CN112115252B

CN112115252B - Intelligent auxiliary writing processing method and device, electronic equipment and storage medium

Info

Publication number: CN112115252B
Application number: CN202010871570.3A
Authority: CN
Inventors: 罗彤; 张学彬; 闻飞祥; 王雨农
Original assignee: Beijing Ronghui Jinxin Information Technology Co ltd
Current assignee: Beijing Ronghui Jinxin Information Technology Co ltd
Priority date: 2020-08-26
Filing date: 2020-08-26
Publication date: 2023-06-02
Anticipated expiration: 2040-08-26
Also published as: CN112115252A

Abstract

The embodiment of the invention discloses an intelligent auxiliary writing processing method, device, electronic equipment and storage medium, which are characterized in that query intention recognition is carried out on writing contents needing quick query auxiliary data, query intention recognition results of the writing contents needing quick query auxiliary data are obtained, further, knowledge graphs are queried according to query intention category recognition results and entity dependency relationship recognition results, auxiliary data needing quick query corresponding to the writing contents are obtained, the auxiliary data needing quick query corresponding to the writing contents are displayed in a designated area of a current writing page, and therefore, the embodiment of the invention can provide intelligent writing assistance in the writing process of a user, analyze and obtain query intention of the user about the writing auxiliary data from the writing contents, and accordingly, the corresponding writing auxiliary data can be obtained from related knowledge graphs according to the query intention, so that the writing efficiency can be effectively improved, and intelligent writing is possible.

Description

Intelligent auxiliary writing processing method and device, electronic equipment and storage medium

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to an intelligent auxiliary writing processing method, an intelligent auxiliary writing processing device, electronic equipment and a storage medium.

Background

Artificial intelligence (Artificial Intelligence, AI for short), which is a new technical science to study, develop theories, methods, techniques and application systems for simulating, extending and expanding human intelligence. Artificial intelligence is a branch of computer science that attempts to understand the nature of intelligence and to produce a new intelligent machine that can react in a similar manner to human intelligence, research in this field including robotics, language recognition, image recognition, natural language processing, and expert systems.

The traditional writing method is mainly characterized in that the data and knowledge are queried and retrieved manually, and the data such as tables, pictures and the like are repeatedly drawn and organized, so that the process is often tedious and inefficient. With the increasing development of deep learning, high-quality writing scenes (macro economy, macro strategy, industry, analysis report and weekly report of companies, financial report comment of companies, government report, academic report and the like) are more dependent on a large amount of data and knowledge, and the whole process of questions, gathering materials, writing, matching, auditing and publishing is performed, so that the production efficiency of articles and reports is low, the quality can be guaranteed, and a plurality of people are difficult to release from heavy labor.

Disclosure of Invention

Because the existing method has the problems, the embodiment of the invention provides an intelligent auxiliary writing processing method, an intelligent auxiliary writing processing device, electronic equipment and a storage medium.

In a first aspect, an embodiment of the present invention provides an intelligent assisted writing processing method, including:

receiving writing contents which are triggered by a user in the writing process of the current writing page and need to quickly inquire auxiliary data through a preset triggering mode;

carrying out query intention recognition on the writing content of the auxiliary data needing to be queried quickly, and obtaining a query intention recognition result of the writing content of the auxiliary data needing to be queried quickly; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities;

acquiring auxiliary data which corresponds to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph;

and displaying auxiliary data which are required to be quickly queried and correspond to the writing content in a designated area of the current writing page so as to be selected and used by a user for the auxiliary data in the writing process.

Further, performing query intention recognition on the writing content requiring the shortcut query auxiliary data, and obtaining a query intention recognition result of the writing content requiring the shortcut query auxiliary data includes:

carrying out query intention type recognition on the writing content needing the shortcut query auxiliary data to obtain a query intention type recognition result;

performing entity identification and entity dependency relationship identification on the writing content needing shortcut query auxiliary data to acquire an entity identification result and an entity dependency relationship identification result; wherein, the entity identification result at least comprises: object entity, time entity and index entity; the entity dependency relationship identification result at least comprises: a combined constraint between the object entity, the time entity, and the index entity.

Further, performing query intention type recognition on the writing content requiring the shortcut query auxiliary data to obtain a query intention type recognition result, including:

inputting the writing content requiring the shortcut inquiry auxiliary data into a preset intention type identification model to acquire an inquiry intention type identification result;

the preset intention type recognition model is obtained by taking a writing content training sample of each known query intention type as sample input data, taking a query intention type corresponding to writing content of each known query intention type as sample output data, and carrying out model training based on a machine learning algorithm; the query intention type recognition result is used for representing the domain branch corresponding to the writing content needing the shortcut query auxiliary data.

Further, performing entity identification and entity dependency relationship identification on the writing content requiring shortcut query auxiliary data, and obtaining an entity identification result and an entity dependency relationship identification result, including:

and carrying out sentence segmentation, word segmentation, part-of-speech tagging and dependency syntactic analysis on the written content needing the quick query auxiliary data to obtain an entity identification result and an entity dependency relationship identification result.

Further, according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph, auxiliary data which corresponds to the writing content and needs to be queried quickly is obtained, and the method comprises the following steps:

determining a knowledge graph of the branch in the corresponding field according to the query intention category recognition result;

acquiring auxiliary data which corresponds to the writing content and needs to be queried rapidly from the knowledge graph of the corresponding field branch according to the entity dependency relationship identification result;

the knowledge graph of the corresponding domain branch stores knowledge and time sequence data of the corresponding domain branch, wherein the knowledge and time sequence data comprises index information of each object at each time in the corresponding domain branch.

determining an extended entity dependency relationship identification result corresponding to the entity dependency relationship identification result according to the entity dependency relationship identification result; wherein, the object entity in the extended entity dependency relationship identification result and the object entity in the entity dependency relationship identification result have a class relationship or a bid relationship; the time entity in the extended entity dependency relationship identification result has an adjacent relationship or a periodic relationship with the time entity in the entity dependency relationship identification result; the index entity in the extended entity dependency relationship identification result has a similar relationship or an opposite relationship with the index entity in the entity dependency relationship identification result;

according to the entity dependency relationship identification result, acquiring accurate auxiliary data which corresponds to the writing content and needs to be quickly queried from the knowledge graph of the corresponding field branch; acquiring relevant auxiliary data which are corresponding to the writing content and need to be quickly queried from the knowledge graph of the corresponding field branch according to the identification result of the extended entity dependency relationship;

Further, the auxiliary data which is required to be quickly queried and corresponds to the writing content is displayed in a designated area of the current writing page so as to be used by a user for selecting the auxiliary data in the writing process, and the method comprises the following steps:

performing rendering processing of different forms on auxiliary data which are corresponding to the written content and need to be quickly queried, and obtaining a rendering result of a text, a table, a chart or rich media of the auxiliary data;

and displaying the rendering results of the text, the table, the chart or the rich media of the auxiliary data in the appointed area of the current writing page so as to be selected and used by a user for the rendering results in the writing process.

In a second aspect, an embodiment of the present invention further provides an intelligent auxiliary writing processing apparatus, including:

the receiving module is used for receiving the writing content which is triggered and sent by a user and needs to quickly inquire auxiliary data in the writing process of the current writing page in a preset triggering mode;

The first acquisition module is used for carrying out query intention recognition on the written content of the auxiliary data needing to be queried quickly and acquiring a query intention recognition result of the written content of the auxiliary data needing to be queried quickly; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities;

the second acquisition module is used for acquiring auxiliary data which is corresponding to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph;

and the display module is used for displaying the auxiliary data which is required to be quickly queried and corresponds to the writing content in the appointed area of the current writing page so as to be selected and used by a user for the auxiliary data in the writing process.

In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements the intelligent aided writing processing method according to the first aspect when executing the computer program.

In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the intelligent aided writing processing method according to the first aspect.

According to the technical scheme, the intelligent auxiliary writing processing method, the device, the electronic equipment and the storage medium provided by the embodiment of the invention are capable of obtaining the query intention recognition result of the writing content requiring the quick query auxiliary data by carrying out query intention recognition on the writing content requiring the quick query auxiliary data, further inquiring the knowledge graph according to the query intention type recognition result and the entity dependency relationship recognition result, obtaining the auxiliary data requiring the quick query corresponding to the writing content, and displaying the auxiliary data requiring the quick query corresponding to the writing content in the appointed area of the current writing page, so that the user can select and use the auxiliary data in the writing process, and therefore, the embodiment of the invention can provide intelligent writing assistance in the user writing process, quickly provide the auxiliary data requiring the query in the user writing process, so that the process of searching the auxiliary data by switching a search platform in the writing process can be omitted, the embodiment of the invention is capable of analyzing the auxiliary data requiring the quick query of the user corresponding to the writing content, thus obtaining the relevant auxiliary data from the writing content, and effectively inquiring the corresponding writing intention and improving the intelligent writing intention. Therefore, according to the intelligent auxiliary writing technical scheme provided by the embodiment of the invention, in the writing process, only the writing content needing to be quickly queried is required to be triggered, and the auxiliary data which is corresponding to the writing content and needs to be quickly queried can be provided for a user, so that the user can directly utilize the corresponding auxiliary data to assist the writing process, and the writing efficiency and the writing effect can be improved.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings that are necessary for the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention and that other drawings can be obtained from these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of an intelligent aided writing processing method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a authoring platform architecture according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of an implementation process of an intelligent aided writing processing method according to an embodiment of the present invention;

fig. 4, fig. 5, fig. 6, fig. 7, fig. 8, and fig. 9 are schematic diagrams of answer query recommendation results corresponding to the writing content, which are obtained according to the query intention recognition result according to the embodiment of the present invention;

FIG. 10 is a system block diagram of an artificial intelligence based intelligent authoring assistant in accordance with one embodiment of the present invention;

FIG. 11 is a schematic diagram of steps of a human-computer interaction method based on an intelligent authoring assistant according to an embodiment of the present invention;

FIG. 12 is a schematic diagram of an intelligent auxiliary authoring processing apparatus according to an embodiment of the present invention;

Fig. 13 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The following describes the embodiments of the present invention further with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.

Fig. 1 shows a flowchart of an intelligent auxiliary writing processing method according to an embodiment of the present invention, and it should be noted that the intelligent auxiliary writing processing method according to the embodiment of the present invention may be implemented based on the writing platform architecture shown in fig. 2. The intelligent auxiliary writing processing method provided by the embodiment of the invention is explained and illustrated in detail below with reference to fig. 1. As shown in fig. 1, the method for processing intelligent auxiliary writing provided by the embodiment of the invention specifically includes the following contents:

step 101: and receiving the writing content which is triggered and sent by the user and needs to quickly inquire auxiliary data in the writing process of the current writing page through a preset triggering mode.

In this step, it is assumed that the user wants to query some auxiliary data during the current page of the page, for example, the user is writing a financial report about Maotai, and at this time the user writes in the content about the business income of the third quarter before Maotai 2019, but because this part belongs to the data-related content, the user will not generally remember and need to query the related content. When the situation is met in the prior art, a user needs to search incomes in the third quarter before 2019 on some search platforms such as a hundred-degree platform or a dog searching platform, and then screening and copying search results are pasted into the writing process, but the processing mode has the defects that the user needs to switch to search pages of hundred degrees and the like to search related contents, further screening and copying are needed, and the user pastes the search results into the writing process, so that the process is very complicated, takes more time for the user, has very poor user experience, and is particularly troublesome for inquiring and screening complex data. In order to solve the problem, as shown in fig. 3, in particular, in the process of writing a current writing page, a user may directly trigger writing content (for example, "business income of the third quarter before the standard 2019") which needs to quickly query auxiliary data and is written by the user on the current page to a server through a preset trigger mode (for example, trigger by a shortcut key ctr+q), and then the server performs query intention recognition on the writing content (business income of the third quarter before the standard 2019) to obtain a query intention recognition result (object: standard; time, three quarters before 2019, index, business income), then according to the query intention recognition result, query a knowledge graph, obtain answer query recommendation results (for example, accurate answers, business income of the first three quarters before the top half of the year 2019 of the Maotai, business income of the first quarter of the year 2019 of the related answers, business income quarter trend of the first year 2019 of the year, etc.), and display the answer query recommendation results in a designated area of the current writing page for users to directly select and use the accurate answers and/or the related answers in the writing process. For example, as shown in FIG. 7, the user may directly select an exact answer in the designated area: revenue from the third quarter before the year 2019 of the cogongrass, and/or related answers, revenue from the first half of the year 2019 of the cogongrass, revenue from the first quarter of the year 2019 of the cogongrass, quarter trend in revenue from the year 2019 of the cogongrass, etc., thereby smoothly completing the authoring process. Further, fig. 8 and 9 show the following about the authoring contents: the exact and relevant recommendations of "three quarter Guizhou Maotai revenue, net profit" prior to 2019 will not be described in detail here.

In addition, in other embodiments of the present invention, the server may further perform different forms of rendering on the accurate answer and/or the relevant answer, and then display the rendered accurate answer and/or the relevant answer in a designated area of the current writing page. For example, the server may answer the exact answer: the business income of the third quarter before 2019 of the Maotai is rendered into a chart format and then displayed in the appointed area of the current writing page, so that a user can directly use the content in the chart format from the appointed area of the current writing page to solve the diagram allocation problem in the writing process, thereby greatly simplifying the writing process and perfectly assisting in completing intelligent writing. Without this rendering process, the user would have to draw the relevant form after getting revenue for the third quarter of the year before the 2019 year of the couchgrass, which would undoubtedly increase the authoring burden and reduce the authoring efficiency. It should be noted that, the specific rendering mode is specifically described in other embodiments, and is not limited to the chart format, and may be rendered into text, form, rich media, etc. in the specified format as required. As another example, referring to the graph rendering results shown in fig. 4, 5 and 6, the recommended results are all the fund flow directions of the cogongrass in the state, wherein, for fig. 4, 5 and 6, the recommended results are recommended in different manners from different dimensions, such as the day 5 dominant inflow of the cogongrass in the state, the day 5 dominant outflow of the cogongrass in the state, the day 5 dominant net inflow of the cogongrass in the state, the day fund flow directions of the cogongrass in the state, the day fund flow trend graphs of the cogongrass in the state, and the like.

Step 102: carrying out query intention recognition on the writing content of the auxiliary data needing to be queried quickly, and obtaining a query intention recognition result of the writing content of the auxiliary data needing to be queried quickly; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship between the entities.

In this step, after the user triggers the sending of the writing content for which the auxiliary data query is required, the query intention recognition needs to be performed on the writing content to recognize the query intention. Generally, the query intent recognition result includes entity dependency relationship recognition results. For example, assume that the authoring contents requiring shortcut-query assistance data are: "Guizhou Maotai 2019 achieves a business total revenue of ___ gigabytes" (where "___" represents auxiliary data in the authored content that requires a quick query), query intent identification can be performed for the authored content: the identification results of entities ("company", "time", "index name") are respectively "Guizhou Maotai", "2019", "incomes", "operating profits", and then the identification results of dependency relationship among the entities are obtained as a triplet ("Guizhou Maotai", "2019", "incomes"). In addition, when the query intention is identified, the category to which the query intention belongs is generally obtained at the same time, and the method can be understood as obtaining the field to which the query intention belongs, for example, determining whether the query intention belongs to a financial auxiliary data query, an economic auxiliary data query, or an educational auxiliary data query, and the like, so that the knowledge graph in the corresponding field can be conveniently queried to obtain the desired auxiliary data.

Step 103: acquiring auxiliary data which corresponds to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph;

in this step, a knowledge graph of the corresponding domain branch may be determined according to the query intent category recognition result, and then auxiliary data requiring quick query corresponding to the writing content may be obtained from the knowledge graph of the corresponding domain branch according to the entity dependency relationship recognition result.

The step 104 of having the association relation comprises the following steps: and displaying auxiliary data which are required to be quickly queried and correspond to the writing content in a designated area of the current writing page so as to be selected and used by a user for the auxiliary data in the writing process.

In this embodiment, after the server queries to obtain the auxiliary data requiring quick query corresponding to the writing content, the server displays the auxiliary data requiring quick query corresponding to the writing content in a specified area of the current writing page (such as a right area of the current writing page), so that the user can drag or paste the auxiliary data located in the specified area directly in the writing process of the current page to perform auxiliary writing, thereby greatly facilitating the user. Therefore, the embodiment of the invention can provide intelligent writing assistance in the user writing process, omits the process of searching auxiliary data and organizing the auxiliary data by switching the search platform in the user writing process, and is more critical in the embodiment of the invention that the query intention of the user about the writing auxiliary data can be obtained by analysis from the writing content, so that the corresponding writing auxiliary data is obtained from the related knowledge graph according to the query intention, thereby effectively improving the writing efficiency and the writing effect and enabling intelligent writing to be possible. Therefore, according to the intelligent auxiliary writing technical scheme provided by the embodiment of the invention, in the writing process, only the writing content needing to be quickly queried is required to be triggered, and the auxiliary data which is corresponding to the writing content and needs to be quickly queried can be provided for a user, so that the user can directly utilize the corresponding auxiliary data to assist the writing process, and the writing efficiency and the writing effect can be improved.

In this embodiment, it should be noted that, a user may connect a text writing page to an intelligent writing robot through a PC or a mobile terminal (an IOS system and an Android system), when the user is in a standby state in a normal writing process, and when the user needs to query a certain item of data when inputting a piece of text, the writing content needing to query auxiliary data may be sent to the writing server in a cursor position through a preset triggering manner, and the writing server performs intent analysis and query on the writing content needing to query auxiliary data, so as to obtain a query result about the auxiliary data, and sends the query result to the text writing page, and the query result is displayed in a designated area in the text writing page, so as to be used directly by the user in the writing process. When the user uses the query result from the designated area, various modes such as dragging, pasting, cutting, copying and the like may be adopted, which is not limited in this embodiment. Query results may also be used from a designated area, for example, by way of class input method interactions. For example, the user can select from top to bottom through the keyboard, and the answer data is quickly inserted by 0-9 shortcut data, space key and selection mode.

In this embodiment, it should be noted that, when the writing content requiring the auxiliary data query is sent to the writing server through a preset triggering manner, the writing content may be triggered by a shortcut key, or may be triggered by a single click or a double click of a mouse, which is not limited in this embodiment, and the preset triggering manner may be freely set according to practical applications, which is not particularly limited herein.

In this embodiment, after acquiring the writing content of the auxiliary data that the user needs to quickly query, the writing server performs query intention recognition on the writing content, thereby acquiring a query intention recognition result. It is to be appreciated that the intent recognition can include recognition analysis of writing requirements that a user is likely to generate in a current writing scenario, wherein the writing requirements can be multiple. And the query intention recognition result is a writing requirement generated by a user under the current writing scene according to the intention recognition.

In this embodiment, the accurate answer and/or the related answer obtained after the knowledge graph is queried according to the query intention recognition result are displayed in the designated area of the current writing page, where the designated area may be any blank of the writing page, and the specific limitation is not specified here. At this time, the user can select and use the auxiliary data for the answer recommendation result in the blank of the writing page. Furthermore, the user can insert the accurate answer and/or the related answer in the designated area in various data expression forms such as numerical values, texts, tables, pictures, data modules and the like through interactive modes such as clicking, selecting, spacing, keyboard shortcut keys and the like.

For example, assume that the user needs to query the authored content of the auxiliary data is: firstly, a user triggers the writing content 'Guizhou Maotai current day fund flow direction' in a preset triggering mode, and query intention recognition is carried out on the writing content 'Guizhou Maotai current day fund flow direction' according to user-triggered operation, so that a query intention recognition result is obtained: the current day, guizhou Maotai, and funding flows, it will be appreciated that query intent recognition results can extract a plurality of word slot content from the authoring intent for providing a plurality of dimensional answer references. Further, according to the query intention recognition result 'current day, guizhou Maotai and fund flow direction', the answer query recommendation result is obtained through the query knowledge graph: the current day fund flow data of the Guizhou Maotai (accurate answer), the current day fund flow data of the Guizhou Maotai (relevant answer) and the current day stock information of the Guizhou Maotai (relevant answer).

According to the technical scheme, the intelligent auxiliary writing processing method provided by the embodiment of the invention is characterized in that the query intention recognition result of the writing content requiring the quick query auxiliary data is obtained by carrying out query intention recognition on the writing content requiring the quick query auxiliary data, and then the knowledge graph is queried according to the query intention type recognition result and the entity dependency relationship recognition result, the auxiliary data requiring the quick query corresponding to the writing content is obtained, and the auxiliary data requiring the quick query corresponding to the writing content is displayed in the appointed area of the current writing page, so that the user can select the auxiliary data for use in the writing process, therefore, the embodiment of the invention can provide intelligent writing assistance in the user writing process, the auxiliary data requiring the query in the writing process of the user can be quickly provided, and the process of searching the auxiliary data by a search platform in the writing process of the user can be omitted. Therefore, according to the intelligent auxiliary writing technical scheme provided by the embodiment of the invention, in the writing process, only the writing content needing to be quickly queried is required to be triggered, and the auxiliary data which is corresponding to the writing content and needs to be quickly queried can be provided for a user, so that the user can directly utilize the corresponding auxiliary data to assist the writing process, and the writing efficiency and the writing effect can be improved.

Based on the content of the foregoing embodiment, in this embodiment, performing query intention recognition on the writing content requiring the shortcut query auxiliary data, and obtaining a query intention recognition result of the writing content requiring the shortcut query auxiliary data includes:

In this embodiment, when query intention recognition is performed on the writing content requiring the shortcut query auxiliary data, and a query intention recognition result of the writing content requiring the shortcut query auxiliary data is obtained, two parts of content need to be obtained, one part is to obtain a query intention type recognition result, and the other part is to obtain an entity dependency relationship recognition result. The purpose of obtaining the query intention category recognition result is to determine the category to which the query intention belongs, for example, whether the query intention belongs to an auxiliary data query of a financial class, an auxiliary data query of an economic class, or an auxiliary data query of an education class, and the like, so that the desired auxiliary data can be conveniently obtained from the knowledge graph of the corresponding field. The objective of obtaining the entity dependency relationship recognition result is to obtain a real query intention, and since the entity dependency relationship recognition result can clearly embody the query intention of writing content requiring quick query auxiliary data, the entity dependency relationship recognition result needs to be obtained, and the entity dependency relationship recognition needs to depend on the entity recognition result and the combination constraint among the entity recognition results.

For example, assume that the authoring content requiring the shortcut query assistance data is: the total incomes and the operating profits of the Guizhou Maotai 2019 are ___ hundred million yuan and ___ hundred million yuan respectively, and the semantic analysis process is as follows: the intention classification model identification result is "company financial index", the entity (company "," time "," index name ") identification result is" Guizhou Maotai "," 2019"," business income "," business profit ", and the dependency relationship identification result is the 1 st triplet (Guizhou Maotai", "2019", "business income"). The 2 nd triplet ("Guizhou Maotai", "2019", "operating profit") can find an answer from the knowledge graph of the corresponding domain by adding the intention classification of "corporate financial index" to the 2 triples identified above.

Based on the content of the foregoing embodiment, in this embodiment, performing query intention type recognition on the writing content requiring shortcut query assistance data, to obtain a query intention type recognition result, includes:

In this embodiment, it should be noted that, as described above, when query intention type recognition is performed on the written content requiring shortcut query assistance data, a query intention type recognition result needs to be obtained. It can be understood that the query intention type recognition result is used for representing the domain branch corresponding to the writing content requiring the shortcut query auxiliary data, so that the domain knowledge graph corresponding to the auxiliary data requiring the shortcut query in the writing content can be determined through the query intention type recognition result.

In order to solve the problem of how to acquire the query intention category recognition result, the embodiment adopts an intelligent intention category recognition model to perform query intention category recognition so as to improve the efficiency and accuracy of query intention category recognition.

Specifically, before using the intent category recognition model, training is required, and during training, a training sample of the writing content of each known query intent category can be used as sample input data, and the query intent category corresponding to the writing content of each known query intent category can be used as sample output data, so that the training is obtained based on a machine learning algorithm.

In this embodiment, it should be noted that the intent type recognition model is generated based on machine learning algorithm training based on the written content of each known query intent type. When the intention type recognition model is generated based on machine learning algorithm training, generally, writing content training samples of known query intention types are taken as sample input data, query intention types corresponding to writing content of the known query intention types are taken as sample output data, and an initial machine learning model is trained until model convergence conditions are met, and then the intention type recognition model is generated. In this embodiment, a CNN or RNN machine learning model may be employed for model training.

Based on the content of the foregoing embodiment, in this embodiment, performing entity identification and entity dependency identification on the authoring content requiring shortcut query assistance data, to obtain an entity identification result and an entity dependency identification result, includes:

In this embodiment, the authoring content is segmented to obtain a plurality of sentences with complete semantics;

for each sentence, the following processing procedure is performed: performing word segmentation and part-of-speech tagging on the sentence to perform entity recognition so as to obtain key entities (object entities) in the sentence, and performing general entity recognition on the sentence so as to at least obtain time entities, condition entities, index entities and the like in the sentence;

and judging the association relation among the time entity, the condition entity, the index entity and the object entity through dependency relation extraction, and further determining an entity dependency relation identification result among the entities.

For example, assume that the authoring content requiring the shortcut query assistance data is: "the total incomes and the operating profits of the Guizhou Maotai 2019 are ___ hundred million yuan and ___ hundred million yuan respectively, which are increased by ___ percent and ____ percent respectively in the same ratio", the semantic analysis process is as follows: the intention classification model identification result is "company financial index", the entity (company "," time "," index name ") identification result is" Guizhou Maotai "," 2019"," business income "," business profit ", and the dependency relationship identification result is the 1 st triplet (Guizhou Maotai", "2019", "business income"). The 2 nd triplet ("Guizhou Maotai", "2019", "operating profit") can find an answer from the knowledge graph of the corresponding domain by adding the intention classification of "corporate financial index" to the 2 triples identified above.

In this embodiment, as shown in fig. 3, when performing semantic analysis on the writing content, the sentence may be first divided, then semantic recognition may be performed on the divided sentence, including performing part-of-speech recognition, entity recognition, relationship recognition, and other information about knowledge maps in related fields, and finally identifying the time entity, the object entity, the index entity, and the semantic dependency relationship between the entities in the writing content.

It can be appreciated that semantic analysis is generally analysis and recognition of text content such as sentences and paragraphs in different dimensions by means of natural language processing, machine learning and the like, including but not limited to analysis and recognition modules such as "word segmentation (word breaking)", "part-of-speech tagging", "entity recognition", "time recognition", "condition recognition", "dependency relationship extraction (dependency syntax analysis)", "intention recognition", "semantic dependency analysis", and the like. The method comprises the steps of firstly cutting a target text (paragraph and long sentence) into a sentence with complete relative semantics, carrying out entity recognition on the sentence, such as key entities of a company organization, a region, a person and the like in the sentence, and then carrying out further universal entity recognition such as time recognition, condition recognition, index recognition and the like, wherein in addition to time and condition, each entity ideally corresponds to entity knowledge in a knowledge graph (knowledge graph) one by one, and then judging the association relationship among the entities, the conditions, the time and the like through dependency relation extraction (dependency syntactic analysis).

It can be understood that, the semantic extraction and recognition generally includes performing semantic coding on characters through a pre-training language model to generate high-latitude semantic vectors, performing sequence labeling and classification tasks such as entity recognition, relation extraction, intent recognition and the like through deep learning models such as Convolutional Neural Networks (CNN), cyclic neural networks (RNN), long and short memory units (LSTM), attention mechanism-based models (such as transformers and derivative models Bert and XLNet based on the same), and performing extraction and recognition processing through models such as support vector machines, bayes, hidden markov, conditional random fields and the like.

Based on the foregoing embodiment, in this embodiment, according to the query intention category recognition result and the entity dependency relationship recognition result, a query knowledge graph is obtained, and auxiliary data, which corresponds to the writing content and needs to be queried quickly, includes:

In this embodiment, it should be noted that, each domain branch corresponds to a corresponding knowledge graph, so when auxiliary data requiring quick query corresponding to the writing content is actually obtained, the knowledge graph of the corresponding domain branch needs to be determined according to the query intention type recognition result, and then the auxiliary data requiring quick query corresponding to the writing content is obtained from the knowledge graph of the domain branch.

It can be understood that knowledge and time sequence data of each domain branch are stored in a knowledge graph of each domain branch, the knowledge and time sequence data comprise index information about each object in each time in the corresponding domain branch, and the entity dependency relationship identification result comprises a combined constraint relationship among an object entity, a time entity and an index entity, so that auxiliary data which are corresponding to the writing content and need to be quickly queried can be obtained from the knowledge graph of the corresponding domain branch according to the entity dependency relationship identification result, and the purpose of intelligent auxiliary writing is achieved.

For example, assume that the authoring content requiring the shortcut query assistance data is:

"the total incomes and the operating profits of the Guizhou Maotai 2019 are ___ hundred million yuan and ___ hundred million yuan respectively, which are increased by ___ percent and ___ percent respectively in the same ratio", the semantic analysis process is as follows: the intention classification model identification result is a 'company financial index', the entity attribute (company ',' time ',' index name ') identification result is' Guizhou Maotai ',' 2019 ',' business income ',' business profit ', and the dependency relationship identification result is a 1 st triplet (Guizhou Maotai', '2019', 'business income'). The 2 nd triplet (Guizhou Maotai "," 2019"," operating profit ") is so that through the above identified 2 triples, finally, the identification result is classified according to the intention of" company financial index ", so that the corresponding auxiliary data can be searched in the knowledge graph corresponding to the" company financial index ", thereby completing the purpose of intelligent auxiliary writing.

In this embodiment, when the auxiliary data requiring quick query corresponding to the writing content is acquired, the accurate auxiliary data requiring quick query corresponding to the writing content (hereinafter, simply referred to as accurate answer) may be acquired, and the relevant auxiliary data requiring quick query corresponding to the writing content (hereinafter, simply referred to as relevant answer) may also be acquired.

It is understood that the exact answer is an answer that matches the query intent recognition result. For example, assume that the query intent recognition results are: the third quarter before 2019, the accurate answer is the answer that exactly matches the third quarter before 2019, that is, the accurate answer is: revenue data for the third quarter of the year before 2019.

It may be understood that the relevant answer refers to an answer having an association relationship with the query intention recognition result. For example, assume that the query intent recognition results are: the third quarter (time) before 2019, and the incomes (indexes), the relevant answer is an answer with an association relationship with incomes of the third quarter before 2019. The answer having the association relationship here may be an answer matching any one or both of the query object, the query time, and the query index. For example, the relevant answer may be: revenue of the first half of the year in the Maotai 2019 (changed query time, but no change in query object and query time), revenue of the first quarter of the year in the Maotai 2019 (changed query time, but no change in query object and query time), revenue of the quarter in the year in the Maotai 2019 (changed query time and query index, but no change in query object), revenue of the third quarter before the year in the wuliangye 2019 (changed query object, but no change in query time and query index), charitable donation of the third quarter before the year in the Maotai 2019 (changed query index, but no change in query time and query object), stock sales of the third quarter before the year in the Maotai 2019 (changed query index, but no change in query time and query object), and so on.

In this step, it should be noted that the meaning of providing an accurate answer is: the user can be assisted in completing the data required for writing. And the meaning of providing the relevant answer is that: the method can help the user to diverge or expand the writing thought and writing content, or help the user to form data contrast in writing, or provide richer data support for the user in writing, and increase the attraction and persuasion of articles. In addition, the meaning of providing the relevant answer is that when the accurate matching result does not exist in the knowledge graph in the writing content of a certain wanted query, the relevant answer can be provided to assist the user as much as possible. Therefore, the embodiment can provide accurate answers and related answers for the user at the same time, namely, the embodiment can provide answers with multiple dimensions for the user to reference, so that more data selections can be provided for the user, and user experience is improved.

It can be appreciated that, in order to provide a relevant answer, in this embodiment, an extended entity dependency relationship recognition result corresponding to the entity dependency relationship recognition result needs to be determined according to the entity dependency relationship recognition result; wherein, the object entity in the extended entity dependency relationship identification result and the object entity in the entity dependency relationship identification result have a class relationship or a bid relationship (such as the relationship between Maotai and wuliangye, or the relationship between submarine fishing and sea fishing shop, or the relationship between Shanghai tobacco and Beijing tobacco, etc.); the time entity in the extended entity dependency relationship identification result and the time entity in the entity dependency relationship identification result have adjacent relationship or periodic relationship (such as relationship between the first quarter and the second quarter, relationship between the first quarter of the last year and the first quarter of the current year, etc.); the index entity in the extended entity dependency relationship recognition result has a similar relationship or an opposite relationship (such as a relationship between total annual income and total annual profit, a relationship between total annual income and annual loss), it can be understood that, because knowledge and time sequence data of corresponding domain branches are stored in the knowledge graph of the corresponding domain branches, the knowledge and time sequence data includes each index information of each object in each time in the corresponding domain branches, so that accurate answers and relevant answers can be provided simultaneously, and accurate auxiliary data which is required to be quickly queried and corresponds to the writing content can be obtained from the knowledge graph of the corresponding domain branches according to the entity dependency relationship recognition result; and acquiring relevant auxiliary data which corresponds to the writing content and needs to be quickly queried from the knowledge graph of the corresponding field branch according to the identification result of the extended entity dependency relationship, so that an accurate answer and a relevant answer can be simultaneously provided for a user to use according to the needs.

Based on the foregoing embodiment, in this embodiment, the displaying, in a specified area of a current writing page, auxiliary data that needs to be queried quickly and corresponds to the writing content, for a user to select the auxiliary data for use in a writing process includes:

In this embodiment, it may be understood that the obtained auxiliary data corresponding to the writing content and requiring quick query may be displayed in a blank of the writing surface in a text, form, chart or rich media rendering form, for example, the obtained auxiliary data corresponding to the writing content and requiring quick query may be automatically drawn to form a multi-specification graph, so as to provide convenience for data insertion and use for the user to select and use the auxiliary data.

In this embodiment, by performing different forms of rendering processing on the auxiliary data which needs to be quickly queried and corresponds to the writing content, the user can select answers with different rendering results, and the user experience is improved while the answer display forms are enriched. The rendering result can be pictures, charts, tables, texts with specified formats and the like, so that various data forms are provided for users, the users can select the data conveniently according to the requirements in the writing process, and the quality and the writing richness of intelligent writing can be improved.

In this embodiment, it should be noted that, in the conventional writing method, data and knowledge are queried and retrieved mainly manually, and data such as tables and pictures are repeatedly drawn and organized, which is often tedious and inefficient. With the increasing development of deep learning, high-quality writing scenes (macro economy, macro strategy, industry, analysis report and weekly report of companies, financial report comment of companies, government report, academic report and the like) are more dependent on a large amount of data, and the whole process of proposition, gathering materials, writing, map matching, auditing and issuing is performed, so that the production efficiency of articles and reports is low, and the quality can be guaranteed, which makes it difficult for a plurality of people to relieve from heavy labor. For this reason, the present embodiment redefines an intelligent content production mode, and the production of this mode can greatly improve the scale and stability of content production. In this mode, machine authoring is finally simplified to article assembly by a series of preforms, and the article production mode is industrially modified in this way, so that the core process and quality control of content production are moved forward to the preform production link. Compared with the traditional writing tool, the intelligent auxiliary writing processing method provided by the embodiment has the following advantages: (1) answer recommendation layer: whether the answer is task answer configuration or question answer configuration, the dynamic parameters can be adjusted to update in real time, self learning is realized through functions of self learning, active reporting and the like, writing support is provided for intention recognition and answer repackaging, and traditional intelligent writing is more dependent on search knowledge and fixed knowledge content. Writing interaction layer: the method meets the normal writing interaction, and the user experiences water to channel, so that the method perfectly meets the interaction requirement of workers. Knowledge management layer: knowledge graph technology is an important component of artificial intelligence technology that describes the relationships of concepts, entities and their keys in the objective world in a structured manner. The knowledge graph providing technology provides a better capability of organizing, managing and understanding mass information of the Internet, and the information of the Internet is expressed in a form closer to the human cognitive world. Therefore, a knowledge graph with semantic processing capability and open interconnection capability is established, and huge application value can be generated in intelligent writing and personalized recommendation.

In this embodiment, as shown in fig. 10, the intelligent writing robot may be connected to a text writing page of a PC or a mobile terminal (IOS system and Android system), in which when a user needs to query a certain item of data in a normal writing process, the user can input the user query text to the writing server by triggering a shortcut key such as ctrl+q at a cursor position when the user inputs a certain piece of text, the writing server firstly performs sentence processing on the input text, then extracts key semantic information in the text, identifies the real intention of the user of the text block, and performs instruction disambiguation and complex entity identification, and finally constructs the text block as a semantic analysis result. Searching corresponding data and data rendering information by utilizing a semantic analysis result, returning the data to the multi-type writing template output module, and finally returning the data to a user according to a rendering format for clicking and inserting the data into a current writing area. As shown in fig. 11, the man-machine interaction method based on the intelligent writing robot comprises the following steps: step 1: opening the intelligent writing assistant text input page. Step 2: text content before a cursor when a user presses a shortcut key (such as ctrl+q, etc.) is received. Step 3: and sending the received text content to the intelligent writing server. Step 4: the intelligent writing service analyzes the text information and outputs a semantic analysis result. Step 5: and querying corresponding data and rendering modes by using semantic analysis results. Step 6: and presenting the queried data to a user in a rendering mode. Step 7: the user selects and clicks on the desired data content, inserting it into the cursor position.

Based on the same inventive concept, another embodiment of the present invention provides an intelligent auxiliary writing processing apparatus, referring to fig. 12, a schematic structural diagram of the intelligent auxiliary writing processing apparatus according to an embodiment of the present invention includes: a receiving module 21, a first acquiring module 22, a second acquiring module 23 and a displaying module 24, wherein:

Since the intelligent auxiliary writing processing device provided in the present embodiment may be used to execute the intelligent auxiliary writing processing method described in the above embodiment, the working principle and the beneficial effects thereof are similar, so that details will not be described in detail herein, and specific details will be referred to the description and the description of the above embodiment.

Based on the same inventive concept, a further embodiment of the present invention provides an electronic device, which is described with reference to fig. 13, and specifically includes the following: a processor 301, a memory 302, a communication interface 303, and a communication bus 304;

wherein, the processor 301, the memory 302, and the communication interface 303 complete communication with each other through the communication bus 304; the communication interface 303 is used for realizing information transmission between devices;

the processor 301 is configured to invoke a computer program in the memory 302, where the processor executes the computer program to implement all the steps of the above-mentioned intelligent aided writing processing method, for example, the processor executes the computer program to implement the following steps: receiving writing contents which are triggered by a user in the writing process of the current writing page and need to quickly inquire auxiliary data through a preset triggering mode; carrying out query intention recognition on the writing content of the auxiliary data needing to be queried quickly, and obtaining a query intention recognition result of the writing content of the auxiliary data needing to be queried quickly; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities; acquiring auxiliary data which corresponds to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph; and displaying auxiliary data which are required to be quickly queried and correspond to the writing content in a designated area of the current writing page so as to be selected and used by a user for the auxiliary data in the writing process.

Based on the same inventive concept, a further embodiment of the present invention provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements all the steps of the user comment generation method described above, for example, the processor implementing the following steps when executing the computer program: receiving writing contents which are triggered by a user in the writing process of the current writing page and need to quickly inquire auxiliary data through a preset triggering mode; carrying out query intention recognition on the writing content of the auxiliary data needing to be queried quickly, and obtaining a query intention recognition result of the writing content of the auxiliary data needing to be queried quickly; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities; acquiring auxiliary data which corresponds to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph; and displaying auxiliary data which are required to be quickly queried and correspond to the writing content in a designated area of the current writing page so as to be selected and used by a user for the auxiliary data in the writing process.

Further, the logic instructions in the memory described above may be implemented in the form of software functional units and stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment of the invention. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the user comment generating method described in the various embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent aided writing processing method is characterized by comprising the following steps:

performing query intention recognition on the writing content requiring the shortcut query auxiliary data to obtain a query intention recognition result of the writing content requiring the shortcut query auxiliary data, wherein the query intention recognition result comprises the following steps:

performing entity identification and entity dependency relationship identification on the writing content needing shortcut query auxiliary data to acquire an entity identification result and an entity dependency relationship identification result;

wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities;

obtaining auxiliary data which corresponds to the writing content and needs to be queried quickly according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph, wherein the auxiliary data comprises the following components:

2. The intelligent aided writing processing method of claim 1, wherein the entity recognition result at least includes: object entity, time entity and index entity; the entity dependency relationship identification result at least comprises: a combined constraint between the object entity, the time entity, and the index entity.

3. The intelligent aided writing processing method of claim 1, wherein the query intention type recognition is performed on the writing content requiring shortcut query assistance data, and the query intention type recognition result is obtained, including:

4. The intelligent aided writing processing method of claim 1, wherein performing entity recognition and entity dependency relationship recognition on the writing content requiring shortcut query assistance data, obtaining an entity recognition result and an entity dependency relationship recognition result, includes:

5. The intelligent aided writing processing method of claim 3, wherein knowledge and time sequence data of the corresponding domain branch are stored in a knowledge graph of the corresponding domain branch, and the knowledge and time sequence data comprise index information of each object at each time in the corresponding domain branch.

6. The method for intelligent auxiliary writing processing according to claim 5, wherein displaying auxiliary data requiring quick query corresponding to the writing content in a designated area of a current writing page for a user to select the auxiliary data for use in the writing process, includes:

7. An intelligent auxiliary writing processing device, which is characterized by comprising:

the first obtaining module is configured to identify a query intention of the writing content requiring the shortcut query auxiliary data, and obtain a query intention identification result of the writing content requiring the shortcut query auxiliary data, where the query intention identification result includes: carrying out query intention type recognition on the writing content needing the shortcut query auxiliary data to obtain a query intention type recognition result; performing entity identification and entity dependency relationship identification on the writing content needing shortcut query auxiliary data to acquire an entity identification result and an entity dependency relationship identification result; wherein the query intention recognition result comprises a query intention category recognition result and an entity dependency relationship recognition result; wherein the entity dependency relationship recognition result is obtained depending on the entity recognition result and the dependency relationship among the entities;

The second obtaining module is configured to obtain auxiliary data, which corresponds to the writing content and needs to be queried quickly, according to the query intention category recognition result and the entity dependency relationship recognition result query knowledge graph, and includes: determining a knowledge graph of the branch in the corresponding field according to the query intention category recognition result; determining an extended entity dependency relationship identification result corresponding to the entity dependency relationship identification result according to the entity dependency relationship identification result; wherein, the object entity in the extended entity dependency relationship identification result and the object entity in the entity dependency relationship identification result have a class relationship or a bid relationship; the time entity in the extended entity dependency relationship identification result has an adjacent relationship or a periodic relationship with the time entity in the entity dependency relationship identification result; the index entity in the extended entity dependency relationship identification result has a similar relationship or an opposite relationship with the index entity in the entity dependency relationship identification result; according to the entity dependency relationship identification result, acquiring accurate auxiliary data which corresponds to the writing content and needs to be quickly queried from the knowledge graph of the corresponding field branch; acquiring relevant auxiliary data which are corresponding to the writing content and need to be quickly queried from the knowledge graph of the corresponding field branch according to the identification result of the extended entity dependency relationship;

8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the intelligent aided writing processing method according to any one of claims 1-6 when executing the program.

9. A non-transitory computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor implements the steps of the intelligent aided authoring processing method of any one of claims 1-6.