Information management method based on intelligent classification and efficient retrieval

Abstract

The invention discloses an information management method based on intelligent classification and efficient retrieval, relates to the technical field of information management, and solves the technical problem of inaccurate retrieval results caused by lack of deep understanding of information semantics and content. Secondary classification is carried out based on semantics, a standard classification result is determined through an integration and screening mechanism, the information classification accuracy is improved, the core content of information is reflected, word segmentation is carried out through a statistical model to determine retrieval heads, the retrieval heads without retrieval content are removed, the retrieval efficiency is improved, the content to be analyzed is screened by calculating the number proportion, and the retrieval efficiency is improved. Retrieval content is screened according to a scientific quantitative standard, information highly related to user requirements can be quickly positioned, the association strength of a retrieval head and retrieval information is accurately analyzed by means of assigning a grammatical position of a pre-selected retrieval head, calculating cosine similarity and the like, and a strong-association retrieval head and a weak-association retrieval head are distinguished.

Description

Technical Field

[0001] The present invention relates to the technical field of information management, and in particular to an information management method based on intelligent classification and efficient retrieval. Background Art

[0002] With the rapid development of information technology, the amount of information has exploded. Whether in academic research, business operations or daily life, people are faced with the challenge of quickly and accurately obtaining the required content from massive information.

[0003] According to the patent application with application number CN202410441933.8, an information management method based on intelligent classification and efficient retrieval is disclosed. By collecting the data that needs to be managed and processed and preprocessing it for subsequent classification and retrieval operations, the burden of manual operation is reduced; feature extraction and text representation are performed on the data; a classification model is constructed to intelligently classify and organize the data for subsequent efficient retrieval and query. Intelligent classification and efficient retrieval can help users find the required information quickly and accurately, and improve work efficiency; according to the user's search conditions and query requirements, data is quickly located and retrieved through the classification model; personalized data recommendations are provided based on the user's interests and preferences so that the user can find the required data. By analyzing the user's behavior and interests, personalized recommendations and search results are provided to meet the user's specific needs; through interaction and feedback with the user, the classification and retrieval results are continuously optimized to improve the accuracy of the retrieval.

[0004] Traditional information retrieval and management methods often rely on simple keyword matching and lack a deep understanding of information semantics and content, resulting in inaccurate retrieval results and imprecise classification. They are unable to meet users' growing demand for efficient and accurate information acquisition. When faced with complex retrieval needs, existing information management systems are unable to quickly locate and provide information that meets user expectations, resulting in a waste of time and resources. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides an information management method based on intelligent classification and efficient retrieval, which solves the problem of inaccurate retrieval results caused by a lack of in-depth understanding of information semantics and content.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an information management method based on intelligent classification and efficient retrieval, the method specifically comprising the following steps: Obtain search information, extract keywords, perform preliminary classification, and then perform secondary classification based on semantics. Integrate the classification results. If the classifications are the same, mark them as standard classification results. If they are different, filter keywords to select standard classification results. The search direction is determined based on the standard classification results. After word segmentation, the word segmentation results are generated by the statistical model, and their indexes are used as search heads. After the search, search heads without content are removed to obtain pre-selected search heads. Obtain pre-selected search headers and search contents, extract common semantics as the content to be analyzed, calculate the ratio of the number of search header types and pre-selected search headers in the content to be analyzed, compare with the preset ratio, and retain the content with a ratio greater than the preset ratio to obtain the retained content; Extract pre-selected search headers from the retained content, calculate cosine similarity with other pre-selected search headers according to grammatical position assignment, and mark those with similarity greater than or equal to a preset similarity as associated pre-selected search headers; Obtain all pre-selected search heads and the number of associated pre-selected search heads, calculate the index correlation strength value corresponding to the pre-selected search head, and compare it with the preset comparison value. If it is greater than the preset comparison value, it is marked as a strongly associated search head, and if it is less than the preset comparison value, it is marked as a weakly associated search head; Get strong and weak correlation search heads and retained content, analyze weak correlation, filter content whose proportion is less than the first preset value, analyze strong correlation, remove the part whose number of weak correlation heads exceeds the second preset value, summarize the filtered content, sort in descending order by the number of word segmentation results, generate and display the search results.

[0007] As a further solution of the present invention, the specific method of screening out the standard classification results is: Acquire the search information corresponding to the search user, extract keywords, and perform preliminary classification based on the keywords to obtain preliminary classification results, then acquire the semantics of the search information, and perform secondary classification based on the semantics of the search information to obtain secondary classification results; Integrate and judge the preliminary classification results and the secondary classification results, obtain the common classification results and mark them as standard classification results. Otherwise, if there is no identical classification result between the two, perform secondary analysis. All keywords and their corresponding semantics are obtained, and the keywords and semantics are matched and screened to obtain pre-selected keywords. At the same time, the preliminary classification results and the secondary classification results are screened based on the pre-selected keywords to obtain standard classification results.

[0008] As a further solution of the present invention, the specific method of obtaining the pre-selected search header is: Obtain the standard classification results corresponding to the search information, and determine the search direction at the same time. Then, perform word segmentation on the search information according to the search direction to obtain the word segmentation result. The obtained index is recorded as the search head and labeled as i, where i=1, 2, ..., j, where j represents the number of search heads. Then, use the search head as the standard to obtain the corresponding search content, which is recorded as Ki, and K=A, B, ..., where A and B represent the corresponding search content information, and based on the search content, eliminate the search head i to obtain the pre-selected search head a, where a=1, 2, ..., b.

[0009] As a further solution of the present invention, the specific method of retaining content greater than a preset proportion to obtain the retained content is: Obtain a pre-selected search header a and the corresponding search content Ka, and simultaneously obtain the common semantic content between the search contents Ka based on the semantics of the search contents, recorded as the content to be analyzed n, where n=1, 2, ..., m, where m represents the number of the content to be analyzed; The types of search headers in the content to be analyzed n are obtained and recorded as search headers to be analyzed. At the same time, the quantitative ratio of the search headers to be analyzed and the pre-selected search headers is calculated, and the quantitative ratio is compared with the preset ratio. The corresponding content to be analyzed is retained to obtain the retained content. Conversely, if the quantitative ratio is less than the preset ratio, the corresponding content to be analyzed is eliminated.

[0010] As a further solution of the present invention, the specific screening method of the associated pre-selected search header is: Get the grammatical position corresponding to the pre-selected search header a, perform value assignment based on different grammatical positions and record it as Ca, calculate the similarity between the pre-selected search headers a, and compare it with the preset similarity. The specific value of the preset similarity is set by the operator, and the pre-selected search headers corresponding to the similarity greater than the preset similarity are marked as associated pre-selected search headers.

[0011] As a further solution of the present invention, the specific method of obtaining the strong correlation search head and the weak correlation search head is: Get all pre-selected search heads a and the number of corresponding associated pre-selected search heads La, and then use the formula Calculate the index association strength value Sa corresponding to the pre-selected search head a, where q is the index adjustment factor, and compare the obtained index association strength value Sa with the preset comparison value Sy; The pre-selected search head a corresponding to the index association strength value Sa greater than the preset comparison value Sy is marked as a strongly associated search head, and the pre-selected search head a corresponding to the index association strength Sa less than the preset comparison value Sy is marked as a weakly associated search head.

[0012] As a further solution of the present invention, the specific method of generating and displaying the search results is: Obtain the number of weakly associated search headers in the retained content, calculate the proportion of weakly associated search headers, and compare the proportion with the first preset value, select and filter the retained content whose proportion of weakly associated search headers is less than the first preset value, and record it as the filtered content; The number of corresponding weakly associated search headers in the retained content is obtained, and the retained content whose number of corresponding weakly associated search headers is greater than a second preset value is eliminated, and the remaining retained content is recorded as filtered content, and the filtered content is displayed and analyzed.

[0013] As a further solution of the present invention, the specific method of displaying and analyzing the screened content is: All the filtered contents are obtained, and the corresponding word segmentation result quantity in the filtered contents is obtained at the same time, and the word segmentation result quantity is sorted from large to small, and the search results are generated, and then the obtained search results are displayed to the corresponding search users.

[0014] The present invention provides an information management method based on intelligent classification and efficient retrieval. Compared with the prior art, it has the following beneficial effects: The present invention extracts keywords for preliminary classification by combining machine learning algorithms, and then performs secondary classification based on semantics. The standard classification results are determined through integration and screening mechanisms, thereby improving the accuracy of information classification, and can more accurately reflect the core content of the information. The statistical model is used for word segmentation to determine the search header, and the search header without search content is eliminated, thereby reducing invalid searches and improving search efficiency. The content to be analyzed is screened by calculating the proportion of quantity, and the search content is screened according to scientific quantitative standards. Information that is highly relevant to user needs can be quickly located. By assigning grammatical positions to pre-selected search headers, calculating cosine similarity, etc., the correlation strength between the search header and the search information is accurately analyzed, and strongly correlated and weakly correlated search headers are distinguished, so as to provide users with more accurate and more demand-oriented search results, thereby improving the efficiency and quality of users in obtaining information. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a diagram of the steps of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] For example, see Figure 1 The present application provides an information management method based on intelligent classification and efficient retrieval, which specifically includes the following steps: Step 1: Obtain the corresponding search information, obtain the keywords in the search information, and preliminarily classify the search information based on the keywords to generate preliminary classification results. The specific classification method is: First, the search information corresponding to the search user is obtained, and the keywords in the search information are extracted. The extraction here is based on the corresponding classification system. For example, machine learning algorithms are used for extraction, such as decision trees, support vector machines, and neural networks. The labeled information data is learned, features and patterns are extracted, and a classification model is constructed. At the same time, the search information is preliminarily classified according to the obtained keywords to obtain a preliminary classification result. The preliminary classification result obtained here may include multiple types. Then, the semantics of the search information is obtained, and secondary classification is performed based on the semantics of the search information to obtain a secondary classification result. The obtained preliminary classification results and secondary classification results are integrated and judged. If there are identical classification results between the two, the common classification results are obtained and marked as standard classification results. Otherwise, if there are no identical classification results between the two, secondary analysis is performed. All keywords and their corresponding semantics are obtained, and the keywords and semantics are matched and filtered to obtain pre-selected keywords. At the same time, the preliminary classification results and the secondary classification results are filtered based on the pre-selected keywords to obtain standard classification results, and it is assumed here that there is only one standard classification result.

[0018] Step 2. Obtain the standard classification results corresponding to the search information, and determine the search direction at the same time. Then, perform word segmentation processing on the search information according to the search direction to obtain the word segmentation results. Specifically, use a statistical model to model the character sequence in the search information, calculate the probability that each character belongs to a different vocabulary, and thus determine the word segmentation results. At the same time, the obtained index is recorded as the search header, and the label is recorded as i, and i=1, 2, ..., j, where j represents the number of search headers. Then, the search is performed based on the search header as the standard to obtain the corresponding search content, and it is recorded as Ki, and K=A, B, ..., where A and B represent the corresponding search content information, and the search header i is eliminated based on the search content to obtain the pre-selected search header a, and a=1, 2, ..., b. The specific elimination method is: eliminate the search headers that do not have the search content, and retain the search headers that have the search content; For example, if the search information is "application of artificial intelligence in medical imaging diagnosis", according to the standard classification system of the database, its corresponding standard classification results may be determined as "computer science-artificial intelligence" and "medicine-medical imaging diagnosis", and the search direction is determined to be research related to the application of artificial intelligence in the medical field; Use statistical models (such as conditional random field models) to model the search information "Application of artificial intelligence in medical imaging diagnosis". By calculating the probability that each character belongs to different words, the word segmentation results are "artificial intelligence", "in", "medical", "imaging", "diagnosis", "in", "of", and "application"; Further, we get the corresponding search heads: search head 1 (i=1): "artificial intelligence", search head 2 (i=2): "medical", search head 3 (i=3): "imaging", search head 4 (i=4): "diagnosis", search head 5 (i=5): "application", now j=5 (the number of search heads), check the search content corresponding to these search heads, in this database, when searching with "in", "in", and "of" as search heads, no relevant academic literature is found (that is, there is no search content), while "artificial intelligence", "medical", "imaging", "diagnosis", and "application" all have corresponding search content. According to the elimination method, the search heads such as "in", "in", and "of" that do not have search content are eliminated.

[0019] Step 3, obtain the pre-selected search header a and the corresponding search content Ka, and at the same time obtain the common semantic content between the search contents Ka based on the semantics of the search contents, record it as the content to be analyzed, and label it as n, and n=1, 2, ..., m, where m represents the number of contents to be analyzed, then obtain the types of search headers in the content to be analyzed n, and record them as the search headers to be analyzed, and calculate the ratio of the number of search headers to be analyzed and the pre-selected search headers, and then sort the content to be analyzed n from large to small according to the ratio of the number; The content n to be analyzed is screened according to the quantity ratio, and the specific screening method is: the quantity ratio is compared with the preset ratio, and the specific value of the preset ratio is set by the operator, which is specifically set according to the number of all search heads. In this application, the value is one-fourth. If the quantity ratio is greater than the preset ratio, and this includes the case where it is greater than the preset ratio, the corresponding content to be analyzed is retained. On the contrary, if the quantity ratio is less than the preset ratio, the corresponding content to be analyzed is eliminated; Step 4: Obtain all the reserved contents to be analyzed and record them as the reserved contents. Then obtain the corresponding pre-selected search header a in the reserved contents. Meanwhile, analyze the correlation strength between the pre-selected search header a and the search information. The specific analysis method is as follows: The grammatical position corresponding to the preselected search head a is obtained, and the grammatical position here is mainly expressed as the position of the undertaking, such as subject, predicate, object or adverbial, and different assignment processing is performed based on different grammatical positions, which is recorded as Ca, and the assignment here is respectively assigned according to the subject, predicate, object and adverbial, for example, the subject is assigned to 3, the predicate is assigned to 2, and the object and adverbial are assigned to 1. Then the associated preselected search head of the preselected search head a is obtained, and here when the associated preselected search head is obtained, the calculation is performed by calculating the corresponding similarity, specifically taking one of the preselected search heads as the analysis object, and then quantizing the analysis object to obtain the corresponding vector, and so on, all the preselected search heads are quantized and the corresponding vectors are obtained, and then the cosine similarity between the analysis object and all the preselected search head vectors is calculated, and the obtained similarity is compared with the preset similarity, and the specific value of the preset similarity is set by the operator; If the similarity is greater than the preset similarity, and this includes the case where the similarity is equal to the preset similarity, the corresponding pre-selected search header is marked as the associated pre-selected search header. Otherwise, if the similarity is less than the preset similarity, no processing is performed; Get all the pre-selected search heads a, and get the number of corresponding associated pre-selected search heads as La, and then use the formula Calculate and obtain the index association strength value Sa corresponding to the pre-selected search head a, where q is the index adjustment factor, and the specific value is adjusted according to the actual situation to highlight or weaken the influence of the quantitative value of the association feature on the association strength, and compare the obtained index association strength value Sa with the preset comparison value Sy; If the index association strength value Sa is greater than the preset comparison value Sy, and here includes the case where it is equal to the preset comparison value, the corresponding pre-selected search head a is marked as a strongly associated search head, otherwise if the index association strength Sa is less than the preset comparison value Sy, the corresponding pre-selected search head a is marked as a weakly associated search head; Step 5, then obtain all the strongly associated search headers and weakly associated search headers, and obtain the corresponding retained content at the same time, and the retained content here is the same as the retained content analyzed in Step 3, and the corresponding retained content of the two is analyzed respectively; Analyze the reserved content corresponding to the weakly associated search header, obtain the number of weakly associated search headers in the reserved content, and calculate the proportion of the weakly associated number of the weakly associated search headers, and compare the obtained proportion of the weakly associated number with a first preset value, and the specific value of the first preset value is set by the operator, select and filter the reserved content corresponding to the weakly associated number proportion less than the first preset value, and record it as the filtered content, and the case where it is equal to the first preset value is not included here; Analyze the reserved content corresponding to the strongly associated search headers, obtain the number of corresponding weakly associated search headers in the reserved content, and remove the reserved content whose number of corresponding weakly associated search headers is greater than a second preset value, and record the remaining reserved content as the screened content; Then, all the filtered contents are obtained, and the corresponding word segmentation result quantity in the filtered contents is obtained, and the word segmentation result quantity is sorted from large to small, and the search results are generated, and then the obtained search results are displayed to the corresponding search users.

[0020] Some of the data in the above formulas are calculated by taking their numerical values, and are not substituted into parameter units for calculation. At the same time, the contents not described in detail in this specification belong to the existing technologies known to those skilled in the art.

[0021] The above embodiments are only used to illustrate the technical method 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 method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An information management method based on intelligent classification and efficient retrieval, characterized in that: The method specifically comprises the following steps: Obtain search information, extract keywords, perform preliminary classification, and then perform secondary classification based on semantics. Integrate the classification results. If the classifications are the same, mark them as standard classification results. If they are different, filter keywords to select standard classification results. The search direction is determined based on the standard classification results. After word segmentation, the word segmentation results are generated by the statistical model, and their indexes are used as search heads. After the search, search heads without content are removed to obtain pre-selected search heads. Obtain pre-selected search headers and search contents, extract common semantics as the content to be analyzed, calculate the ratio of the number of search header types and pre-selected search headers in the content to be analyzed, compare with the preset ratio, and retain the content with a ratio greater than the preset ratio to obtain the retained content; Extract pre-selected search headers from the retained content, calculate cosine similarity with other pre-selected search headers according to grammatical position assignment, and mark those with similarity greater than or equal to a preset similarity as associated pre-selected search headers; Obtain all pre-selected search heads and the number of associated pre-selected search heads, calculate the index correlation strength value corresponding to the pre-selected search head, and compare it with the preset comparison value. If it is greater than the preset comparison value, it is marked as a strongly associated search head, and if it is less than the preset comparison value, it is marked as a weakly associated search head; Get strong and weak correlation search heads and retained content, analyze weak correlation, filter content whose proportion is less than the first preset value, analyze strong correlation, remove the part whose number of weak correlation heads exceeds the second preset value, summarize the filtered content, sort in descending order by the number of word segmentation results, generate and display the search results.

2. The information management method based on intelligent classification and efficient retrieval according to claim 1, characterized in that: The specific method of filtering out the standard classification results is: Acquire the search information corresponding to the search user, extract keywords, and perform preliminary classification based on the keywords to obtain preliminary classification results, then acquire the semantics of the search information, and perform secondary classification based on the semantics of the search information to obtain secondary classification results; Integrate and judge the preliminary classification results and the secondary classification results, obtain the common classification results and mark them as standard classification results. Otherwise, if there is no identical classification result between the two, perform secondary analysis. All keywords and their corresponding semantics are obtained, and the keywords and semantics are matched and screened to obtain pre-selected keywords. At the same time, the preliminary classification results and the secondary classification results are screened based on the pre-selected keywords to obtain standard classification results.

3. The information management method based on intelligent classification and efficient retrieval according to claim 1 is characterized in that: The specific method of obtaining the pre-selected search header is: Obtain the standard classification results corresponding to the search information, and determine the search direction at the same time. Then, perform word segmentation on the search information according to the search direction to obtain the word segmentation result. The obtained index is recorded as the search head and labeled as i, where i=1, 2, ..., j, where j represents the number of search heads. Then, use the search head as the standard to obtain the corresponding search content, which is recorded as Ki, and K=A, B, ..., where A and B represent the corresponding search content information, and based on the search content, eliminate the search head i to obtain the pre-selected search head a, where a=1, 2, ..., b.

4. The information management method based on intelligent classification and efficient retrieval according to claim 1 is characterized in that: The specific method of retaining content greater than the preset proportion to obtain the retained content is: Obtain a pre-selected search header a and the corresponding search content Ka, and simultaneously obtain the common semantic content between the search contents Ka based on the semantics of the search contents, recorded as the content to be analyzed n, where n=1, 2, ..., m, where m represents the number of the content to be analyzed; The types of search headers in the content to be analyzed n are obtained and recorded as search headers to be analyzed. At the same time, the quantitative ratio of the search headers to be analyzed and the pre-selected search headers is calculated, and the quantitative ratio is compared with the preset ratio. The corresponding content to be analyzed is retained to obtain the retained content. Conversely, if the quantitative ratio is less than the preset ratio, the corresponding content to be analyzed is eliminated.

5. The information management method based on intelligent classification and efficient retrieval according to claim 1 is characterized in that: The specific screening method of the associated pre-selected search header is: Get the grammatical position corresponding to the pre-selected search header a, perform value assignment based on different grammatical positions and record it as Ca, calculate the similarity between the pre-selected search headers a, and compare it with the preset similarity. The specific value of the preset similarity is set by the operator, and the pre-selected search headers corresponding to the similarity greater than the preset similarity are marked as associated pre-selected search headers.

6. The information management method based on intelligent classification and efficient retrieval according to claim 1, characterized in that: The specific method of obtaining the strong correlation search head and the weak correlation search head is: Get all pre-selected search heads a and the number of corresponding associated pre-selected search heads La, and then use the formula Calculate the index association strength value Sa corresponding to the pre-selected search head a, where q is the index adjustment factor, and compare the obtained index association strength value Sa with the preset comparison value Sy; The pre-selected search head a corresponding to the index association strength value Sa greater than the preset comparison value Sy is marked as a strongly associated search head, and the pre-selected search head a corresponding to the index association strength Sa less than the preset comparison value Sy is marked as a weakly associated search head.

7. The information management method based on intelligent classification and efficient retrieval according to claim 1 is characterized in that: The specific method of generating and displaying the search results is: Obtain the number of weakly associated search headers in the retained content, calculate the proportion of weakly associated search headers, and compare the proportion with the first preset value, select and filter the retained content whose proportion of weakly associated search headers is less than the first preset value, and record it as the filtered content; The number of corresponding weakly associated search headers in the retained content is obtained, and the retained content whose number of corresponding weakly associated search headers is greater than a second preset value is eliminated, and the remaining retained content is recorded as filtered content, and the filtered content is displayed and analyzed.

8. The information management method based on intelligent classification and efficient retrieval according to claim 7 is characterized in that: The specific method of displaying and analyzing the screened content is as follows: All the filtered contents are obtained, and the corresponding word segmentation result quantity in the filtered contents is obtained at the same time, and the word segmentation result quantity is sorted from large to small, and the search results are generated, and then the obtained search results are displayed to the corresponding search users.

Images