navigation method based on the standard chapter of the power grid according to claim 1, Features: The method of realizing semantic understanding by using natural language processing technology comprises the following steps: Identify relationships between entities from text; Apply dependency parsing to identify phrase structures in sentences; Build a topic-keyword mapping table to associate keywords with predefined topics.

3. The intelligent content navigation method based on the standard chapter of the power grid as claimed in claim 2, Features: The construction of the topic-keyword mapping table and associating keywords with predefined topics includes the following steps: Search the index for chapters based on the identified topics and keywords; Use vector space model to calculate the similarity between query and chapter; Apply semantic retrieval technology to calculate semantic similarity.

4. The intelligent content navigation method based on the standard chapter of the power grid as claimed in claim 3, Features: The sections searched in the index based on the identified topics and keywords include: The LDA algorithm is used to extract the topic of the document set, where the probability of the LDA generation process is: Among them, t is a word, z is a topic, d is a document, and N 0 is the number of keywords selected for each topic; K is the number of topics; The TF-IDF algorithm is used to extract keywords for each topic, where the TF-IDF calculation formula is: TF-IDF(t,d,D)=TF(t,d)×IDF(t,D) TF(t,d)=c t / c 0 IDF(t,D)=log(z d / z 0 ) Among them, c t is the number of occurrences of word t in document d, c 0 is the total number of words in document d; z d is the total number of documents, z 0 is the number of documents containing word t; select the top / bottom with the highest TF-IDF value for each topic 0 words and build a mapping table.

5. The intelligent content navigation method based on the standard chapter of the power grid as claimed in claim 4, Features: The method of calculating the similarity between the query and the chapter using the vector space model includes: Use the inverted index structure to segment each chapter and build a word-chapter mapping; According to the input keywords, relevant topics are found from the mapping table, and the expanded query q is: q'=q∪{all keywords related to the topic q}; Find all chapters containing any word in q' and return the candidate chapter set C; Use the TF-IDF weighted bag-of-words model to represent documents and queries, with a document vector: d = [w1, w2, ..., wn], where wi is the TF-IDF weight of word i; For each candidate chapter c, calculate the cosine similarity with the query q', where the cosine similarity formula is: Among them, n 0 is the number of elements contained in q' and c; the candidate chapters are sorted in descending order according to the cosine similarity.

6. The intelligent content navigation method based on the standard chapter of the power grid according to claim 5, Features: The establishment of a personalized recommendation system to recommend relevant chapters based on user history records includes the following steps: Analyze users’ historical search history and browsing behavior to understand their preferences and interests; Recommend other chapters and topics related to the user's interests based on the content of the chapters the user has viewed.

7. The intelligent content navigation method based on the standard chapter of the power grid according to claim 6, Features: Analyzing the user's historical search records and browsing behaviors to understand preferences and interests includes: Collect data on search history and browsing behavior to calculate the user's interest in a topic: I(u,z)=ω 1 ×S(u,z)+ω 2 ×B(u,z) Among them, u is the user, S(u,z) is the search frequency, B(u,z) is the browsing time, ω 1 and ω 2 is the weight parameter; Apply a linear decay function to reduce the impact of the old behavior: D(z)=1-(with time -needle time ) / max time Where D(z) is the time decay factor, cu time is the current time, ac time is the time when the behavior occurs, max time is the maximum time span; Other sections and topics recommended by the author that are relevant to his interests include: Recommend similar chapters based on the chapters the user has recently viewed: R u ={c|max(J(c,v)),v∈V(u)} Where V(u) is the set of chapters recently browsed by user u; R u is the set of chapters recommended for user u; c is the candidate chapter; J(c,v) is the Jaccard similarity between the candidate chapter c and the chapter v most recently browsed by the user.

8. An intelligent content navigation system based on power grid standard chapters, based on the intelligent content navigation method based on power grid standard chapters according to any one of claims 1 to 7, Features: Also includes, Classification module, used to classify power grid standard chapters by subject; The labeling module is used to add appropriate labels to each chapter; Intelligent search engine for quickly locating relevant chapters based on user input; Semantic understanding module, used to understand the user's query intention; Personalized recommendation module, used to recommend relevant chapters based on user history records; The interactive interface module is used to provide an intuitive and friendly user interface.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. Features: When the processor executes the computer program, the steps of the intelligent content navigation method based on power grid standard chapters according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, Features: When the computer program is executed by a processor, the steps of the intelligent content navigation method based on power grid standard chapters according to any one of claims 1 to 7 are implemented.

Images