Intelligent meeting management optimization method, device and equipment and readable storage medium

Abstract

The application discloses an intelligent conference management optimization method and device, equipment and a readable storage medium. The intelligent conference management optimization method comprises the following steps: acquiring conference content text, extracting each key text word in the conference content text; based on each key text word, the conference content text is divided into each target text; based on the text information of each target text, attribute labels are given to each target text to obtain each label identification text; according to the correlation between the text attribute label of each label identification text and the target participant, each label identification text is pushed to the corresponding target participant. The application solves the technical problem of low conference efficiency.

Description

Technical Field

[0001] This application relates to the field of big data technology in financial technology (Fintech), and in particular to a method, apparatus, device and readable storage medium for optimizing intelligent meeting management. Background Technology

[0002] With the continuous development of fintech, especially internet fintech, more and more technologies (such as distributed systems and artificial intelligence) are being applied in the financial field. However, the financial industry is also placing higher demands on technology, such as on the distribution of tasks to be completed.

[0003] With the continuous development of computer technology, remote conferencing software has become increasingly sophisticated. Participants can use this software to hold remote meetings from different locations, i.e., remote conferences. However, this method obviously requires participants from different locations to log in to the remote conferencing software at the same time to achieve a simultaneous remote meeting. However, during a remote meeting, participants may be unable to attend on time for various reasons, affecting meeting efficiency. At the same time, some participants may only be relevant to a portion of the meeting content, but these participants are required to attend the entire meeting, which also affects meeting efficiency. Summary of the Invention

[0004] The main objective of this application is to provide an intelligent meeting management optimization method, apparatus, device, and readable storage medium, aiming to solve the technical problem of low meeting efficiency in the prior art.

[0005] To achieve the above objectives, this application provides an intelligent meeting management optimization method, which is applied to an intelligent meeting management optimization device, and includes:

[0006] Obtain the meeting content text and extract key words from the meeting content text;

[0007] Based on the aforementioned key words, the meeting content text is segmented into target texts;

[0008] Based on the text information of each target text, attribute tags are assigned to each target text to obtain each tag identifier text;

[0009] Based on the correlation between the text attribute tags of each of the aforementioned tag identifiers and the meeting target, each of the aforementioned tag identifiers is pushed to the corresponding meeting target.

[0010] This application also provides an intelligent meeting management optimization device, which is a virtual device and is applied to an intelligent meeting management optimization equipment. The intelligent meeting management optimization device includes:

[0011] The keyword extraction module is used to obtain the meeting content text and extract key words from the meeting content text;

[0012] The text segmentation module is used to segment the meeting content text into target texts based on the aforementioned key text words;

[0013] The attribute label assignment module is used to assign attribute labels to each of the target texts based on the text information of each target text, so as to obtain the label identifier text;

[0014] The meeting content push module is used to push each labeled text to the corresponding meeting target based on the relevance between the text attribute tags of each labeled text and the meeting target.

[0015] This application also provides an intelligent meeting management optimization device, which is a physical device. The intelligent meeting management optimization device includes: a memory, a processor, and a program of the intelligent meeting management optimization method stored in the memory and executable on the processor. When the program of the intelligent meeting management optimization method is executed by the processor, it can implement the steps of the intelligent meeting management optimization method as described above.

[0016] This application also provides a computer-readable storage medium storing a program for implementing an intelligent meeting management optimization method. When the program for the intelligent meeting management optimization method is executed by a processor, it implements the steps of the intelligent meeting management optimization method as described above.

[0017] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the intelligent meeting management optimization method described above.

[0018] This application provides an intelligent meeting management optimization method, apparatus, device, and readable storage medium. Compared to existing technologies that use remote conferencing software to achieve remote meetings at the same time but different locations, this application first obtains the meeting content text, extracts key text words from the meeting content text, segments the meeting content text into target texts based on the key text words, assigns attribute tags to each target text based on the text information of each target text, obtains tagged texts, and pushes each tagged text to the corresponding target text based on the correlation between the text attribute tags of each tagged text and the meeting target. Since the meeting content text is segmented into multiple target texts, different times can be freely selected. The system distributes multiple target texts labeled with tags to participants in different locations. Participants can freely choose when to process their respective meeting content without having to gather together for the meeting. Each participant only needs to focus on the target text labeled with their own tag, without needing to pay attention to the entire meeting content. Furthermore, the correlation between the text attribute tags of the labeled text and the meeting target ensures that each participant can accurately obtain the target text (meeting content) they need to focus on. Therefore, it overcomes the technical shortcomings of remote meetings, where participants may not be able to attend on time for various reasons, or some participants may only be relevant to a part of the meeting content, but these participants are required to participate in the entire meeting, thus improving meeting efficiency. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the first embodiment of the intelligent meeting management optimization method of this application;

[0022] Figure 2 This is a flowchart illustrating the second embodiment of the intelligent meeting management optimization method of this application;

[0023] Figure 3 This is a flowchart illustrating the third embodiment of the intelligent meeting management optimization method of this application;

[0024] Figure 4This is a schematic diagram of the conference system processing logic in the intelligent conference management optimization method of this application;

[0025] Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application.

[0026] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0027] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0028] This application provides an intelligent meeting management optimization method. In the first embodiment of the intelligent meeting management optimization method of this application, refer to... Figure 1 The intelligent meeting management optimization method includes:

[0029] Step S10: Obtain the meeting content text and extract key words from the meeting content text;

[0030] Step S20: Based on each of the aforementioned key words, the meeting content text is segmented into target texts;

[0031] Step S30: Based on the text information of each target text, attribute tags are assigned to each target text to obtain each tag identifier text;

[0032] Step S40: Based on the correlation between the text attribute tags of each of the aforementioned tag identifier texts and the meeting target, push each of the aforementioned tag identifier texts to the corresponding meeting target.

[0033] In this embodiment, it should be noted that the in-meeting management optimization information is applied to the intelligent meeting system. The meeting initiator inputs the meeting content text into the intelligent meeting system, which automatically generates tag identification texts corresponding to the meeting content text and distributes these tag identification texts to their respective target participants. The meeting content text is the main content of the meeting in text form. The text information of the target text can be the semantic information of key words in the target text, or it can be the attribute feature information of the target text. The attribute feature information can be one or more of the following: the industry type involved in the text content, the technology type involved in the text content, the business type involved in the text content, or the personnel involved in the text content.

[0034] Specifically, the system receives the meeting content text input by the meeting initiator, segments the meeting content text into words to obtain individual text words, selects key text words from among the text words based on their importance, segments the meeting content text into target texts based on the position of each key text word in the meeting content text, assigns attribute tags to each target text based on the semantic feature information of each target text, and obtains tagged identification texts with text attribute tags. The system also obtains personnel tags for each participating target, and distributes the tagged identification texts to the corresponding participating targets based on the correlation between personnel tags and text attribute tags.

[0035] In one possible implementation, the importance of the text words is determined by their frequency of occurrence in the meeting content text. The step of selecting key text words from among the text words based on their importance includes:

[0036] The frequency of each of the text words in the meeting content text is obtained, and the text words with a frequency greater than a preset frequency threshold are selected as key text words.

[0037] In another possible implementation, the importance of the text words is represented by their TF-IDF (term frequency–inverse document frequency) characteristic value in the meeting content text. The step of selecting key text words from among the text words based on their importance includes:

[0038] Calculate the TF-IDF feature value of each of the text words in the meeting content text, and select the text words whose TF-IDF feature value is greater than the preset TF-IDF feature threshold as key text words.

[0039] In one possible implementation, the step of distributing the tag identification text to the corresponding participant target based on the correlation between the personnel tag and the text attribute tag includes:

[0040] Calculate the tag similarity between each text attribute tag and each of the aforementioned personnel tags. Based on the tag similarity, determine the text distribution relationship between each tag identifier text and each participant. Based on the text distribution relationship, distribute the tag identifier text to the corresponding participant target.

[0041] The step of determining the text distribution relationship between each tag identifier text and each participant based on the similarity of each tag includes:

[0042] Select the highest label similarity among all the label similarities, and determine the distribution relationship between the label identifier text corresponding to the highest label similarity and the corresponding participant; remove the similarity corresponding to the participant whose distribution relationship has been determined from all the label similarities, and return to the execution step: select the highest label similarity among all the label similarities, and determine the distribution relationship between the label identifier text corresponding to the highest label similarity and the corresponding participant, until all participants have label identifier text with distribution relationship.

[0043] Step S20 includes:

[0044] Step S21: Search for synonyms of the key text words in the meeting content text;

[0045] Step S22: Determine the target text range based on the position of the key words and synonyms in the meeting content text;

[0046] Step S23: Based on the target text range, divide the meeting content text into corresponding target texts.

[0047] In this embodiment, it should be noted that the synonymous text words are text words that have the same semantic meaning as the corresponding key text words.

[0048] Specifically, each preset synonym corresponding to the key text word is obtained; based on each preset synonym, the meeting content text is traversed to obtain the synonym text words of the key text word in the meeting content text; taking the position of the key text word and the synonym text words in the meeting content text as the origin, a preset text area is divided as the target text area; and the content text in the target text area of ​​the meeting content text is taken as the target text.

[0049] In one implementable manner, the step of dividing a preset text region as the target text range, using the positions of the key words and the synonyms in the meeting content text as the origin, includes:

[0050] Using the line positions of the key words and synonyms in the meeting content text as the origin, a preset text area extending forward by a first preset number of lines and backward by a second preset number of lines is defined as the target text area. The first preset number of lines can be set to the number of lines from the origin forward to the paragraph break, or it can be a pre-set fixed number of lines. Similarly, the second preset number of lines can be set to the number of lines from the origin backward to the paragraph break, or it can be a pre-set fixed number of lines. For example, a text area extending 3 lines forward and 3 lines backward from the origin could be defined as the target text area.

[0051] In another feasible embodiment, the step of dividing a preset text region as the target text range, using the positions of the key text words and the synonym text words in the meeting content text as the origin, includes:

[0052] Using the positions of the key words and synonyms in the meeting content text as the origin, a preset text area extending forward by a first preset number of words and backward by a second preset number of words is defined as the target text area. The first preset number of lines can be set to the number of words from the origin forward to the paragraph break, or it can be a pre-set fixed number of words. Similarly, the second preset number of lines can be set to the number of words from the origin backward to the paragraph break, or it can be a pre-set fixed number of words. For example, a text area extending 3 lines forward (100 words) and backward (80 words) from the origin could be defined as the target text area.

[0053] Wherein, the text information includes the semantic information of the key text words of the target text, and step S30 includes:

[0054] Step S31: Perform semantic recognition on each of the key text words to obtain word semantic information;

[0055] Step S32: Based on the mapping relationship between word semantic information and text attribute tags, attribute tags are assigned to the target text corresponding to each key text word to obtain each tag-identified text.

[0056] In this embodiment, it should be noted that the word semantic information includes at least the semantic recognition result of the key text word.

[0057] Specifically, by inputting the word vectors corresponding to each of the key text words into a preset semantic recognition model, semantic recognition is performed on each of the key text words to obtain the semantic recognition results corresponding to each of the key text words; based on the mapping relationship between word semantic information and text attribute labels, the text attribute label corresponding to each semantic recognition result is determined, and the text attribute label corresponding to each semantic recognition result is assigned to the corresponding target text to obtain each of the label identification texts, wherein the label identification text is the target text carrying the text attribute label, and a one-to-one correspondence can be set between the text attribute label, the label identification text, and the semantic recognition result, and the text attribute label can be set as a text attribute label vector. For example, assuming the text attribute label is (a, b, c), where a identifies the person associated with the target text as Zhang San, b identifies the technology associated with the target text as Java technology, and c identifies the key business of the target text as financial business.

[0058] Furthermore, it should be noted that the content of technical research and development meetings varies significantly during the research and development process. Although meeting tags can be generated by drawing on historical experience, for example, by using knowledge images or machine learning models, these tags often do not match the content due to its variability. In contrast, this embodiment typically identifies the semantics of keywords to obtain semantic recognition results. Based on these results, text attribute tags are directly assigned to each target text without relying on historical experience. That is, meeting tags are generated by analyzing the textual word features of the meeting content itself, ensuring the current timeliness of the tags and improving their accuracy. Meetings can then be conducted based on these more accurate tags, thus improving meeting efficiency.

[0059] Furthermore, after step S40, the following is also included:

[0060] The system obtains feedback information from each participating target regarding the received tag identification text. Based on this feedback, it performs a global score on the local meeting to obtain a global score result. This global score result is then pushed to the strong stakeholders of the meeting. The feedback information can be one or more of the following: confirmation immediacy, one-time information completeness, and interactive participation coverage. Confirmation immediacy can be determined by the time it takes for participating targets to confirm the received tag identification text. One-time information completeness can be determined by each participating target scoring the received tag identification text. Interactive participation coverage is determined by detecting whether each participating target provides feedback on the received tag identification text.

[0061] This application provides an intelligent meeting management optimization method. Compared to existing technologies that use remote conferencing software to achieve remote meetings at the same time but different locations, this application first obtains the meeting content text, extracts key text words from the meeting content text, segments the meeting content text into target texts based on the key text words, and then assigns attribute tags to each target text based on the text information of each target text to obtain tagged texts. Based on the correlation between the text attribute tags of each tagged text and the meeting target, each tagged text is pushed to the corresponding meeting target. Since the meeting content text is segmented into multiple target texts, the tagged texts can be pushed to different targets at different times. Multiple target texts labeled with tags are distributed to participants in different locations. Participants can freely choose when to process their respective meeting content without having to gather together for the meeting. Each participant only needs to focus on the target text labeled with their own tag, without needing to pay attention to the entire meeting content. Furthermore, the relevance between the text attribute tags of the labeled text and the meeting target ensures that each participant can accurately obtain the target text (meeting content) they need to focus on. Therefore, it overcomes the technical shortcomings of remote meetings, where participants may not be able to attend on time for various reasons, or some participants may only be relevant to a part of the meeting content, but these participants are required to participate in the entire meeting, thus improving meeting efficiency.

[0062] It is understood that, in this embodiment of the invention, the improved meeting efficiency is not only that the overall meeting process does not need to wait for all personnel to arrive before it can begin, thus improving the punctuality of the meeting start and ensuring the efficient conduct of the meeting, but also that each participant does not need to participate in the entire process, as relevant information will be pushed to them in a timely manner in subsequent meetings, thereby improving the meeting participation efficiency of the participants.

[0063] Furthermore, referring to Figure 2 Based on the first embodiment of this application, in another embodiment of this application, the text information includes the significant attribute features of the target text. Step S30, based on the text information of each target text, assigns attribute tags to each target text to obtain each tag-identified text, further includes:

[0064] Step A10: Extract text attribute features from each of the target texts to obtain the attribute feature set corresponding to each of the target texts;

[0065] In this embodiment, it should be noted that the text attribute feature is a feature that identifies text attributes. The text attribute can be one or more of the following attributes: industry attribute, technical attribute, business attribute, and personnel attribute. The industry attribute is the industry type reflected by the text content of the target text, the technical attribute is the type of technology used reflected by the text content of the target text, the business attribute is the type of business reflected by the text content of the target text, and the industry attribute is the personnel type or personnel name associated with the text content of the target text.

[0066] Specifically, the text keywords corresponding to each target text are classified to obtain the text word classification result for each target text. Based on the text word classification result, the text attribute features corresponding to each target text are determined to obtain the attribute feature set for each target text. The attribute feature set includes at least one text attribute feature.

[0067] In one feasible approach, the step of determining the text attribute features corresponding to each target text based on the text word classification results, and obtaining the attribute feature set corresponding to each target text, includes:

[0068] Based on the mapping relationship between classification labels and text attribute features, the text attribute features corresponding to each classification label in the text word classification results are queried to obtain the attribute feature set corresponding to the target text. For example, assuming that the classification labels are X1, X2, X3, Y1, Y2, Y3 and S1, then the text attribute features are X, Y and S, respectively.

[0069] Step A20: Remove the non-significant attribute features from each of the attribute feature sets to obtain each significant attribute feature set;

[0070] In this embodiment, specifically, a text classification model is constructed using the attribute feature sets corresponding to each target text. Non-significant attribute features are removed from each attribute feature set to obtain significant attribute feature sets. The non-significant attribute features are text attribute features whose contribution to the construction of the text classification model is less than a preset contribution threshold. For example, assuming that each attribute feature set includes a total of two features X and Y, if the text classification model is constructed using feature X, and the classification accuracy of the text classification model for each target text is 20%, and the classification accuracy of the text classification model using features X and Y is 80%, then it is proven that the contribution of feature Y to the text classification model is 75%, and the contribution of feature X to the text classification model is 25%. If the preset contribution threshold is set to 30%, then feature X is a non-significant attribute feature, and feature Y is a significant attribute feature.

[0071] Step A30: Based on the feature value of the significant attribute features in each significant attribute feature set, attribute labels are assigned to the target text corresponding to each significant attribute feature set to obtain the labeled text.

[0072] In this embodiment, it should be noted that the salient attribute feature set includes at least one salient attribute feature.

[0073] Specifically, the following steps are performed for each significant attribute feature set: Based on the classification labels of the text keywords corresponding to each significant attribute feature in the target text, the feature value corresponding to each significant attribute feature is determined; based on the feature value of each significant attribute feature corresponding to the target text, the target text is assigned corresponding text attribute labels to obtain the label identifier text corresponding to the target text. For example, assuming the classification labels of each text keyword in the target text are X1, Y1, and S1, where X represents the industry attribute label, Y represents the technology attribute label, and S represents the financial attribute label, X has 6 possible values ​​representing 6 different industries, Y has 4 possible values ​​representing 4 different technology types, and S has 2 possible values ​​representing whether the developed product is a financial product. Therefore, X corresponds to 6 text attribute labels, Y corresponds to 4 text attribute labels, and S corresponds to 2 text attribute labels. Thus, based on the values ​​of X1, Y1, and S1, the target text can be assigned corresponding text attribute labels.

[0074] Step A20 further includes:

[0075] Step A21: Determine the attribute feature value corresponding to each text attribute feature in each attribute feature set;

[0076] Step A22: Calculate the variance of each attribute feature value corresponding to each text attribute feature to obtain the feature variance corresponding to each text attribute feature;

[0077] Step A23: Based on the variance of each feature, filter out non-significant attribute features from the text attribute features;

[0078] Step A24: Remove the non-significant attribute features from each of the attribute feature sets to obtain each of the significant attribute feature sets.

[0079] In this embodiment, it should be noted that the attribute feature value is the specific text attribute corresponding to the target text. For example, assuming the text attribute feature is an industry attribute feature X, if the corresponding attribute feature value is x1, it means that the target text is associated with the pharmaceutical industry; if the corresponding attribute feature value is x2, it means that the target text is associated with the internet industry. If the target text does not possess a certain text attribute feature, the attribute feature value of the target text under that text attribute feature can be set to 0.

[0080] Specifically, the feature value of each text attribute feature in each of the attribute feature sets is obtained to obtain the attribute feature value corresponding to each target text. Then, the variance of the attribute feature value of each text attribute feature under each target text is calculated to obtain the feature variance corresponding to each text attribute feature. Text attribute features with feature variance less than a preset feature variance threshold are regarded as non-significant features. Non-significant attribute features in each of the attribute feature sets are removed to obtain each of the significant attribute feature sets. It should be noted that the smaller the feature variance, the smaller the difference between each target text on the text attribute feature corresponding to the feature variance. Therefore, the contribution of the text attribute feature corresponding to the feature variance to distinguish each target text is smaller, and the difference in label assignment based on the text attribute feature is smaller.

[0081] This application provides a method for assigning differentiated labels to target texts based on feature variance. Specifically, it involves extracting text attribute features from each target text to obtain an attribute feature set corresponding to each target text, and then determining the attribute feature value corresponding to each text attribute feature in each attribute feature set. The variance of each attribute feature value corresponding to each text attribute feature is calculated to obtain the feature variance corresponding to each text attribute feature. Based on the feature variance, non-significant attribute features are filtered from each text attribute feature. The non-significant attribute features in each attribute feature set are then removed to obtain each significant attribute feature set. This achieves the goal of extracting non-significant attribute features from each text attribute feature by calculating feature variance. The smaller the feature variance, the smaller the difference between target texts on the text attribute feature corresponding to the feature variance. Therefore, the text attribute feature corresponding to the feature variance contributes less to distinguishing each target text. Based on the feature value of the significant attribute feature in each significant attribute feature set, attribute labels are assigned to the target text corresponding to each significant attribute feature set to obtain each labeled text. It enables the assignment of differentiated attribute labels to each target text, thereby improving the accuracy of attribute label assignment to target text.

[0082] Furthermore, referring to Figure 3 Based on the first embodiment of this application, in another embodiment of this application, before step S40, the intelligent meeting management optimization method further includes:

[0083] Step C10: Determine if there is any reviewer's comment content on the tag identifier text;

[0084] Step C20: If it exists, update the text content of the tag identifier text according to the comment content to obtain new target text, and return to the execution step: assign attribute tags to each target text based on the text information of each target text to obtain each tag identifier text.

[0085] In this embodiment, specifically, it is determined whether there is a reviewer's comment on the tagged text. If so, the comment is concatenated with the text content of the tagged text to obtain a new target text, and the process returns to the execution step: assigning attribute tags to each target text based on its text information to obtain each tagged text. If not, the process waits for the execution step: based on the relevance between the text attribute tags of each tagged text and the meeting target, each tagged text is pushed to the corresponding meeting target.

[0086] Step C20 further includes:

[0087] Step C21: Determine whether the reviewer is a strong stakeholder in this meeting;

[0088] Step C22: If the reviewer is the strong stakeholder, then the text content of the tag identifier text is updated according to the comment content to obtain the new target text;

[0089] Step C23: If the reviewer is not a strong stakeholder, calculate the correlation confidence between the reviewer and the tag identification text;

[0090] Step C24: If the relevance confidence level is greater than the preset relevance confidence level threshold, then based on the comment content, update the text content of the tag identifier text to obtain a new target text.

[0091] In this embodiment, it should be noted that the strong stakeholder is a person who has a strong relevance to this meeting. For example, the strong stakeholder may be a leader in the R&D position or the initiator of the meeting. The relevance credibility refers to the credibility of the evaluation content of the reviewer corresponding to the tag identification text.

[0092] Specifically, it is determined whether the reviewer is a strong stakeholder in this meeting. If the reviewer is a strong stakeholder, the reviewer directly concatenates the review content with the text content of the tag identification text to obtain a new target text. If the reviewer is not a strong stakeholder, the correlation confidence between the reviewer and the tag identification text is calculated. If the correlation confidence is greater than a preset correlation confidence threshold, the text content of the tag identification text is updated based on the review content to obtain a new target text. If the correlation confidence is not greater than the preset correlation confidence threshold, the text content of the tag identification text is not updated, and the process waits for the following step: based on the correlation between the text attribute tags of each tag identification text and the meeting target, each tag identification text is pushed to the corresponding meeting target.

[0093] The step of calculating the correlation credibility between the reviewer and the tag identification text includes:

[0094] Based on the text attribute tags corresponding to the tag identifier text, the product module to be developed corresponding to the tag identifier text is determined; a first credibility is generated based on the correlation between the reviewer and the product module to be developed; the historical meeting category tags of the reviewer's historical participation in meetings are obtained, and the current meeting category tag corresponding to the current meeting is obtained; the number of meeting category tags in each historical meeting category tag that are consistent with the current meeting category tag is counted to obtain the number of similar meeting participations; a second credibility is generated based on the number of similar meeting participations; the credibility with the larger value between the first credibility and the second credibility is selected as the relevant credibility. For example, if the reviewer is the R&D leader of the product module to be developed, the first credibility is set to 100%; if the reviewer is the R&D assistant of the product module to be developed, the first credibility is set to 60%; if the number of similar meeting participations is no more than 5, the second credibility is set to 0%; if the number of similar meeting participations is greater than 5 and less than 10, the second credibility is set to 50%; if the number of similar meeting participations is no less than 10, the second credibility is set to 80%. Furthermore, the content of this comment and the tag text can be recorded in a preset knowledge graph as historical data for reference.

[0095] like Figure 4The diagram shown is a schematic of the meeting system processing logic in an embodiment of this application. The main body of the meeting content or the main material of the meeting content is the meeting content text. "Supplement" is a text attribute tag assigned to the target text. "Correction" refers to the process of updating the text content of the tag-identified text based on the comment content. The tag to which the content belongs is a text attribute tag. Module 1, Module 2, Module 3 and Module 4 are the tag-identified texts of each.

[0096] This application provides an intelligent meeting management method based on review comments. Specifically, after obtaining the text of each tag identifier, a reviewer comments on each tag identifier text. When it is determined that there is a reviewer's comment on the tag identifier text, the text content of the tag identifier text is updated online in real time according to the comment content to obtain new target text, and the execution step is returned: based on the text information of each target text, attribute tags are assigned to each target text to obtain each tag identifier text, until the participants with review authority have no objection to the text content, then each tag identifier text is distributed to the corresponding participants, thereby further ensuring the accuracy of the text attribute tags of the tag identifier text. Distributing the tag identifier text based on more accurate text attribute tags can improve the matching degree between the distributed text content and the corresponding participants, reduce the probability of tag identifier text distribution errors, and thus improve meeting efficiency.

[0097] Reference Figure 5 , Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application.

[0098] like Figure 5 As shown, the intelligent meeting management optimization device may include: a processor 1001, such as a CPU, a memory 1005, and a communication bus 1002. The communication bus 1002 is used to establish communication between the processor 1001 and the memory 1005. The memory 1005 may be a high-speed RAM or a stable, non-volatile memory, such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0099] Optionally, the intelligent meeting management optimization device may also include a rectangular user interface, a network interface, a camera, RF (Radio Frequency) circuitry, sensors, hard disk circuitry, a WiFi module, etc. The rectangular user interface may include a display screen and an input submodule such as a keyboard; optionally, the rectangular user interface may also include standard wired or wireless interfaces. The network interface may optionally include standard wired or wireless interfaces (such as a Wi-Fi interface).

[0100] Those skilled in the art will understand that Figure 5 The structure of the intelligent meeting management optimization device shown in the figure does not constitute a limitation on the intelligent meeting management optimization device. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0101] like Figure 5 As shown, the memory 1005, as a computer storage medium, may include an operating system, a network communication module, and an intelligent meeting management and optimization program. The operating system is a program that manages and controls the hardware and software resources of the intelligent meeting management and optimization device, supporting the operation of the intelligent meeting management and optimization program and other software and / or programs. The network communication module is used to enable communication between the various components within the memory 1005, as well as communication with other hardware and software in the intelligent meeting management and optimization system.

[0102] exist Figure 5 In the intelligent meeting management optimization device shown, the processor 1001 is used to execute the intelligent meeting management optimization program stored in the memory 1005 to implement the steps of the intelligent meeting management optimization method described above.

[0103] The specific implementation method of the intelligent meeting management optimization device in this application is basically the same as the embodiments of the intelligent meeting management optimization method described above, and will not be repeated here.

[0104] This application embodiment also provides an intelligent meeting management optimization device, which is applied to an intelligent meeting management optimization equipment, and includes:

[0105] The keyword extraction module is used to obtain the meeting content text and extract key words from the meeting content text;

[0106] The text segmentation module is used to segment the meeting content text into target texts based on the aforementioned key text words;

[0107] The attribute label assignment module is used to assign attribute labels to each of the target texts based on the text information of each target text, so as to obtain the label identifier text;

[0108] The meeting content push module is used to push each labeled text to the corresponding meeting target based on the relevance between the text attribute tags of each labeled text and the meeting target.

[0109] Optionally, the text information includes the semantic information of the key words in the target text, and the attribute tag assignment module is further used for:

[0110] Semantic recognition is performed on each of the key text words to obtain word semantic information;

[0111] Based on the mapping relationship between word semantic information and text attribute tags, attribute tags are assigned to the target text corresponding to each key text word to obtain the tag-identified text.

[0112] Optionally, the text information includes salient attribute features of the target text, and the attribute label assignment module is further used for:

[0113] Text attribute features are extracted from each of the target texts to obtain the attribute feature set corresponding to each of the target texts;

[0114] Each set of attribute features is then removed from the non-significant attribute features to obtain the set of significant attribute features.

[0115] Based on the feature value of the significant attribute features in each of the significant attribute feature sets, attribute labels are assigned to the target text corresponding to each of the significant attribute feature sets to obtain the labeled text.

[0116] Optionally, the attribute label assignment module is further used for:

[0117] Determine the attribute feature value corresponding to each text attribute feature in each of the attribute feature sets;

[0118] Calculate the variance of each attribute feature value corresponding to each text attribute feature to obtain the feature variance corresponding to each text attribute feature;

[0119] Based on the variance of each feature, non-significant attribute features are selected from the text attribute features.

[0120] By removing non-significant attribute features from each of the attribute feature sets, the significant attribute feature sets are obtained.

[0121] Optionally, the text segmentation module is further configured to:

[0122] Search for synonyms of the key words in the meeting content text;

[0123] The target text range is determined based on the positions of the key words and synonyms in the meeting content text;

[0124] Based on the target text range, the corresponding target text is divided into sections within the meeting content text.

[0125] Optionally, the intelligent meeting management optimization device is further used for:

[0126] Determine if there are any comments from reviewers regarding the text identified by the tag;

[0127] If it exists, then based on the comment content, update the text content of the tag identifier text to obtain new target text, and return to the execution step: based on the text information of each target text, assign attribute tags to each target text to obtain each tag identifier text.

[0128] Optionally, the intelligent meeting management optimization device is further used for:

[0129] Determine whether the reviewer is a strong stakeholder in this meeting;

[0130] If the reviewer is the strong stakeholder, then the text content of the tag identifier text is updated based on the review content to obtain the new target text;

[0131] If the reviewer is not a strong stakeholder, then calculate the correlation confidence between the reviewer and the tag identification text;

[0132] If the relevance confidence level is greater than a preset relevance confidence level threshold, then the text content of the tag identifier text is updated based on the comment content to obtain a new target text.

[0133] The specific implementation of the intelligent meeting management optimization device of this application is basically the same as the embodiments of the intelligent meeting management optimization method described above, and will not be repeated here.

[0134] This application provides a computer-readable storage medium that stores one or more programs, which can be executed by one or more processors to implement the steps of the intelligent meeting management optimization method described above.

[0135] The specific implementation of the readable storage medium in this application is basically the same as the embodiments of the above-described intelligent meeting management optimization method, and will not be repeated here.

[0136] This application provides a computer program product, which includes one or more computer programs. The one or more computer programs can be executed by one or more processors to implement the steps of the intelligent meeting management optimization method described above.

[0137] The specific implementation of the computer program product in this application is basically the same as the embodiments of the above-mentioned intelligent meeting management optimization method, and will not be repeated here.

[0138] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. A method for optimizing intelligent meeting management, characterized in that, The intelligent meeting management optimization method includes: Obtain the meeting content text and extract key words from the meeting content text; Based on the aforementioned key words, the meeting content text is segmented into target texts; Based on the text information of each target text, text attribute tags are assigned to each target text to obtain each tag identifier text; Based on the correlation between the text attribute tags of each of the aforementioned tag identifiers and the meeting target, each of the aforementioned tag identifiers is pushed to the corresponding meeting target; The step of segmenting the meeting content text into target texts based on the aforementioned keywords includes: Search for synonyms of the key words in the meeting content text; The target text range is determined based on the positions of the key words and synonyms in the meeting content text; Based on the target text range, the corresponding target text is divided into sections within the meeting content text; The step of pushing each labeled text to the corresponding participant target based on the relevance between the text attribute tags of each labeled text and the participant target includes: obtaining the personnel tags of each participant target, calculating the tag similarity between each text attribute tag and each personnel tag, determining the text distribution relationship between each labeled text and each participant based on the tag similarity, and distributing the labeled text to the corresponding participant target based on the text distribution relationship; wherein, the step of determining the text distribution relationship between each labeled text and each participant based on the tag similarity includes: selecting the highest tag similarity among the tag similarities, determining the distribution relationship between the labeled text corresponding to the highest tag similarity and the corresponding participant; removing the similarity corresponding to the participant whose distribution relationship has been determined from the tag similarities among the tag similarities, and returning to the execution step: selecting the highest tag similarity among the tag similarities, determining the distribution relationship between the labeled text corresponding to the highest tag similarity and the corresponding participant, until each participant has a labeled text with a distribution relationship.

2. The intelligent meeting management optimization method as described in claim 1, characterized in that, The text information includes the semantic information of the key words in the target text. The step of assigning attribute tags to each of the target texts based on their text information to obtain the tagged text includes: Semantic recognition is performed on each of the key text words to obtain word semantic information; Based on the mapping relationship between word semantic information and text attribute tags, attribute tags are assigned to the target text corresponding to each key text word to obtain the tag-identified text.

3. The intelligent meeting management optimization method as described in claim 1, characterized in that, The text information includes the salient attribute features of the target text. The step of assigning attribute tags to each of the target texts based on their text information to obtain the tagged text includes: Text attribute features are extracted from each of the target texts to obtain the attribute feature set corresponding to each of the target texts; Each set of attribute features is then removed from the non-significant attribute features to obtain the set of significant attribute features. Based on the feature value of the significant attribute features in each of the significant attribute feature sets, attribute labels are assigned to the target text corresponding to each of the significant attribute feature sets to obtain the labeled text.

4. The intelligent meeting management optimization method as described in claim 3, characterized in that, The step of removing non-significant attribute features from each of the attribute feature sets to obtain each significant attribute feature set includes: Determine the attribute feature value corresponding to each text attribute feature in each of the attribute feature sets; Calculate the variance of each attribute feature value corresponding to each text attribute feature to obtain the feature variance corresponding to each text attribute feature; Based on the variance of each feature, non-significant attribute features are selected from the text attribute features. By removing non-significant attribute features from each of the attribute feature sets, the significant attribute feature sets are obtained.

5. The intelligent meeting management optimization method as described in any one of claims 1-4, characterized in that, Before the step of pushing each tag identifier text to the corresponding participant target based on the correlation between the text attribute tags of each tag identifier text and the participant target, the intelligent meeting management optimization method further includes: Determine if there are any comments from reviewers regarding the text identified by the tag; If it exists, then based on the comment content, update the text content of the tag identifier text to obtain new target text, and return to the execution step: based on the text information of each target text, assign attribute tags to each target text to obtain each tag identifier text.

6. The intelligent meeting management optimization method as described in claim 5, characterized in that, The step of updating the text content of the tag identifier text based on the comment content to obtain the new target text includes: Determine whether the reviewer is a strong stakeholder in this meeting; If the reviewer is the strong stakeholder, then the text content of the tag identifier text is updated based on the review content to obtain the new target text; If the reviewer is not a strong stakeholder, then calculate the correlation confidence between the reviewer and the tag identification text; If the relevance confidence level is greater than a preset relevance confidence level threshold, then the text content of the tag identifier text is updated based on the comment content to obtain a new target text.

7. An intelligent meeting management optimization device, characterized in that, The intelligent meeting management optimization device includes: The keyword extraction module is used to obtain the meeting content text and extract key words from the meeting content text; The text segmentation module is used to segment the meeting content text into target texts based on the aforementioned key text words; The attribute label assignment module is used to assign attribute labels to each of the target texts based on the text information of each target text, so as to obtain the label identifier text; The meeting content push module is used to push each of the labeled texts to the corresponding meeting target based on the relevance between the text attribute tags of each labeled text and the meeting target; The text segmentation module is specifically used to query the synonyms of the key words in the meeting content text; determine the target text range based on the position of the key words and the synonyms in the meeting content text; and divide the meeting content text into corresponding target text based on the target text range. The meeting content push module is specifically used to obtain personnel tags for each participating target, calculate the tag similarity between each text attribute tag and each personnel tag, determine the text distribution relationship between each tag identifier text and each participating person based on the tag similarity, distribute the tag identifier text to the corresponding participating target based on the text distribution relationship, and select the highest tag similarity among the tag similarities to determine the distribution relationship between the tag identifier text corresponding to the highest tag similarity and the corresponding participating person; remove the similarity corresponding to the participating person whose distribution relationship has been determined from the tag similarities and return to the execution step: select the highest tag similarity among the tag similarities to determine the distribution relationship between the tag identifier text corresponding to the highest tag similarity and the corresponding participating person, until each participating person has a tag identifier text with a distribution relationship.

8. An intelligent meeting management optimization device, characterized in that, The intelligent meeting management optimization device includes: a memory, a processor, and a program stored in the memory for implementing the intelligent meeting management optimization method. The memory is used to store programs that implement intelligent meeting management optimization methods; The processor is used to execute a program that implements the intelligent meeting management optimization method to implement the steps of the intelligent meeting management optimization method as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program for implementing the intelligent meeting management optimization method, which is executed by a processor to implement the steps of the intelligent meeting management optimization method as described in any one of claims 1 to 6.

Images