CN108062955B - Intelligent report generation method, system and equipment - Google Patents
Intelligent report generation method, system and equipment Download PDFInfo
- Publication number
- CN108062955B CN108062955B CN201711318355.5A CN201711318355A CN108062955B CN 108062955 B CN108062955 B CN 108062955B CN 201711318355 A CN201711318355 A CN 201711318355A CN 108062955 B CN108062955 B CN 108062955B
- Authority
- CN
- China
- Prior art keywords
- conversation
- report
- text
- voice
- recording
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000004458 analytical method Methods 0.000 claims abstract description 33
- 230000008569 process Effects 0.000 claims abstract description 21
- 238000004590 computer program Methods 0.000 claims description 8
- 238000003780 insertion Methods 0.000 claims description 6
- 230000037431 insertion Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000012163 sequencing technique Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 239000000463 material Substances 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 230000006399 behavior Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010223 real-time analysis Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B27/00—Editing; Indexing; Addressing; Timing or synchronising; Monitoring; Measuring tape travel
- G11B27/10—Indexing; Addressing; Timing or synchronising; Measuring tape travel
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Telephonic Communication Services (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses an intelligent report generation method, a system and equipment, wherein the method comprises the following steps: recording in the outside meeting process to obtain corresponding meeting voice; performing text analysis on the conversation voice to obtain a corresponding conversation text; and acquiring the report name, and acquiring the conversation report with the name of the report and the content including the conversation text. In the technical scheme disclosed by the embodiment of the invention, the recording is carried out in the process of the external conversation to obtain the corresponding conversation voice, the text analysis is carried out on the conversation voice to obtain the corresponding conversation text, and the conversation report with the report name and the content including the conversation text is finally formed, so that the recording of the conversation content and the generation of the corresponding report are realized through the recording and the text analysis without manually recording the conversation content and generating the corresponding report by a recording person, and the generated report avoids the situations of losing the conversation content or making mistakes in the recording during the manual recording, and can completely and accurately embody the conversation content.
Description
Technical Field
The present invention relates to the field of speech recognition and text analysis technologies, and in particular, to a method, a system, and a device for generating an intelligent report.
Background
In the process of important meeting such as meeting or talking, people often need to record the meeting content and generate a corresponding report for subsequent inquiry and use. At present, a special recording person is usually arranged to record the conversation content and generate a corresponding report, but because the voice speed is higher than the recording speed of the recording person, the situation that the recording person cannot record in time and then loses part of the conversation content and the recording is wrong due to the wrong listening to the conversation content easily occurs, and finally the generated report lacks part of the conversation content and is not accurate enough.
In summary, how to provide a technical solution capable of completely and accurately generating a report indicating conversation contents is an urgent problem to be solved by those skilled in the art.
Disclosure of Invention
The invention aims to provide an intelligent report generation method, an intelligent report generation system and intelligent report generation equipment, so that the generated report can completely and accurately reflect conversation contents.
In order to achieve the above purpose, the invention provides the following technical scheme:
an intelligent report generation method, comprising:
recording in the outside meeting process to obtain corresponding meeting voice;
performing text analysis on the conversation voice to obtain a corresponding conversation text;
and acquiring a report name, and acquiring a session report with the name of the report and the content including the session text.
Preferably, after obtaining the conversation text and before obtaining the conversation report, the method further includes:
dividing the conversation voice and the conversation text into a plurality of sub-conversation voices and sub-conversation texts by taking a preset time interval as a unit, correcting the corresponding sub-conversation texts by using a crowdsourcing mode and manually based on the sub-conversation voices, and combining the corrected sub-conversation texts to obtain the corrected conversation text.
Preferably, after obtaining the modified conversation text, the method further includes:
and performing incremental learning on a model for realizing the conversation voice analysis by taking the conversation voice and the corrected conversation text as training corpora, so as to realize the conversation voice analysis by using the model subjected to the incremental learning.
Preferably, after obtaining the conversation text and before obtaining the conversation report, the method further includes:
importing the conversation text into an editable area;
and determining the associated information which is associated with the session text in the information prestored in the database, and inserting the content selected from the associated information by the outside into the session text under the triggering of the outside.
Preferably, determining the associated information in the information pre-stored in the database and associated with the conversation text includes:
identifying entities, attributes and events of the conversation text to obtain entity tuples containing the entities, the attributes and the events of the conversation text;
and constructing the entity tuples into sub-graphs, inputting the sub-graphs into a knowledge graph pre-stored in a database for sub-graph matching, if the similarity obtained by matching is greater than a preset similarity threshold value, determining that the matching is successful, and sequencing and displaying the successfully matched information in the knowledge graph as associated information according to the similarity from large to small.
Preferably, the displaying the information successfully matched in the knowledge-graph includes:
and displaying the information successfully matched in the knowledge graph in a floating bullet box form.
Preferably, after the content selected from the associated information by the outside is inserted into the conversation text under the outside trigger, the method further includes:
and recording the sub-image matching mode between the conversation text and the externally selected content, the conversation text and the externally selected content to a record set.
Preferably, after obtaining the meeting report with the name of the report and the content including the meeting text, the method further includes:
performing text editing processing on the content included in the conversation report under external triggering;
generating corresponding keywords based on the representation learning of the conversation report and combining a topic model, and adding the keywords into the conversation report;
after the keyword is added to the conversation report, the method further comprises the following steps:
and saving the conversation report as a template for calling the template to be used as the conversation report template.
An intelligent report generating system comprising:
a recording module to: recording in the outside meeting process to obtain corresponding meeting voice;
an analysis module to: performing text analysis on the conversation voice to obtain a corresponding conversation text;
a report generation module to: and acquiring a report name, and acquiring a session report with the name of the report and the content including the session text.
An intelligent report generating device comprising:
a memory for storing a computer program;
a processor for implementing the steps of the intelligent report generation method as claimed in any one of the above when said computer program is executed.
The invention provides an intelligent report generation method, a system and equipment, wherein the method comprises the following steps: recording in the outside meeting process to obtain corresponding meeting voice; performing text analysis on the conversation voice to obtain a corresponding conversation text; and acquiring a report name, and acquiring a session report with the name of the report and the content including the session text. In the technical scheme disclosed by the embodiment of the invention, the recording is carried out in the process of the external conversation to obtain the corresponding conversation voice, the text analysis is carried out on the conversation voice to obtain the corresponding conversation text, and the conversation report with the report name and the content including the conversation text is finally formed, so that the recording of the conversation content and the generation of the corresponding report are realized through the recording and the text analysis without manually recording the conversation content and generating the corresponding report by a recording person, and the generated report avoids the situations of losing the conversation content or making mistakes in the recording during the manual recording, and can completely and accurately embody the conversation content.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of an intelligent report generation method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an intelligent report generation system according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flowchart of an intelligent report generation method according to an embodiment of the present invention is shown, where the method includes:
s11: and recording in the outside meeting process to obtain corresponding meeting voice.
In the external meeting process, the voice generated in the meeting process can be recorded to obtain the corresponding meeting voice, specifically, a system for realizing the method disclosed by the application can be provided with a recording button corresponding to the recording, the initial state of the recording button is 'start recording', the recording is started after the 'start recording' is clicked by the outside, and the recording button is changed into 'pause recording'; and stopping the recording after clicking the 'recording pause', wherein the recording button is changed into 'recording start', and the correspondingly obtained recording is the conversation voice. In addition, the voice can be automatically stored, namely the voice of the meeting can be automatically stored after being obtained; the voice can be manually stored, namely the system can be provided with a 'storage' button, so that the voice obtained by recording is stored by clicking the button from the outside after the recording is started and finished. In addition, the recording in the conversation process can be real-time, namely, as long as the recording is needed, the recording is realized in real time and the conversation voice corresponding to the current moment is obtained.
It should be noted that, in the recording process, the waveform corresponding to the recording may also be displayed by a time axis for external access.
S12: and carrying out text analysis on the conversation voice to obtain a corresponding conversation text.
The system corresponding to the method provided by the invention can also be provided with a voice analysis button, so that when the button is clicked by the outside, the text analysis is carried out on the conversation voice obtained by recording; the invention can also be set as automatic real-time analysis, namely real-time text analysis is carried out on the conversation voice obtained by recording in the recording process, which is within the protection scope of the invention. In addition, an analytic progress bar can be displayed in the process of analyzing the conversation voice, so that external workers can know the progress of voice analysis through the progress bar. It should be noted that, in the present application, if only the conversation voice needs to be saved without any text parsing, the conversation voice can be saved only by clicking a preset "temporary parsing" button or the "save" button from the outside, and all saved voice corresponding lists are updated correspondingly, which is convenient for the query and use of the staff.
In addition, in the process of analyzing the conversation voice in real time, the conversation voice can be divided into different paragraphs or corresponding marks can be made according to the pause of the conversation voice, the pause can be a pause which determines that no person speaks in the recording process and the holding time is not less than the preset pause duration which is preset according to actual needs, namely, the voice corresponding text between two pauses can be a paragraph or a preset pause mark is added at each pause position; the method can also divide paragraphs or make corresponding marks according to the change of the speaking person in the conversation voice, namely, different persons can be distinguished by judging the tone, frequency and the like when the persons speak, so that the voice corresponding text of the same person is used as one paragraph or a preset person-changing mark is added when the persons speak every time, the number of the persons participating in the conversation can be determined according to the tone, frequency and the like when the persons speak, and identification information such as a number or a name is set for each person, so that the identification information of the person is added in front of the corresponding text when the person speaks. Of course, the pauses and the character replacement can also be considered comprehensively, for example, the corresponding text of the speech between every two pauses is used as a paragraph while the corresponding character replacing mark is added when the character speech is replaced each time, and the like, and the invention is within the protection scope of the invention.
S13: and acquiring the report name, and acquiring the conversation report with the name of the report and the content including the conversation text.
It should be noted that the report name may be a dialog box popped up by the user when clicking the "save" button for the external editing of the report name, and a default report name may be preset, such as "XX company"; the report names can also be selected from the report names which are stored in the database in advance through intelligent association, wherein the report names with higher association degree with the current conversation text are the current report names, and the association degree can be that the more times the report names appear in the conversation text, the higher the corresponding association degree is proved, and the like. Through the method, the session report with the report name and the report content is obtained for subsequent use, query and the like.
In the technical scheme disclosed by the embodiment of the invention, the recording is carried out in the process of the external conversation to obtain the corresponding conversation voice, the text analysis is carried out on the conversation voice to obtain the corresponding conversation text, and the conversation report with the report name and the content including the conversation text is finally formed, so that the recording of the conversation content and the generation of the corresponding report are realized through the recording and the text analysis without manually recording the conversation content and generating the corresponding report by a recording person, and the generated report avoids the situations of losing the conversation content or making mistakes in the recording during the manual recording, and can completely and accurately embody the conversation content.
The intelligent report generation method provided by the embodiment of the present invention may further include, after obtaining the conversation text and before obtaining the conversation report:
dividing the conversation voice and the conversation text into a plurality of sub-conversation voices and sub-conversation texts by taking a preset time interval as a unit, correcting the corresponding sub-conversation texts by using a crowdsourcing mode and manually based on the sub-conversation voices, and combining the corrected sub-conversation texts to obtain a corrected conversation text.
The preset time interval can be set according to actual needs, for example, 10 minutes, the conversation voice and the conversation text are divided into a plurality of sub-conversation voices and a plurality of corresponding sub-conversation texts by taking the preset time interval as a unit, wherein the sub-conversation voices and the sub-conversation texts are in one-to-one correspondence, and the time periods of the sub-conversation voices and the sub-conversation texts which are corresponding to each other are the same. The corresponding sub-conversation text is corrected manually based on the sub-conversation voice by using a crowdsourcing mode, so that the accuracy of the corrected conversation text is ensured, and important information leakage is effectively avoided due to the crowdsourcing mode. Specifically, modifying the corresponding sub-conversation text with a human based on the sub-conversation speech using a crowdsourcing approach may include: and sending the obtained sub-conversation voice and the corresponding sub-conversation text to different terminals (each sub-conversation voice and the corresponding sub-conversation text correspond to one terminal), wherein the corresponding staff of each terminal manually intervenes and corrects the received sub-conversation text based on the received sub-conversation voice, for example, whether the content of the received sub-conversation record is the same as the content of the corresponding sub-conversation text is manually listened back, and the content of the sub-conversation record is taken as the criterion for correcting different places, so that a record of correction can be kept on the sub-conversation text or a process of directly recording the correction in another storage space is carried out.
The method for generating an intelligent report according to the embodiment of the present invention may further include, after obtaining the modified conversation text:
and performing incremental learning on the model for realizing the conversation voice analysis by taking the conversation voice and the corrected conversation text as training corpora, so as to realize the conversation voice analysis by using the model after the incremental learning. Because the corrected conversation text is more accurate compared with the conversation text before correction and more accords with the corresponding conversation voice, the conversation voice and the corrected conversation text can be used as training corpora to carry out incremental learning on the model for realizing voice analysis, so that the model is realized through the incremental learning when the voice analysis is required to be realized subsequently, and the accuracy of automatic voice analysis is improved.
The intelligent report generation method provided by the embodiment of the present invention may further include, after obtaining the conversation text and before obtaining the conversation report:
importing the conversation text into an editable area;
and determining the associated information which is associated with the conversation text in the information prestored in the database, and inserting the content selected from the associated information by the outside into the conversation text under the triggering of the outside.
The method disclosed by the application can be used for setting an editable area such as a note bar in a corresponding system, importing the corresponding text into the editable area (specifically, the end of the editable area) when the whole or a part of the conversation text is clicked from the outside, and automatically importing the text into the editable area after the conversation text is generated, which is within the protection scope of the invention. And the encodable area is used as a text editing tool and has the text editing functions of basic fonts, paragraphs, colors, labels and the like. The preset database stores corresponding information, the information in the database selected from the outside can be inserted into the position appointed by the outside in the conference text under the trigger of the outside, and particularly, the selection, the insertion and the like of the information needing to be inserted can be realized by clicking an 'insertion button' from the outside; and the associated information which is associated with the conversation text in the database can be automatically screened out so as to be checked by the outside or the information which needs to be inserted into the conversation text is selected from the database, and the information is inserted into the conversation text under the trigger of the outside, and if the section of voice comprises the current stockholder of the company, the information corresponding to the stockholder can be used as the associated information to be selected by the outside. Thereby speeding up the accurate generation of reports by such intelligent association prompt functionality.
The method for generating an intelligent report according to the embodiments of the present invention determines, in information pre-stored in a database, associated information associated with a conversation text, and may include:
identifying entities, attributes and events of the session text to obtain entity tuples containing the entities, the attributes and the events of the session text;
and constructing entity tuples into subgraphs, inputting the subgraphs into a knowledge graph pre-stored in a database for subgraph matching, if the similarity obtained by matching is greater than a preset similarity threshold value, determining that the matching is successful, and sequencing and displaying the successfully matched information in the knowledge graph as associated information according to the similarity from large to small.
The method comprises the steps of carrying out dynamic real-time analysis on a conversation text, including word segmentation, part of speech recognition, named entity recognition and syntactic dependency analysis on the conversation text, so as to obtain an entity and a component structure contained in the text, wherein the component structure comprises attributes and events of the entity in the conversation text, and specifically, the entity in the application is defined as an independently-existing participant of financial security activities, including enterprises, investors and the like; the entity attributes are static expressions of the entity elements, and are described by taking an enterprise as an example, the attributes of the enterprise comprise business information, industry attributes, associated enterprises, individuals, key agents, financial attributes and the like, wherein the key agent attributes of the enterprise comprise actual control persons, and the actual control persons can comprise stockholders of the enterprise and persons who are not the stockholders of the enterprise but can govern the enterprise behavior through investment relations, protocols or other arrangements; the event is an event corresponding to the entity, which is described by taking an enterprise as an example, the event of the entity is a set of dynamic behaviors of the enterprise, including financing, business activities, information disclosure and the like, and the event itself also has corresponding attributes, such as time, place, people, event origin, event passing, result and influence of the event; thus, by identifying the entity, the attribute, the concept and the event contained in the session text, the tuple form containing the entity, the attribute and the event in the session text, namely the entity tuple is obtained; inputting sub-graphs constructed by entity tuples into a knowledge graph stored in a database for sub-graph matching, if the matching similarity exceeds a preset similarity threshold which is preset according to actual needs, determining that the content is matched, and further sequencing and displaying the matched content serving as associated information according to the similarity from large to small, wherein a plurality of sub-graphs are stored in the knowledge graph, sequentially comparing the sub-graphs of the entity tuples with each sub-graph in the knowledge graph, and if the similarity obtained by comparison is greater than the preset similarity threshold, indicating that the matching is successful. Thereby facilitating the external checking and use. It should be noted that, generally, fuzzy matching of subgraphs may be performed first when performing subgraph matching, so that a plurality of candidate subgraphs corresponding to the subgraph corresponding to the conversation text may be generated, then, the similarity between the subgraph corresponding to the conversation text and each node and attribute in each candidate subgraph is determined, and the subgraph candidate subgraph with the highest similarity and higher than the preset similarity threshold is determined as the final subgraph successfully matched.
The method for generating an intelligent report, provided by the embodiment of the present invention, displays information successfully matched in a knowledge graph, and may include:
and displaying the information successfully matched in the knowledge graph in a floating bullet box form.
The display of the information which is successfully matched can be achieved in a floating pop-up box mode, particularly, after the matching is determined to be successful, the entity part in the talking text can be highlighted, the information which needs to be displayed is displayed in the floating pop-up box mode, the floating pop-up box can be in a closing mode by default, and the floating pop-up box is displayed when the highlight entity is selected outside (such as a mouse is moved to the highlight entity). The content of the floating popup box can be closed under the external trigger after being browsed by a user, and the information can also be added into the conversation text after any information is selected from the outside. Therefore, the method and the device are convenient for the user to check and insert the related information, and can avoid excessive occupation of page resources.
The method for generating an intelligent report provided by the embodiment of the present invention may further include, after inserting the content selected from the associated information by the outside into the conversation text under the outside trigger:
recording the sub-image matching mode between the conversation text and the externally selected content, the conversation text and the externally selected content into a record set.
After the content needing to be added into the session text is selected by the outside, the session text, the externally selected content and the corresponding subgraph matching mode can be recorded into a record set, wherein the externally selected content and the subgraph matching mode between the session texts can comprise a subgraph corresponding to the externally selected content, a subgraph corresponding to the session text and the similarity between the two subgraphs; therefore, when a certain conversation text needs to be matched again, the subgraph matching mode corresponding to the conversation text can be found from the record set, if a plurality of found subgraph matching modes are provided, the externally selected content corresponding to the subgraph matching mode with the highest similarity is selected as the content corresponding to the conversation text and is returned to the external user in a display mode and the like. The finding of the sub-graph matching pattern may be implemented by related representation learning, similarity calculation, classification method, and the like in deep learning, which is not specifically limited herein.
The method for generating an intelligent report according to the embodiments of the present invention may further include, after obtaining a conversation report whose name is a report name and whose content includes a conversation text:
performing text editing processing on the content included in the conversation report under the external trigger;
corresponding keywords are generated based on the representation learning of the conversation report and combined with the topic model, and the keywords are added into the conversation report.
After the meeting report is obtained, the meeting report supports basic text editing capability, and can carry out text editing processing such as adding, deleting, adjusting various formats and the like, so that a user can finally adjust the meeting report into a report form required by the user. In addition, corresponding keywords, such as dimensions of industries, areas and the like, can be generated by combining the presentation learning of report names and contents of the conversation reports with the topic model, and the keywords are added into the conversation reports, so that the users can conveniently search and classify the reports based on the keywords.
The method for generating an intelligent report according to the embodiment of the present invention may further include, after adding the keyword to the conversation report:
the conversation report is saved as a template for calling the template to be used as the conversation report template.
The conversation report is saved as a template, and the system can also save one or more common standard templates, so that any saved template can be called when the conversation report needs to be generated, the content is sufficient, and a new report can be quickly formed. The stored template of the session report only includes the format of the session report, such as typesetting, directory structure, etc., and does not include the substantial content of the session text, etc., so as to form a new session report by putting other content in the template.
In addition, notes or texts input from the outside can be received and stored, so that when the contents need to be inserted into the conversation text, the outside can select corresponding contents from the stored notes or texts to realize the insertion.
According to the method and the system, a user is supported to import notes or texts to serve as candidate materials for generating reports, the system establishes corresponding relations among the catalogues in the template, the catalogues in the candidate materials, the paragraphs in the template and the paragraphs in the candidate materials according to the stored similarity of the catalogues in the template and the catalogues in the candidate materials, so that the catalogues and the paragraphs of the candidate materials are added to corresponding positions in the template, the report generation is automatically carried out, and the format is intelligently adjusted.
An embodiment of the present invention further provides an intelligent report generating system, as shown in fig. 2, which may include:
a recording module 11, configured to: recording in the outside meeting process to obtain corresponding meeting voice;
a parsing module 12 for: performing text analysis on the conversation voice to obtain a corresponding conversation text;
a report generation module 13, configured to: and acquiring the report name, and acquiring the conversation report with the name of the report and the content including the conversation text.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
a correction module to: after the conversation text is obtained and before the conversation report is obtained, the conversation voice and the conversation text are divided into a plurality of sub-conversation voices and sub-conversation texts by taking a preset time interval as a unit, the corresponding sub-conversation texts are corrected manually based on the sub-conversation voices in a crowdsourcing mode, and the corrected sub-conversation texts are combined to obtain a corrected conversation text.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
a learning module to: after the corrected conversation text is obtained, incremental learning is carried out on the model for realizing conversation voice analysis by taking the conversation voice and the corrected conversation text as training corpora, so that the model for realizing the incremental learning is used for realizing the conversation voice analysis.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
an insertion module to: after obtaining the conversation text and before obtaining the conversation report, importing the conversation text into an editable area; and determining the associated information which is associated with the conversation text in the information prestored in the database, and inserting the content selected from the associated information by the outside into the conversation text under the triggering of the outside.
In the system for generating an intelligent report provided in an embodiment of the present invention, the insertion module may include:
a matching unit for: identifying entities, attributes and events of the session text to obtain entity tuples containing the entities, the attributes and the events of the session text; and constructing entity tuples into subgraphs, inputting the subgraphs into a knowledge graph pre-stored in a database for subgraph matching, if the similarity obtained by matching is greater than a preset similarity threshold value, determining that the matching is successful, and sequencing and displaying the successfully matched information in the knowledge graph as associated information according to the similarity from large to small.
In the system for generating an intelligent report provided in an embodiment of the present invention, the matching unit may include:
a display subunit to: and displaying the information successfully matched in the knowledge graph in a floating bullet box form.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
a recording module to: after the content selected from the associated information by the outside is inserted into the conversation text under the triggering of the outside, a subgraph matching mode between the conversation text and the content selected by the outside, the conversation text and the content selected by the outside are recorded into a record set.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
an editing module to: after obtaining the meeting report with the name of the report and the content including the meeting text, performing text editing processing on the content included in the meeting report under the external trigger;
a keyword extraction module to: after a conversation report with the name of the report and the content including conversation text is obtained, corresponding keywords are generated based on the representation learning of the conversation report and the topic model, and the keywords are added into the conversation report.
The system for generating an intelligent report provided by the embodiment of the present invention may further include:
a save module to: after the keywords are added to the conversation report, the conversation report is saved as a template for calling the template to be used as the conversation report template.
An embodiment of the present invention further provides an intelligent report generating device, which may include:
a memory for storing a computer program;
a processor for implementing the steps of the intelligent report generation method as claimed in any one of the above when said computer program is executed.
An embodiment of the present invention further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the intelligent report generation method described in any one of the above may be implemented.
For a description of a relevant part in an intelligent report generation system, an intelligent report generation device, and a computer-readable storage medium provided in the embodiments of the present invention, reference is made to detailed descriptions of a corresponding part in an intelligent report generation method provided in the embodiments of the present invention, and details are not described herein again. In addition, parts of the technical solutions provided in the embodiments of the present invention that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail, so as to avoid redundant description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (9)
1. An intelligent report generation method, comprising:
recording in the outside meeting process to obtain corresponding meeting voice;
performing text analysis on the conversation voice to obtain a corresponding conversation text;
acquiring a report name, and acquiring a conversation report with the name of the report and the content including the conversation text;
obtaining a report name, comprising:
selecting a report name with the association degree reaching the corresponding high requirement with the conversation text from report names prestored in a database as the report name of the conversation text; the more the message names stored in the database appear in the conversation text, the higher the corresponding association degree is;
after obtaining the conversation text and before obtaining the conversation report, the method further includes:
importing the conversation text into an editable area;
and determining the associated information which is associated with the session text in the information prestored in the database, and inserting the content selected from the associated information by the outside into the session text under the triggering of the outside.
2. The method of claim 1, wherein obtaining the conversation text and obtaining the conversation report further comprise:
dividing the conversation voice and the conversation text into a plurality of sub-conversation voices and sub-conversation texts by taking a preset time interval as a unit, correcting the corresponding sub-conversation texts by using a crowdsourcing mode and manually based on the sub-conversation voices, and combining the corrected sub-conversation texts to obtain the corrected conversation text.
3. The method of claim 2, wherein after obtaining the modified conversation text, further comprising:
and performing incremental learning on a model for realizing the conversation voice analysis by taking the conversation voice and the corrected conversation text as training corpora, so as to realize the conversation voice analysis by using the model subjected to the incremental learning.
4. The method of claim 1, wherein determining the associated information in the pre-stored information in the database that has an association with the conversation text comprises:
identifying entities, attributes and events of the conversation text to obtain entity tuples containing the entities, the attributes and the events of the conversation text;
and constructing the entity tuples into sub-graphs, inputting the sub-graphs into a knowledge graph pre-stored in a database for sub-graph matching, if the similarity obtained by matching is greater than a preset similarity threshold value, determining that the matching is successful, and sequencing and displaying the successfully matched information in the knowledge graph as associated information according to the similarity from large to small.
5. The method of claim 4, wherein displaying the information that the matching in the knowledge-graph was successful comprises:
and displaying the information successfully matched in the knowledge graph in a floating bullet box form.
6. The method of claim 4, wherein after inserting the external content selected from the associated information into the conversation text under the external trigger, the method further comprises:
and recording the sub-image matching mode between the conversation text and the externally selected content, the conversation text and the externally selected content to a record set.
7. The method of claim 1, wherein after obtaining the conversation report with the report name and the conversation text, the method further comprises:
performing text editing processing on the content included in the conversation report under external triggering;
generating corresponding keywords based on the representation learning of the conversation report and combining a topic model, and adding the keywords into the conversation report;
after the keyword is added to the conversation report, the method further comprises the following steps:
and saving the conversation report as a template for calling the template to be used as the conversation report template.
8. An intelligent report generation system, comprising:
a recording module to: recording in the outside meeting process to obtain corresponding meeting voice;
an analysis module to: performing text analysis on the conversation voice to obtain a corresponding conversation text;
a report generation module to: acquiring a report name, and acquiring a conversation report with the name of the report and the content including the conversation text;
the report generating module acquires a report name, and comprises:
selecting a report name with the association degree reaching the corresponding high requirement with the conversation text from report names prestored in a database as the report name of the conversation text; the more the message names stored in the database appear in the conversation text, the higher the corresponding association degree is;
the system further comprises:
an insertion module to: after obtaining the conversation text and before obtaining the conversation report, importing the conversation text into an editable area; and determining the associated information which is associated with the session text in the information prestored in the database, and inserting the content selected from the associated information by the outside into the session text under the triggering of the outside.
9. An intelligent report generating device, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the intelligent report generation method according to any of claims 1 to 7 when executing said computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711318355.5A CN108062955B (en) | 2017-12-12 | 2017-12-12 | Intelligent report generation method, system and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711318355.5A CN108062955B (en) | 2017-12-12 | 2017-12-12 | Intelligent report generation method, system and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108062955A CN108062955A (en) | 2018-05-22 |
| CN108062955B true CN108062955B (en) | 2020-12-11 |
Family
ID=62138309
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711318355.5A Active CN108062955B (en) | 2017-12-12 | 2017-12-12 | Intelligent report generation method, system and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108062955B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766433B (en) * | 2018-12-06 | 2023-05-09 | 湖南科创信息技术股份有限公司 | Automatic cataloging method and system and computer readable storage medium |
| CN109918628A (en) * | 2019-01-17 | 2019-06-21 | 平安科技（深圳）有限公司 | Report-generating method, device, equipment and storage medium based on speech recognition |
| CN110457424A (en) * | 2019-06-28 | 2019-11-15 | 谭浩 | Generate method, computer readable storage medium and the terminal device of interview report |
| CN110543559A (en) * | 2019-06-28 | 2019-12-06 | 谭浩 | Method for generating interview report, computer-readable storage medium and terminal device |
| CN110751950A (en) * | 2019-10-25 | 2020-02-04 | 武汉森哲地球空间信息技术有限公司 | Police conversation voice recognition method and system based on big data |
| CN111368515B (en) * | 2020-03-02 | 2021-01-26 | 中国农业科学院农业信息研究所 | Industry dynamic interactive report generation method and system based on PDF document fragmentation |
| CN111507758B (en) * | 2020-04-09 | 2023-10-31 | 深圳传世智慧科技有限公司 | Investigation method, device, system and server based on semantic analysis |
| CN112862431A (en) * | 2021-01-19 | 2021-05-28 | 中国建设银行股份有限公司 | Method and device for generating conversation record, electronic equipment and readable storage medium |
| CN113421572B (en) * | 2021-06-23 | 2024-02-02 | 平安科技（深圳）有限公司 | Real-time audio dialogue report generation method and device, electronic equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106057193A (en) * | 2016-07-13 | 2016-10-26 | 深圳市沃特沃德股份有限公司 | Conference record generation method based on telephone conference and device |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6415929B2 (en) * | 2014-10-30 | 2018-10-31 | 株式会社東芝 | Speech synthesis apparatus, speech synthesis method and program |
| CN105810206A (en) * | 2014-12-30 | 2016-07-27 | 富泰华工业（深圳）有限公司 | Meeting recording device and method thereof for automatically generating meeting record |
| CN105810207A (en) * | 2014-12-30 | 2016-07-27 | 富泰华工业（深圳）有限公司 | Meeting recording device and method thereof for automatically generating meeting record |
| US10460030B2 (en) * | 2015-08-13 | 2019-10-29 | International Business Machines Corporation | Generating structured meeting reports through semantic correlation of unstructured voice and text data |
| CN105845129A (en) * | 2016-03-25 | 2016-08-10 | 乐视控股（北京）有限公司 | Method and system for dividing sentences in audio and automatic caption generation method and system for video files |
| CN106847256A (en) * | 2016-12-27 | 2017-06-13 | 苏州帷幄投资管理有限公司 | A kind of voice converts chat method |
| CN107147564A (en) * | 2017-05-09 | 2017-09-08 | 胡巨鹏 | Real-time speech recognition error correction system and identification error correction method based on cloud server |
-
2017
- 2017-12-12 CN CN201711318355.5A patent/CN108062955B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106057193A (en) * | 2016-07-13 | 2016-10-26 | 深圳市沃特沃德股份有限公司 | Conference record generation method based on telephone conference and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108062955A (en) | 2018-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108062955B (en) | Intelligent report generation method, system and equipment | |
| US11455475B2 (en) | Human-to-human conversation analysis | |
| US9014363B2 (en) | System and method for automatically generating adaptive interaction logs from customer interaction text | |
| CN111666746B (en) | Conference summary generation method and device, electronic equipment and storage medium | |
| US8527263B2 (en) | Method and system for automatically generating reminders in response to detecting key terms within a communication | |
| US20170169822A1 (en) | Dialog text summarization device and method | |
| JP4215792B2 (en) | CONFERENCE SUPPORT DEVICE, CONFERENCE SUPPORT METHOD, AND CONFERENCE SUPPORT PROGRAM | |
| CN110493019B (en) | Automatic generation method, device, equipment and storage medium of conference summary | |
| US20190384813A1 (en) | Generating summaries and insights from meeting recordings | |
| US10613825B2 (en) | Providing electronic text recommendations to a user based on what is discussed during a meeting | |
| US20200137224A1 (en) | Comprehensive log derivation using a cognitive system | |
| JP5212604B2 (en) | Risk detection system, risk detection method and program thereof | |
| US9697246B1 (en) | Themes surfacing for communication data analysis | |
| US20130253932A1 (en) | Conversation supporting device, conversation supporting method and conversation supporting program | |
| CA3115974C (en) | Presentation assistance device for calling attention to words that are forbidden to speak | |
| US20190384854A1 (en) | Generating summaries and insights from meeting recordings | |
| US20160189103A1 (en) | Apparatus and method for automatically creating and recording minutes of meeting | |
| CN111063355A (en) | Conference record generation method and recording terminal | |
| US20190295199A1 (en) | Intelligent legal simulator | |
| WO2017041377A1 (en) | Method and device for generating calendar reminding information | |
| CN111222837A (en) | Intelligent interviewing method, system, equipment and computer storage medium | |
| US20060129398A1 (en) | Method and system for obtaining personal aliases through voice recognition | |
| US20200320134A1 (en) | Systems and methods for generating responses for an intelligent virtual | |
| US20230163988A1 (en) | Computer-implemented system and method for providing an artificial intelligence powered digital meeting assistant | |
| CN116741178A (en) | Manuscript generation method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |