CN117494800A - Task execution method and device based on constructed nuclear fusion knowledge graph - Google Patents

Abstract

The specification discloses a task execution method and device based on a constructed nuclear fusion knowledge graph, which can judge the correlation information between entity name information according to a target model trained by corpus in the nuclear fusion field after the entity name information is acquired by corpus, and does not need to manually judge the correlation information. And after the correlation information is obtained, a nuclear fusion knowledge graph can be further constructed, and the experimental task is subjected to task analysis through the nuclear fusion knowledge graph, so that the time required by the task analysis is greatly reduced, and the task execution efficiency is improved.

Description

Task execution method and device based on constructed nuclear fusion knowledge graph

Technical Field

The specification relates to the field of artificial intelligence and the field of nuclear fusion, in particular to a task execution method and device based on a constructed nuclear fusion knowledge graph.

Background

In recent years, as the conventional energy source is increasingly unable to meet the energy source demand of human beings, the field of nuclear fusion is rapidly developed to meet the increasing energy source demand. In recent years, magnetic confinement fusion research is the most possible way to realize nuclear fusion, but with the increasing progress of technology, the cost and difficulty of experiments are also higher and higher, and the magnetic confinement fusion research is the biggest bottleneck in nuclear fusion experimental research. The improvement of the experiment cost and the difficulty makes the knowledge demand of professionals on physical knowledge increase, but the uniqueness of the nuclear fusion field makes the difficulty of professionals in learning the knowledge in the field extremely high, and the personal knowledge quantity can not solve all problems in the nuclear fusion experiment.

Therefore, how to effectively analyze the task results in the field of nuclear fusion to improve the subsequent task execution efficiency is a problem to be solved.

Disclosure of Invention

The specification provides a task execution method and device based on a constructed nuclear fusion knowledge graph, so as to partially solve the problems existing in the prior art.

The technical scheme adopted in the specification is as follows:

the specification provides a task execution method based on a constructed nuclear fusion knowledge graph, which comprises the following steps:

acquiring name information of each entity in the nuclear fusion field;

inputting the entity name information into a preset first target model aiming at each entity name information, so as to output text description content corresponding to the entity name information through the first target model;

determining the correlation information among the entity name information according to the text description content corresponding to each entity name information;

each entity name information corresponds to different nodes respectively, and edges among the nodes are determined according to the correlation information among the entity name information, so that a knowledge graph in the nuclear fusion field is constructed;

and carrying out task analysis on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field to obtain an analysis result, and executing target tasks according to the analysis result.

Optionally, acquiring name information of each entity in the nuclear fusion field specifically includes:

acquiring corpus data in the nuclear fusion field;

determining non-subclass entities from the corpus data as candidate entities;

inputting entity name information corresponding to each candidate entity into a preset second target model for each candidate entity, and outputting a judging result of judging whether the candidate entity is an entity in the nuclear fusion field through the second target model;

and determining each entity in the nuclear fusion field from each candidate entity according to the judging result output by the second target model aiming at each candidate entity, and acquiring entity name information of each determined entity in the nuclear fusion field.

Optionally, for each entity name information, the entity name information is input into a preset first target model, so that text description content corresponding to the entity name information is output through the first target model, and specifically includes:

inputting the entity name information and the determined prompt word for the entity name information into a preset first target model aiming at each entity name information, so that the first target model outputs text description content corresponding to the entity name information based on the prompt word;

And if the text description content corresponding to the entity name information is not the target description content corresponding to the entity name information, acquiring a modified prompt word aiming at the entity name information, and re-inputting the entity name information and the modified prompt word into the first target model, so that the first target model outputs the text description content corresponding to the entity name information based on the modified prompt word until the first target model outputs the target description content corresponding to the entity name information.

Optionally, determining the correlation information between the entity name information according to the text description content corresponding to each entity name information specifically includes:

inputting the two entity name information, text description contents corresponding to the two entity name information and prompt words for describing the correlation between the two entity name information into a preset third target model aiming at any two entity name information, so that the third target model determines and outputs the correlation information between the two entity name information according to the prompt words for describing the correlation between the two entity name information.

Optionally, the method further comprises:

judging whether the correlation information between the two entity name information output by the third target model is the target correlation information between the two entity name information or not;

and responding to the fact that the correlation information between the two entity name information is not the target correlation information between the two entity name information, acquiring a modified prompt word for the two entity name information, and re-inputting text description contents corresponding to the two entity name information and the modified prompt word into the third target model, so that the third target model outputs the correlation information between the two entity name information based on the modified prompt word until the third target model outputs the target correlation information between the two entity name information.

The specification provides a task execution device based on a constructed nuclear fusion knowledge graph, which comprises:

the acquisition module is used for acquiring information of each entity in the nuclear fusion field;

the first determining module is used for inputting the entity name information into a preset first target model aiming at each entity name information so as to output text description content corresponding to the entity name information through the first target model;

The second determining module is used for determining the correlation information among the entity names according to the text description content corresponding to each entity name information;

the third determining module is used for respectively corresponding different nodes according to the name information of each entity and determining edges among the nodes according to the correlation information among the name information of each entity so as to construct a knowledge graph in the nuclear fusion field;

and the control module is used for carrying out task analysis on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field to obtain an analysis result, and executing target tasks according to the analysis result.

Optionally, the acquisition module is specifically configured to,

acquiring corpus data in the nuclear fusion field;

determining non-subclass entities from the corpus data as candidate entities;

inputting entity name information corresponding to each candidate entity into a preset second target model for each candidate entity, and outputting a judging result of judging whether the candidate entity is an entity in the nuclear fusion field through the second target model;

and determining each entity in the nuclear fusion field from each candidate entity according to the judging result output by the second target model aiming at each candidate entity, and acquiring entity name information of each determined entity in the nuclear fusion field.

Optionally, the first determining module is specifically configured to,

inputting the entity name information and the determined prompt word for the entity name information into a preset first target model aiming at each entity name information, so that the first target model outputs text description content corresponding to the entity name information based on the prompt word;

and if the text description content corresponding to the entity name information is not the target description content corresponding to the entity name information, acquiring a modified prompt word aiming at the entity name information, and re-inputting the entity name information and the modified prompt word into the first target model, so that the first target model outputs the text description content corresponding to the entity name information based on the modified prompt word until the first target model outputs the target description content corresponding to the entity name information.

The present specification provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements a task execution method based on a constructed nuclear fusion knowledge graph.

The specification provides an electronic device, which comprises a processor and a computer program stored on a memory and capable of running on the processor, wherein the task execution method based on the constructed nuclear fusion knowledge graph is realized when the processor executes the program.

The above-mentioned at least one technical scheme that this specification adopted can reach following beneficial effect:

in the task execution method based on the constructed nuclear fusion knowledge graph provided by the specification, each entity name information in the nuclear fusion field is acquired, text description content corresponding to each entity name information and correlation information among the entity name information are acquired for each entity name information, each entity name information corresponds to different nodes, edges of each node are determined according to the correlation information among the entity name information, the knowledge graph in the nuclear fusion field is constructed, further, task analysis is performed on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field, and target tasks are executed according to the obtained analysis results.

According to the method, after entity name information is obtained through corpus, the correlation information among the entity name information can be judged according to the target model trained by corpus in the nuclear fusion field, and a judgment process of the entity name information is not needed to be manually carried out. And after the correlation information is obtained, a nuclear fusion knowledge graph can be further constructed, and the experimental task is subjected to task analysis through the nuclear fusion knowledge graph, so that the time required by the task analysis is greatly reduced, and the task execution efficiency is improved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the specification, illustrate and explain the exemplary embodiments of the present specification and their description, are not intended to limit the specification unduly. In the drawings:

fig. 1 is a schematic flow chart of a task execution method based on a constructed nuclear fusion knowledge graph provided in the present specification;

FIG. 2 is a flow chart for constructing a knowledge graph provided in the present specification;

FIG. 3 is a schematic diagram of a knowledge graph provided in the present specification;

fig. 4 is a schematic structural diagram of a task execution device based on a constructed nuclear fusion knowledge graph provided in the present specification;

fig. 5 is a schematic structural diagram of an electronic device corresponding to fig. 1 provided in the present specification.

Detailed Description

For the purposes of making the objects, technical solutions and advantages of the present specification more apparent, the technical solutions of the present specification will be clearly and completely described below with reference to specific embodiments of the present specification and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present specification. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.

The following describes in detail the technical solutions provided by the embodiments of the present specification with reference to the accompanying drawings.

Fig. 1 is a schematic flow chart of a task execution method based on a constructed nuclear fusion knowledge graph, provided in the present specification, including:

s101: and acquiring name information of each entity in the nuclear fusion field.

The execution main body of the task execution method based on the constructed nuclear fusion knowledge graph provided in the present specification may be a terminal device such as a notebook computer or a desktop computer, or may be a client or a server installed in the terminal device, and for convenience of description, only the terminal device is taken as the execution main body as an example, and the task execution method based on the constructed nuclear fusion knowledge graph provided in the present specification is described below. In the field of nuclear fusion, the task analysis mode for performing the experimental task at present is often limited to manual work, and the manual analysis of the nuclear fusion experimental task has high requirements on the knowledge storage of experimenters, which is difficult to realize in the field of nuclear fusion. That is, due to the limitation of knowledge in the field of nuclear fusion, the existing method cannot perform accurate task analysis on the nuclear fusion experimental task. Therefore, how to accurately analyze the experimental task is a urgent problem for the field of nuclear fusion.

Therefore, the specification provides a construction method of the nuclear fusion knowledge graph, and task analysis can be carried out on experimental tasks according to the constructed knowledge graph. The present nuclear fusion knowledge graph does not exist in the field of nuclear fusion, because the construction of the nuclear fusion knowledge graph requires entity name information in the field of nuclear fusion and correlation relations among the entity name information, and it is difficult for the field of nuclear fusion to acquire the entity name information and correlation relations among the entity name information sufficient to construct the nuclear fusion knowledge graph.

Based on the above problems, the present specification provides a solution, that is, entity name information is obtained in advance, then a correlation relationship between the entity name information is obtained through a target model, and a knowledge graph under the nuclear fusion field is constructed based on the determined correlation relationship between the entity name information.

In the process of constructing the nuclear fusion knowledge graph, the terminal equipment can acquire corpus data in the nuclear fusion field from public data sets, open data sources, manual input data and other channels, wherein after the corpus data in the nuclear fusion field is acquired, the corpus data can be processed, and particularly, the corpus data can be segmented in a semantic analysis mode so as to extract name information of each entity from the acquired corpus data.

It should be noted that, the knowledge graph constructed in the present specification may be implemented based on two different contexts, and firstly, there is no related knowledge graph in the nuclear fusion domain before that, then the knowledge graph in the nuclear fusion domain needs to be constructed through the acquired name information of each entity and the determined correlation relationship between the name information of each entity; the related knowledge maps in the nuclear fusion field are stored before, but the content of the stored knowledge maps is less, so that the knowledge maps constructed in the specification can be understood as the knowledge maps which are expanded on the original basis, and the expanded knowledge maps are regarded as the knowledge maps constructed by the method provided by the specification.

For the second case, the obtained entity name information may be extracted from the existing nuclear fusion knowledge graph, and some corpus data may be obtained on the basis of the obtained entity name information, so as to extract some entity name information that is not recorded in the existing knowledge graph from the corpus data. Then, the entity name information acquired in this case can be understood to include entity name information extracted from the existing knowledge graph, and entity name information extracted from the acquired corpus data.

It should be noted that, after the entity name information is obtained, the obtained entity name information needs to be filtered, because even in the related corpus of the nuclear fusion corpus, the extracted entity name information may not be an entity in the field of nuclear fusion, which is not beneficial to the construction of the nuclear fusion knowledge graph.

Based on this, in the present specification, the terminal device may input the obtained entity name information into a preset second target model, so as to determine whether the input entity name information is an entity in the fusion domain through the second target model, if so, the entity name information is reserved, otherwise, the entity name information is removed. Wherein the second object model may be obtained by supervised training.

In the process of training the second target model, sample entity name information and label information corresponding to each sample entity name information can be acquired first, wherein the label information is used for indicating whether the pre-marked sample entity name information is entity name information in the nuclear fusion field or not. And then, the sample entity name information can be input into a second target model to be trained, and the second target model can output a judgment result aiming at the sample entity name information. And finally, training a second target model by taking the deviation between the minimized judging result and the label information corresponding to the sample entity name information as an optimization target.

Since similar entities may exist in the entities retained by the second object model, the acquired entity name information may optionally be filtered before being input into the second object model.

Specifically, before extracting the entity name information from the corpus data in the nuclear fusion field, a relationship tree is established through the potential relationship between the entity name information. The relationship tree comprises a root node, a child node and a leaf node, and an entity corresponding to the leaf node represents a non-subclass entity without subclasses, namely, the entity which is not duplicated or similar is guaranteed. Therefore, entity name information corresponding to leaf nodes in the relationship tree can be used as each candidate entity. And then, the terminal equipment can input the determined candidate entities into the second target model so as to determine entity name information of each entity in the nuclear fusion field from the candidate entities.

Of course, the filtering of the acquired entity name information may also be performed after the second object model determines the acquired entity name information. That is, which entity name information is the entity name information in the nuclear fusion field is screened out through the second target model, and then the entity name information without subclass is further screened out from the entity name information.

S102: and inputting the entity name information into a preset first target model aiming at each entity name information so as to output text description content corresponding to the entity name information through the first target model.

After acquiring the name information of each entity in the nuclear fusion field through the related corpus data in the nuclear fusion field, the terminal equipment can input the acquired name information of the entity into a preset first target model so as to acquire text description content corresponding to the name information of the entity through the first target model.

The first target model can be obtained through training in a supervised training mode. In the training process of the first target model, sample entity name information and label texts corresponding to the sample entity name information can be acquired first, wherein the label texts are used for representing standard interpretation texts corresponding to the predetermined sample entity name information. The sample entity name information may then be input into a first target model to be trained, which may then output text description content for the sample entity name information. And finally, training the first target model by taking the deviation between the text description content output by the minimized first target model and the label text corresponding to the sample entity name information as an optimization target.

Of course, in the present specification, the first target model may also be an existing large language model, and then the terminal device may call the large language model through a preset interface, so as to obtain text description content corresponding to each entity name information through text output capability of the large language model.

S103: and determining the correlation information among the entity name information according to the text description content corresponding to each entity name information.

After obtaining entity name information of each candidate entity and text description content corresponding to the entity name information, the terminal equipment can obtain correlation information among the entity name information through a preset third target model, so that a nuclear fusion knowledge graph is constructed.

The terminal device may use the two entity name information as an entity pair, where the entity pair includes the entity name information of the two entities and text description contents corresponding to the two entity name information.

For each entity pair, the terminal equipment respectively inputs each entity pair into a third target model, so that an explanation text of the correlation of the two entity name information in the entity pair is obtained, the correlation relation between the two entity name information is judged through the explanation text of the correlation, and then the correlation relation between the two entity name information is output. The terminal device may obtain the correlation relationship between all the entity name information through the third object model, where the indicated correlation information in the specification may be used to indicate whether there is a positive correlation, a negative correlation, or no correlation between the entities corresponding to the two entity name information.

The third target model may be trained by a supervised training approach. In the training process of the third target model, first, a sample entity pair and label correlation information corresponding to each sample entity pair can be obtained. The sample entity pair may then be input into a third target model to be trained, which then outputs correlation information for the sample entity pair. And finally, training the third target model by taking the deviation between the correlation information output by the minimized third target model and the label correlation information corresponding to the sample entity pair as an optimization target.

Of course, in the present specification, the third target model may be an existing large language model, and then the terminal device may call the large language model through a preset interface, so as to obtain the correlation information between the name information of each entity through the text output capability of the large language model.

S104: and determining edges among the nodes according to the correlation information among the entity name information so as to construct a knowledge graph in the nuclear fusion field.

After acquiring the name information of each entity and the correlation information among the name information of each entity, establishing the relation among the name information of each entity, and further constructing a knowledge graph in the field of nuclear fusion. The present disclosure provides a knowledge graph construction flowchart, as shown in fig. 2.

In fig. 2, data collection is a corpus collection process, entity recognition is a process of extracting entity name information from the corpus, relation extraction is a process of obtaining correlation information between the entity name information, entity linking is to link each entity name information to a corresponding node in a knowledge graph, and finally, a graph structure is constructed, that is, the nodes are connected by using edges representing the correlation information between the entity name information, so that the knowledge graph is obtained.

The specification also provides a schematic diagram of the knowledge graph, as shown in fig. 3.

In fig. 3, the existence of edges between entity name information represents that the connected entity name information has a correlation relationship (including a positive correlation relationship and a negative correlation relationship). For example, as can be seen from fig. 3, there is a positive correlation between the entity name information a and the entity name information B in fig. 3, and there is a negative correlation between the entity name information a and the entity name information C in fig. 3.

For the two entity name information with no connection edge in fig. 3, it represents that there is no positive correlation or no negative correlation between the two entity name information, that is, there is no correlation between the two entity name information. For example, as can be seen from fig. 3, there is no correlation between the entity name information C and the entity name information F.

S105: and carrying out task analysis on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field to obtain an analysis result, and executing target tasks according to the analysis result.

After the nuclear fusion knowledge graph is constructed, task analysis on experimental tasks can be realized. Taking the problem of experiment occurrence during the experiment task process as an example, after acquiring experiment task data, the terminal equipment can perform task analysis on the experiment task so as to determine which entity name information corresponding to the cause of the experiment problem exists through the constructed knowledge graph of the nuclear fusion field, and further recommend the entity name information to the terminal equipment used by the experimenters.

The above description presents a use scenario in the experimental process, and the present specification does not specifically limit the use scenario. It should be noted that, the target models (i.e., the first target model, the second target model, and the third target model) used in the process of constructing the nuclear fusion knowledge graph provide training methods in the specification, but in the actual application process, the existing general large language model can be used, and a large amount of corpus integrated by the large language model can be utilized, so that the richness of the constructed nuclear fusion knowledge graph is improved.

However, no matter which language model is used, there may be a problem that the output result does not match the input data, so that, optionally, the target language model may output a result that matches the input data more by the prompt word.

The output result of the target model is not matched with the input data, the target model particularly shows the constraint on the output result, the target model can output related results according to the input data, and some of the output results are needed for constructing the nuclear fusion knowledge graph, but also contain unnecessary results. Therefore, the output result of the target model can be limited by using the prompt word, so that the result with higher matching degree can be output, and if the existing general large language model is used, the randomness parameter of the large language model can be additionally changed in the process of calling the large language model through the interface, so that the output result of the general large language model is restrained.

It should be noted that, in the actual use of the target model, the prompt word often cannot make the target model directly output the target result when the target model is used for the first time. Therefore, in the use of the prompting word, when the result output by the target model is not the target result, the prompting word needs to be modified, that is, the target model outputs the target result by iterating the prompting word for a plurality of times.

Taking the third object model as an example, inputting the entity pair and the prompt word for describing the correlation between the two entity name information in the entity pair into the third object model, so that the third object model determines and outputs the correlation information between the two entity name information in the entity pair according to the prompt word for describing the correlation between the two entity name information in the entity pair.

After obtaining the correlation information between the two entity name information, text description contents corresponding to the two entity name information and a prompt word for describing the correlation between the two entity name information can be input into a preset third target model, so that the third target model determines and outputs the correlation information between the two entity name information according to the prompt word for describing the correlation between the two entity name information. And then, judging the correlation information output by the third target model, if the correlation information output by the third target model is determined to be the target correlation information between the two entity name information, outputting the correlation information, and if the correlation information is not the target correlation information, modifying the prompt word.

Wherein, in case that the correlation information between the two obtained entity name information is not the target correlation information, the hint words of the two entity name information can be modified. After modifying the prompting word, re-inputting the modified prompting word, the two entity name information and the text description content corresponding to the two entity name information into the third target model, so that the third target model outputs the correlation information of the two entity name information determined based on the modified prompting word until the output result of the third target model is the target correlation information.

For another example, after the first target model obtains the entity name information, the entity name information and the determined prompt word for the entity name information may be input into a preset first target model, so that the first target model determines text description content corresponding to the entity name information according to the determined prompt word for the entity name information. And then, judging the text description content output by the first target model, if the text description content output by the first target model is determined to be the target text description content of the entity name information, outputting the text description content, and if not, modifying the prompt word.

Wherein, in the case that the text description content corresponding to the obtained entity name information is not the target text description content, the prompt word of the entity name information can be modified. After modifying the prompting word, re-inputting the modified prompting word and the entity name information into the first target model, so that the first target model outputs text description content of the entity name information determined based on the modified prompting word until the output result of the first target model is the target text description content corresponding to the entity name information.

The task execution method based on the constructed nuclear fusion knowledge graph is based on the same thought, and the corresponding device, storage medium and electronic equipment are also provided.

Fig. 4 is a schematic structural diagram of a task execution device based on a constructed nuclear fusion knowledge graph according to an embodiment of the present disclosure, where the device includes:

the acquisition module 401 is used for acquiring information of each entity in the field of nuclear fusion;

a first determining module 402, configured to input, for each entity name information, the entity name information into a preset first target model, so as to output, through the first target model, text description content corresponding to the entity name information;

A second determining module 403, configured to determine relevance information between the entity names according to text description content corresponding to each entity name information;

a third determining module 404, configured to respectively correspond to different nodes with each entity name information, and determine edges between the nodes according to correlation information between the entity name information, so as to construct a knowledge graph in the nuclear fusion field;

and the control module 405 is configured to perform task analysis on the experimental task in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field, obtain an analysis result, and execute a target task according to the analysis result.

Optionally, the collecting module 401 is specifically configured to obtain corpus data in the nuclear fusion field; determining non-subclass entities from the corpus data as candidate entities; inputting entity name information corresponding to each candidate entity into a preset second target model for each candidate entity, and outputting a judging result of judging whether the candidate entity is an entity in the nuclear fusion field through the second target model; and determining each entity in the nuclear fusion field from each candidate entity according to the judging result output by the second target model aiming at each candidate entity, and acquiring entity name information of each determined entity in the nuclear fusion field.

Optionally, the first determining module 402 is specifically configured to, for each entity name information, input the entity name information and the determined prompting word for the entity name information into a preset first target model, so that the first target model outputs text description content corresponding to the entity name information based on the prompting word; and if the text description content corresponding to the entity name information is not the target description content corresponding to the entity name information, acquiring a modified prompt word aiming at the entity name information, and re-inputting the entity name information and the modified prompt word into the first target model, so that the first target model outputs the text description content corresponding to the entity name information based on the modified prompt word until the first target model outputs the target description content corresponding to the entity name information.

Optionally, the second determining module 403 is specifically configured to input, for any two pieces of entity name information, the two pieces of entity name information, text description content corresponding to the two pieces of entity name information, and a prompt word for describing a correlation between the two pieces of entity name information into a preset third target model, so that the third target model determines and outputs correlation information between the two pieces of entity name information according to the prompt word for describing the correlation between the two pieces of entity name information.

Optionally, the second determining module 403 is specifically configured to determine whether the correlation information between the two entity name information output by the third target model is target correlation information between the two entity name information; and responding to the fact that the correlation information between the two entity name information is not the target correlation information between the two entity name information, acquiring a modified prompt word for the two entity name information, and re-inputting text description contents corresponding to the two entity name information and the modified prompt word into the third target model, so that the third target model outputs the correlation information between the two entity name information based on the modified prompt word until the third target model outputs the target correlation information between the two entity name information.

The present disclosure also provides a computer readable storage medium storing a computer program which, when executed by a processor, is operable to perform the task execution method provided in fig. 1 and based on the constructed nuclear fusion knowledge graph.

Based on the task execution method based on the constructed nuclear fusion knowledge graph shown in fig. 1, the embodiment of the specification also provides a structural schematic diagram of the electronic device shown in fig. 5. At the hardware level, as in fig. 5, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile storage, although it may include hardware required for other services. The processor reads the corresponding computer program from the nonvolatile memory to the memory and then runs the computer program to realize the task execution method based on the constructed nuclear fusion knowledge graph as shown in the figure 1.

Of course, other implementations, such as logic devices or combinations of hardware and software, are not excluded from the present description, that is, the execution subject of the following processing flows is not limited to each logic unit, but may be hardware or logic devices.

In the 90 s of the 20 th century, improvements to one technology could clearly be distinguished as improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) or software (improvements to the process flow). However, with the development of technology, many improvements of the current method flows can be regarded as direct improvements of hardware circuit structures. Designers almost always obtain corresponding hardware circuit structures by programming improved method flows into hardware circuits. Therefore, an improvement of a method flow cannot be said to be realized by a hardware entity module. For example, a programmable logic device (Programmable Logic Device, PLD) (e.g., field programmable gate array (Field Programmable Gate Array, FPGA)) is an integrated circuit whose logic function is determined by the programming of the device by a user. A designer programs to "integrate" a digital system onto a PLD without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Moreover, nowadays, instead of manually manufacturing integrated circuit chips, such programming is mostly implemented by using "logic compiler" software, which is similar to the software compiler used in program development and writing, and the original code before the compiling is also written in a specific programming language, which is called hardware description language (Hardware Description Language, HDL), but not just one of the hdds, but a plurality of kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), lava, lola, myHDL, PALASM, RHDL (Ruby Hardware Description Language), etc., VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog are currently most commonly used. It will also be apparent to those skilled in the art that a hardware circuit implementing the logic method flow can be readily obtained by merely slightly programming the method flow into an integrated circuit using several of the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable logic controllers, and embedded microcontrollers, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic of the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller may thus be regarded as a kind of hardware component, and means for performing various functions included therein may also be regarded as structures within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.

The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer chip or entity, or by a product having a certain function. One typical implementation is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being functionally divided into various units, respectively. Of course, the functions of each element may be implemented in one or more software and/or hardware elements when implemented in the present specification.

It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present description is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the specification. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.

It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.

The foregoing is merely exemplary of the present disclosure and is not intended to limit the disclosure. Various modifications and alterations to this specification will become apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, or the like, which are within the spirit and principles of the present description, are intended to be included within the scope of the claims of the present description.

Claims (10)

1. The task execution method based on the constructed nuclear fusion knowledge graph is characterized by comprising the following steps of:

acquiring name information of each entity in the nuclear fusion field;

inputting the entity name information into a preset first target model aiming at each entity name information, so as to output text description content corresponding to the entity name information through the first target model;

Determining the correlation information among the entity name information according to the text description content corresponding to each entity name information;

each entity name information corresponds to different nodes respectively, and edges among the nodes are determined according to the correlation information among the entity name information, so that a knowledge graph in the nuclear fusion field is constructed;

and carrying out task analysis on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field to obtain an analysis result, and executing target tasks according to the analysis result.

2. The method of claim 1, wherein obtaining the name information of each entity in the field of nuclear fusion specifically comprises:

acquiring corpus data in the nuclear fusion field;

determining non-subclass entities from the corpus data as candidate entities;

inputting entity name information corresponding to each candidate entity into a preset second target model for each candidate entity, and outputting a judging result of judging whether the candidate entity is an entity in the nuclear fusion field through the second target model;

and determining each entity in the nuclear fusion field from each candidate entity according to the judging result output by the second target model aiming at each candidate entity, and acquiring entity name information of each determined entity in the nuclear fusion field.

3. The method of claim 1, wherein for each entity name information, inputting the entity name information into a preset first target model, so as to output text description content corresponding to the entity name information through the first target model, and specifically comprising:

inputting the entity name information and the determined prompt word for the entity name information into a preset first target model aiming at each entity name information, so that the first target model outputs text description content corresponding to the entity name information based on the prompt word;

and if the text description content corresponding to the entity name information is not the target description content corresponding to the entity name information, acquiring a modified prompt word aiming at the entity name information, and re-inputting the entity name information and the modified prompt word into the first target model, so that the first target model outputs the text description content corresponding to the entity name information based on the modified prompt word until the first target model outputs the target description content corresponding to the entity name information.

4. The method of claim 1, wherein determining the correlation information between the entity name information according to the text description content corresponding to each entity name information, specifically comprises:

Inputting the two entity name information, text description contents corresponding to the two entity name information and prompt words for describing the correlation between the two entity name information into a preset third target model aiming at any two entity name information, so that the third target model determines and outputs the correlation information between the two entity name information according to the prompt words for describing the correlation between the two entity name information.

5. The method of claim 4, wherein the method further comprises:

judging whether the correlation information between the two entity name information output by the third target model is the target correlation information between the two entity name information or not;

and responding to the fact that the correlation information between the two entity name information is not the target correlation information between the two entity name information, acquiring a modified prompt word for the two entity name information, and re-inputting text description contents corresponding to the two entity name information and the modified prompt word into the third target model, so that the third target model outputs the correlation information between the two entity name information based on the modified prompt word until the third target model outputs the target correlation information between the two entity name information.

6. The task execution device based on the constructed nuclear fusion knowledge graph is characterized by comprising:

the acquisition module is used for acquiring information of each entity in the nuclear fusion field;

the first determining module is used for inputting the entity name information into a preset first target model aiming at each entity name information so as to output text description content corresponding to the entity name information through the first target model;

the second determining module is used for determining the correlation information among the entity names according to the text description content corresponding to each entity name information;

the third determining module is used for respectively corresponding different nodes according to the name information of each entity and determining edges among the nodes according to the correlation information among the name information of each entity so as to construct a knowledge graph in the nuclear fusion field;

and the control module is used for carrying out task analysis on experimental tasks in the nuclear fusion field according to the constructed knowledge graph in the nuclear fusion field to obtain an analysis result, and executing target tasks according to the analysis result.

7. The apparatus of claim 6, wherein the acquisition module is configured to,

Acquiring corpus data in the nuclear fusion field;

determining non-subclass entities from the corpus data as candidate entities;

inputting entity name information corresponding to each candidate entity into a preset second target model for each candidate entity, and outputting a judging result of judging whether the candidate entity is an entity in the nuclear fusion field through the second target model;

and determining each entity in the nuclear fusion field from each candidate entity according to the judging result output by the second target model aiming at each candidate entity, and acquiring entity name information of each determined entity in the nuclear fusion field.

8. The apparatus of claim 6, wherein the first determining means is specifically configured to,

inputting the entity name information and the determined prompt word for the entity name information into a preset first target model aiming at each entity name information, so that the first target model outputs text description content corresponding to the entity name information based on the prompt word;

and if the text description content corresponding to the entity name information is not the target description content corresponding to the entity name information, acquiring a modified prompt word aiming at the entity name information, and re-inputting the entity name information and the modified prompt word into the first target model, so that the first target model outputs the text description content corresponding to the entity name information based on the modified prompt word until the first target model outputs the target description content corresponding to the entity name information.

9. A computer-readable storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the method of any of the preceding claims 1-5.

10. An electronic device comprising a processor and a computer program stored on a memory and executable on the processor, characterized in that the processor implements the method of any of the preceding claims 1-5 when executing the program.