CN112015973A - Relation reasoning method and terminal for heterogeneous network - Google Patents

Abstract

The invention provides a relation reasoning method and a terminal of a heterogeneous network, wherein the method comprises the following steps: receiving query information input by a user; reasoning the query information in the heterogeneous network based on a preset reasoning strategy to obtain a target reasoning result; the heterogeneous network comprises a basic network layer and a high-level relation network layer, wherein the basic network layer takes people, affairs, places and objects as network nodes, and network layers are established by taking human-to-human, human-to-affairs, human-to-ground and human-to-object as network relations; the high-level relation network layer is a network layer with probability weight established by mining generated relations. The embodiment of the invention does not need to establish a plurality of network topologies aiming at different types of data, thereby reducing the difficulty of reasoning.

Description

A relational reasoning method and terminal for heterogeneous networks

technical field

The present invention relates to the technical field of communications, and in particular, to a relational reasoning method and terminal for heterogeneous networks.

Background technique

In the context of the popularization of the Internet, the online and offline behavior data of users is also increasing day by day, and PB-level data is generated every day. In policing scenarios, screening and reasoning of suspicious information from massive data has become a technical and business problem. If the data types are classified as people-things-land-things, a complex and heterogeneous network topology is generated every day. In order to gain insight into user intentions, analyze user behavior, and analyze the relationship between users and all things, in the prior art, a common relational reasoning method is relational reasoning based on isomorphic networks. However, for nodes of different types (heterogeneous), multiple network topologies need to be constructed, so reasoning is difficult.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide a relational reasoning method and terminal for heterogeneous networks, so as to solve the problem of relatively difficult reasoning.

In a first aspect, an embodiment of the present invention provides a relational reasoning method for heterogeneous networks, including:

Receive user input query information;

Inferring the query information in a heterogeneous network based on a preset inference strategy to obtain a target inference result;

Wherein, the preset reasoning strategy includes at least one of event reasoning strategy, rule reasoning strategy and conditional reasoning strategy, and the heterogeneous network includes a basic network layer and a high-level relational network layer, and the basic network layer is based on human, Things, places, and things are network nodes, and the network layer is established with people-to-people, people-to-things, people-to-land, and people-to-things. Network layer.

In a second aspect, an embodiment of the present invention further provides a terminal, including:

a receiving module for receiving query information input by a user;

an inference module, configured to infer the query information in a heterogeneous network based on a preset inference strategy to obtain a target inference result;

Wherein, the preset reasoning strategy includes at least one of event reasoning strategy, rule reasoning strategy and conditional reasoning strategy, and the heterogeneous network includes a basic network layer and a high-level relational network layer, and the basic network layer is based on human, Things, places, and things are network nodes, and the network layer is established with people-to-people, people-to-things, people-to-land, and people-to-things. Network layer.

In a third aspect, an embodiment of the present invention further provides a terminal, including a processor, a memory, and a computer program stored on the memory and executable on the processor, where the computer program is executed by the processor The steps to implement the above relational reasoning method for heterogeneous networks.

In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the above method for relational reasoning in a heterogeneous network.

In the embodiment of the present invention, since different types of data are uniformly abstracted into network nodes and relationships to form a heterogeneous network, and according to the query information input by the user, at least one of an event reasoning strategy, a rule reasoning strategy, and a conditional reasoning strategy is used. The corresponding reasoning results are obtained by querying in the heterogeneous network, so as to realize the relational reasoning of the heterogeneous network. There is no need to build multiple network topologies for different types of data, which reduces the difficulty of reasoning.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without creative labor.

FIG. 1 is a flowchart of a relational reasoning method for heterogeneous networks provided by an embodiment of the present invention;

2 is a schematic structural diagram of a Rete algorithm in a relational reasoning method for heterogeneous networks provided by an embodiment of the present invention;

3 is an architecture diagram of a rule inference strategy of a relational reasoning method for heterogeneous networks provided by an embodiment of the present invention;

4 is an example diagram of a conditional reasoning engine in a relational reasoning method for heterogeneous networks provided by an embodiment of the present invention;

5 is a structural diagram of a terminal provided by an embodiment of the present invention;

FIG. 6 is a structural diagram of a terminal provided by another embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Referring to FIG. 1, FIG. 1 is a flowchart of a relationship reasoning method for heterogeneous networks provided by an embodiment of the present invention. As shown in FIG. 1, the method includes the following steps:

Step 101, receiving query information input by a user;

Step 102, inferring the query information in a heterogeneous network based on a preset inference strategy, to obtain a target inference result;

Wherein, the preset reasoning strategy includes at least one of event reasoning strategy, rule reasoning strategy and conditional reasoning strategy, and the heterogeneous network includes a basic network layer and a high-level relational network layer, and the basic network layer is based on human, Things, places, and things are network nodes, and the network layer is established with people-to-people, people-to-things, people-to-land, and people-to-things. Network layer.

In the embodiment of the present invention, data can be organized and managed based on a graph database (large-scale heterogeneous network that has been graphed), with people, things, places, and things as network nodes, and people to people, people to things, people The basic network layer is established for the network relationship with the land and people and things. For the relationship generated by mining, such as suspected gang relationship, the second layer is an advanced relationship network layer with probability weights. For example, user A watched a movie D in shopping mall C, user A, shopping mall C and movie D can be set as network nodes in the basic network layer, and the relationship between user A and other network nodes constitutes the hierarchical relationship of the network . In addition, user A and user B take a flight at the same time, and the probability that user A and user B travel together is determined as the first preset percentage in the advanced relationship network layer. Specifically, the determination of the relationship and the determination of the probability weight may be determined according to different strategies, which will not be further described herein.

The above event reasoning strategy refers to analyzing specific facts. For example, in an optional embodiment, both input conditions and data can be regarded as events (subject predicate object (SPO) triples), and reasoning The process of transforming into a process in which the input SPOs are matched in the base network layer. Specifically, the query information input by the user is the input SPO. At this time, the query information input by the user needs to be semantically parsed to obtain a query statement of the SPO structure that the system can recognize, and then query to obtain corresponding results.

The above rule inference strategy refers to filtering out data matching the rules corresponding to the query information in the basic network layer.

The above conditional reasoning strategy is based on Bayesian reasoning, which performs probabilistic reasoning on high-level networks.

In the embodiment of the present invention, since different types of data are uniformly abstracted into network nodes and relationships to form a heterogeneous network, and according to the query information input by the user, at least one of an event reasoning strategy, a rule reasoning strategy, and a conditional reasoning strategy is used. The corresponding reasoning results are obtained by querying in the heterogeneous network, so as to realize the relational reasoning of the heterogeneous network. There is no need to build multiple network topologies for different types of data, which reduces the difficulty of reasoning.

It should be noted that the above-mentioned preset reasoning strategies may include some or all of event reasoning strategies, rule reasoning strategies and conditional reasoning strategies. Each of these inference strategies corresponds to an outcome. Specifically, one or more of the event reasoning strategy, the rule reasoning strategy, and the conditional reasoning strategy may be used for reasoning according to the query information input by the user, for example, similar to querying specific facts (eg, the father of user A's wife) can be obtained by reasoning through the event reasoning strategy. When similarly querying possible events (inquiring about people known to user A and user B), three inference results can be obtained according to the event inference strategy, the rule inference strategy and the conditional inference strategy, and the three inference results together constitute the above target inference results.

In an optional embodiment, when the preset reasoning strategy includes the event reasoning strategy, the above step 102 includes:

Step 1021, performing semantic analysis on the query information to obtain a query statement composed of subject-verb-object SPO triples;

Step 1022, matching in the basic network layer according to the query statement to obtain a matching result;

Wherein, the target inference result includes the matching result.

In this embodiment, the event reasoning strategy mainly converts the expression of events (query information) into the form of triples, and traverses the basic network layer through the event reasoning engine to obtain the final reasoning conclusion. The event inference engine is a component embedded in the application, which realizes the separation of business decisions from the application code, saves various complex if-else, and judges and uses predefined semantic modules to write business decisions. Receive data input from users, interpret business rules, and make business decisions based on business rules.

The above event reasoning engine has done a lot of optimization work mainly based on the RETE algorithm, which provides an efficient implementation of the expert system. The Rete algorithm is a fast forward rule matching algorithm. The black box explanation of the system shown in Figure 2: the left side is the addition and deletion of rules; the right side is the addition and deletion of facts. The rule is composed of left and right components, the left component is composed of rule units or function units, and the right component is composed of action units. The firing of a rule, the fact that the condition on the left is satisfied, triggers the action on the right.

It should be understood that in the Rete algorithm, related technologies may be referred to for the structure of the Rete network, for example, the Rete network includes an alpha part and a beta part, and the alpha part performs constant tests on the WME. The output of the alpha node is saved in alpha memories (AM). For example, alpha memories for the condition (<x>on<y>) will store WMEs with fields equal to on. The beta part mainly contains join nodes and beta memories. Join nodes are used to perform consistency checks on variable bindings between conditions. Beta memories hold instances of partial productions (combinations of WMEs that satisfy the conditions of some, but not all, of the productions). These partial instantiations are called tokens. The Alpha network performs all detections involving one WME (greater than, equal to, less than, variable, fixed), and the Beta network performs detections involving more than two WMEs.

In another optional embodiment, when the preset reasoning strategy includes the rule reasoning strategy, the step 102 includes:

Step 1023, identifying the query information to obtain a first fact, where the first fact includes the entity corresponding to the network node, the relationship information of the entity, and the attribute information of the entity;

Step 1024: Filter the first fact according to the attribute constraint condition, and obtain a second fact in the first fact that satisfies the attribute constraint condition;

Step 1025, based on the basic network layer, transmit the first fact in the rule network to obtain a rule inference result;

Wherein, the target inference result includes the rule inference result, and the rule network is established according to target rules matching the query information in a preset rule base.

For example, in the above query information, the information of user E needs to be queried, and the above first fact may include singer E, actor E, and writer E. If the above attribute constraints are singers, then actor E and writer E can be filtered out according to the attribute constraints. The attribute constraint condition may be determined according to the above query information. For example, when querying the information of user E who performs on the same stage as singer A, the attribute constraint can be determined to be singer. In addition, the above attribute constraint conditions may also include other attribute constraint information, which is not further limited herein.

Further, based on the foregoing embodiment, in this embodiment, in the process of transmitting the first fact in the rule network, the method further includes:

According to the attribute constraint conditions, attribute constraints are performed on the intermediate facts generated during the process of transmitting the first fact in the rule network.

It should be understood that, in this embodiment of the present invention, an intermediate fact will be generated in the process of transmitting the first fact in the rule network, and the intermediate fact may include the entity corresponding to the network node, the The relationship information of the entity and the attribute information of the entity. After adding new intermediate facts, the intermediate facts can also be constrained according to the attribute constraints, that is, after filtering out the facts that do not meet the attribute constraints in the intermediate facts, the intermediate facts continue to be transmitted in the rule network until the final result is obtained. Fact, the final fact is the rule inference result.

In this embodiment of the present invention, the facts introduced into the network are dynamically constrained, so that the time for fact reasoning can be reduced and the efficiency of reasoning can be improved.

The specific implementation of the rule inference strategy can be referred to as shown in FIG. 3 , and each process module in FIG. 3 is described in detail below:

When performing rule parsing, you can read the target rules matching the query information in the rule base one by one, and parse the rules according to the representation method of the rules, so as to convert the target rules into the data format required by the rule service system RuleBase, and call the interface of the rule service system. Add rules for the rule service system to build a rule network.

During the validity check, the first fact and the intermediate fact generated during the execution of the rule are checked for validity through attribute constraints.

The core module of the inference strategy is the rule inference engine, which completes the efficient matching of facts and rules. A variety of rule matching algorithms can be used for specific implementation, mainly including two stages of rule editing and runtime execution. The rule editor uses the rules in the rule base to generate a network of rules, and the runtime execution propagates the fact of joining in the network.

The processing process of the query module Query is to transmit the facts to the rule service system to trigger the derivation of the rules.

There are many kinds of rule instances in the rule network, and the management module (for example, Agenda can be used. Agenda is a distributed timed task management system based on mongodb database implemented by nodejs.) The execution order of the activated rule instances can be determined according to a certain strategy, and execute the rule instance. At the same time, Agenda can also perform conflict resolution. For specific conflict resolution solutions, reference may be made to related technologies, which will not be repeated here.

Since the addition of infrequent data will generate a lot of intermediate data for inference, the garbage collection module is used to clean up the intermediate data/memory resources of inference in batches. Specifically, it can be recycled regularly.

Further, in yet another embodiment, when the above-mentioned preset reasoning strategy includes the conditional reasoning strategy, the step 102 includes:

Using the conditional reasoning engine of Bayesian reasoning, the conditional reasoning result corresponding to the query information is obtained by reasoning from the basic network layer and the advanced relational network layer;

Wherein, the target reasoning result includes the rule reasoning result. As shown in FIG. 4 , the conditional reasoning engine includes a multi-tree propagation inference engine, a cluster tree propagation inference engine, a method engine based on combinatorial optimization, and a method engine based on search. and at least one of the Monte Carlo Mante Carlo algorithm engines.

In the embodiment of the present invention, the above-mentioned multi-tree propagation inference engine, clique-tree propagation inference engine, and method engine based on combinatorial optimization are used to realize accurate inference, that is, it is desirable to calculate the exact value of the marginal distribution or conditional distribution of the target variable. The computational complexity of the class algorithm grows exponentially with the size of very large cliques, so it is only applicable when the size of the Bayesian network is small.

The search-based method engine and the Mante Carlo algorithm engine are used to realize approximate inference. When the scale of the network is large, approximate inference is often used. The approximate inference algorithm can obtain the approximate solution of the original problem with low time complexity.

Specifically, the specific implementation of the multi-tree propagation inference engine, the clique tree propagation inference engine, the method engine based on combinatorial optimization, the method engine based on search, and the Mante Carlo algorithm engine can refer to related technologies, and will not be described in detail here.

It should be noted that the various optional implementation manners introduced in the embodiments of the present invention may be implemented in combination with each other, or may be implemented independently, which are not limited by the embodiments of the present invention.

Referring to FIG. 5, FIG. 5 is a structural diagram of a terminal provided by an embodiment of the present invention. As shown in FIG. 5, the terminal 500 includes:

A receiving module 501, configured to receive query information input by a user;

an inference module 502, configured to infer the query information in a heterogeneous network based on a preset inference strategy to obtain a target inference result;

Wherein, the preset reasoning strategy includes at least one of event reasoning strategy, rule reasoning strategy and conditional reasoning strategy, and the heterogeneous network includes a basic network layer and a high-level relational network layer, and the basic network layer is based on human, Things, places, and things are network nodes, and the network layer is established with people-to-people, people-to-things, people-to-land, and people-to-things. Network layer.

Optionally, when the preset reasoning strategy includes the event reasoning strategy, the reasoning module 502 includes:

a parsing unit, configured to perform semantic parsing on the query information to obtain a query statement composed of subject-verb-object SPO triples;

a matching unit, configured to match in the basic network layer according to the query statement to obtain a matching result;

Wherein, the target inference result includes the matching result.

Optionally, when the preset reasoning strategy includes the rule reasoning strategy, the reasoning module 502 includes:

an identification unit, configured to identify the query information to obtain a first fact, where the first fact includes an entity corresponding to the network node, relationship information of the entity, and attribute information of the entity;

a control unit, configured to filter the first fact according to attribute constraints, and obtain a second fact in the first fact that satisfies the attribute constraints;

an inference unit, configured to transmit the first fact in the rule network based on the basic network layer to obtain a rule inference result;

Wherein, the target inference result includes the rule inference result, and the rule network is established according to target rules matching the query information in a preset rule base.

Optionally, the control unit is further configured to, in the process of transmitting the first fact in the rule network, in the process of transmitting the first fact in the rule network according to the attribute constraint condition. The resulting intermediate facts are subject to attribute constraints.

Optionally, when the preset reasoning strategy includes the conditional reasoning strategy, the reasoning module 502 is configured to: use a conditional reasoning engine of Bayesian reasoning to infer from the basic network layer and the advanced relational network layer. the conditional reasoning result corresponding to the query information;

Wherein, the target inference result includes the rule inference result, and the conditional inference engine includes a multi-tree propagation inference engine, a clique tree propagation inference engine, a combinatorial optimization-based method engine, a search-based method engine, and a Monte Carlo Mante Carlo method engine. At least one of the algorithm engines.

The terminal provided in the embodiment of the present invention can implement each process implemented by the terminal in the method embodiments of FIG. 1 to FIG. 4 , and to avoid repetition, details are not described here.

FIG. 6 is a schematic diagram of a hardware structure of a terminal implementing various embodiments of the present invention.

The terminal 600 includes but is not limited to: a radio frequency unit 601, a network module 602, an audio output unit 603, an input unit 604, a sensor 605, a display unit 606, a user input unit 607, an interface unit 608, a memory 609, a processor 610, and a power supply 611 and other components. Those skilled in the art can understand that the terminal structure shown in FIG. 6 does not constitute a limitation on the terminal, and the terminal may include more or less components than the one shown, or combine some components, or arrange different components. In the embodiment of the present invention, the terminal includes but is not limited to a mobile phone, a tablet computer, a notebook computer, a palmtop computer, a vehicle-mounted terminal, a wearable device, a pedometer, and the like.

The processor 610 is configured to receive query information input by a user;

Inferring the query information in a heterogeneous network based on a preset inference strategy to obtain a target inference result;

Wherein, the preset reasoning strategy includes at least one of event reasoning strategy, rule reasoning strategy and conditional reasoning strategy, and the heterogeneous network includes a basic network layer and a high-level relational network layer, and the basic network layer is based on human, Things, places, and things are network nodes, and the network layer is established with people-to-people, people-to-things, people-to-land, and people-to-things. Network layer.

Optionally, when the preset reasoning strategy includes the event reasoning strategy, the processor 610 is configured to: perform semantic analysis on the query information to obtain a query statement composed of subject-verb-object SPO triples; according to the The query sentence is matched in the basic network layer, and a matching result is obtained; wherein, the target inference result includes the matching result.

Optionally, when the preset inference strategy includes the rule inference strategy, the processor 610 is configured to:

Identifying the query information to obtain a first fact, where the first fact includes an entity corresponding to the network node, relationship information of the entity, and attribute information of the entity;

Filter the first fact according to the attribute constraint, and obtain a second fact in the first fact that satisfies the attribute constraint;

Based on the basic network layer, the first fact is transmitted in the rule network to obtain a rule inference result;

Wherein, the target inference result includes the rule inference result, and the rule network is established according to target rules matching the query information in a preset rule base.

Optionally, the processor 610 is further configured to: in the process of transmitting the first fact in the rule network, generate the first fact in the process of transmitting the first fact in the rule network according to the attribute constraint condition attribute constraints on the intermediate facts.

Optionally, when the preset reasoning strategy includes the conditional reasoning strategy, the processor 610 is configured to: use a conditional reasoning engine of Bayesian reasoning to infer the obtained information from the basic network layer and the advanced relational network layer. The conditional inference result corresponding to the query information; wherein, the target inference result includes the rule inference result, and the conditional inference engine includes a multi-tree propagation inference engine, a cluster tree propagation inference engine, a method engine based on combinatorial optimization, and a search-based inference engine. At least one of the method engine and the Monte Carlo Mante Carlo algorithm engine.

In the embodiment of the present invention, since different types of data are uniformly abstracted into network nodes and relationships to form a heterogeneous network, and according to the query information input by the user, at least one of an event reasoning strategy, a rule reasoning strategy, and a conditional reasoning strategy is used. The corresponding reasoning results are obtained by querying in the heterogeneous network, so as to realize the relational reasoning of the heterogeneous network. There is no need to build multiple network topologies for different types of data, which reduces the difficulty of reasoning.

It should be understood that, in this embodiment of the present invention, the radio frequency unit 601 may be used for receiving and sending signals during sending and receiving of information or during a call. Specifically, after receiving the downlink data from the base station, it is processed by the processor 610; The uplink data is sent to the base station. Generally, the radio frequency unit 601 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, and the like. In addition, the radio frequency unit 601 can also communicate with the network and other devices through a wireless communication system.

The terminal provides the user with wireless broadband Internet access through the network module 602, such as helping the user to send and receive emails, browse web pages, and access streaming media.

The audio output unit 603 may convert audio data received by the radio frequency unit 601 or the network module 602 or stored in the memory 609 into audio signals and output as sound. Also, the audio output unit 603 may also provide audio output related to a specific function performed by the terminal 600 (eg, call signal reception sound, message reception sound, etc.). The audio output unit 603 includes a speaker, a buzzer, a receiver, and the like.

The input unit 604 is used to receive audio or video signals. The input unit 604 may include a graphics processor (Graphics Processing Unit, GPU) 6041 and a microphone 6042, and the graphics processor 6041 captures images of still pictures or videos obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode data is processed. The processed image frames may be displayed on the display unit 606 . The image frames processed by the graphics processor 6041 may be stored in the memory 609 (or other storage medium) or transmitted via the radio frequency unit 601 or the network module 602 . The microphone 6042 can receive sound and can process such sound into audio data. The processed audio data can be converted into a format that can be transmitted to a mobile communication base station via the radio frequency unit 601 for output in the case of a telephone call mode.

Terminal 600 also includes at least one sensor 605, such as light sensors, motion sensors, and other sensors. Specifically, the light sensor includes an ambient light sensor and a proximity sensor, wherein the ambient light sensor can adjust the brightness of the display panel 6061 according to the brightness of the ambient light, and the proximity sensor can turn off the display panel 6061 and/or when the terminal 600 is moved to the ear. or backlight. As a type of motion sensor, the accelerometer sensor can detect the magnitude of acceleration in all directions (generally three axes), and can detect the magnitude and direction of gravity when stationary, and can be used to identify the terminal posture (such as horizontal and vertical screen switching, related games, The sensor 605 may also include a fingerprint sensor, a pressure sensor, an iris sensor, a molecular sensor, a gyroscope, a barometer, a hygrometer, a thermometer, an infrared Sensors, etc., will not be repeated here.

The display unit 606 is used to display information input by the user or information provided to the user. The display unit 606 may include a display panel 6061, and the display panel 6061 may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

The user input unit 607 may be used to receive input numerical or character information, and generate key signal input related to user settings and function control of the terminal. Specifically, the user input unit 607 includes a touch panel 6071 and other input devices 6072 . The touch panel 6071, also referred to as a touch screen, can collect the user's touch operations on or near it (such as the user's finger, stylus, etc., any suitable object or accessory on or near the touch panel 6071). operate). The touch panel 6071 may include two parts, a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into contact coordinates, and then sends it to the touch controller. To the processor 610, the command sent by the processor 610 is received and executed. In addition, the touch panel 6071 can be implemented in various types such as resistive, capacitive, infrared, and surface acoustic waves. In addition to the touch panel 6071 , the user input unit 607 may also include other input devices 6072 . Specifically, other input devices 6072 may include, but are not limited to, physical keyboards, function keys (such as volume control keys, switch keys, etc.), trackballs, mice, and joysticks, which are not described herein again.

Further, the touch panel 6071 can be covered on the display panel 6061. When the touch panel 6071 detects a touch operation on or near it, it transmits it to the processor 610 to determine the type of the touch event, and then the processor 610 determines the type of the touch event according to the touch The type of event provides a corresponding visual output on the display panel 6061. Although in FIG. 6, the touch panel 6071 and the display panel 6061 are used as two independent components to realize the input and output functions of the terminal, in some embodiments, the touch panel 6071 and the display panel 6061 can be integrated to form a Realize the input and output functions of the terminal, which is not limited here.

The interface unit 608 is an interface for connecting an external device to the terminal 600 . For example, external devices may include wired or wireless headset ports, external power (or battery charger) ports, wired or wireless data ports, memory card ports, ports for connecting devices with identification modules, audio input/output (I/O) ports, video I/O ports, headphone ports, and more. The interface unit 608 may be used to receive input (eg, data information, power, etc.) from an external device and transmit the received input to one or more elements within the terminal 600 or may be used to communicate between the terminal 600 and the external device. transfer data between.

The memory 609 may be used to store software programs as well as various data. The memory 609 may mainly include a stored program area and a stored data area, wherein the stored program area may store an operating system, an application program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; Data created by the use of the mobile phone (such as audio data, phone book, etc.), etc. Additionally, memory 609 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The processor 610 is the control center of the terminal, uses various interfaces and lines to connect various parts of the entire terminal, and executes by running or executing the software programs and/or modules stored in the memory 609, and calling the data stored in the memory 609. Various functions of the terminal and processing data, so as to monitor the terminal as a whole. The processor 610 may include one or more processing units; preferably, the processor 610 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem The processor mainly handles wireless communication. It can be understood that, the above-mentioned modulation and demodulation processor may not be integrated into the processor 610.

The terminal 600 may also include a power supply 611 (such as a battery) for supplying power to various components. Preferably, the power supply 611 may be logically connected to the processor 610 through a power management system, so as to manage charging, discharging, and power consumption management through the power management system. Function.

In addition, the terminal 600 includes some unshown functional modules, which will not be repeated here.

Preferably, an embodiment of the present invention further provides a terminal, including a processor 610, a memory 609, a computer program stored in the memory 609 and running on the processor 610, and the computer program is implemented when the processor 610 executes it. The various processes of the above embodiments of the method for relational reasoning in heterogeneous networks can achieve the same technical effect, and are not repeated here to avoid repetition.

Embodiments of the present invention further provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, each process of the above embodiment of the method for relational reasoning in heterogeneous networks is implemented, and can To achieve the same technical effect, in order to avoid repetition, details are not repeated here. The computer-readable storage medium is, for example, a read-only memory (Read-Only Memory, ROM for short), a random access memory (Random Access Memory, RAM for short), a magnetic disk or an optical disk, and the like.

Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of the present invention.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described systems, devices and units may refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

In the embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solutions in the embodiments of the present invention.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as independent products, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk and other mediums that can store program codes.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (12)

1. A relationship inference method for a heterogeneous network, comprising:

receiving query information input by a user;

reasoning the query information in the heterogeneous network based on a preset reasoning strategy to obtain a target reasoning result;

the heterogeneous network comprises a basic network layer and a high-level relation network layer, wherein the basic network layer takes people, affairs, places and objects as network nodes, and network layers are established by taking human-to-human, human-to-affairs, human-to-ground and human-to-object as network relations; the high-level relation network layer is a network layer with probability weight established by mining generated relations.

2. The method according to claim 1, wherein when the preset inference policy includes the event inference policy, the inferring the query information in the heterogeneous network based on the preset inference policy, and obtaining a target inference result includes:

performing semantic analysis on the query information to obtain a query statement consisting of a Supper-predicate object (SPO) triple;

matching in the basic network layer according to the query statement to obtain a matching result;

wherein the target inference result comprises the matching result.

3. The method according to claim 1, wherein when the preset inference policy includes the rule inference policy, the inferring the query information in the heterogeneous network based on the preset inference policy, and obtaining a target inference result includes:

identifying the query information to obtain a first fact, wherein the first fact comprises an entity corresponding to the network node, relationship information of the entity and attribute information of the entity;

screening the first facts according to attribute constraint conditions to obtain second facts meeting the attribute constraint conditions in the first facts;

transmitting the first fact in a rule network based on the basic network layer to obtain a rule reasoning result;

and the target inference result comprises the rule inference result, and the rule network is established according to a target rule matched with the query information in a preset rule base.

4. The method of claim 3, wherein in communicating the first fact in a regular network, the method further comprises:

and performing attribute constraint on intermediate facts generated in the process of transmitting the first fact in the regular network according to the attribute constraint conditions.

5. The method according to claim 1, wherein when the preset inference policy includes the conditional inference policy, the inferring the query information in the heterogeneous network based on the preset inference policy, and obtaining a target inference result includes:

a conditional inference engine of Bayesian inference is utilized to infer from the basic network layer and the advanced relationship network layer to obtain a conditional inference result corresponding to the query information;

wherein the target inference result comprises the rule inference result, and the conditional inference engine comprises at least one of a multi-tree propagation inference engine, a clique-tree propagation inference engine, a combinatorial optimization-based method engine, a search-based method engine, and a Monte Carlo algorithm engine.

6. A terminal, comprising:

the receiving module is used for receiving query information input by a user;

the reasoning module is used for reasoning the query information in the heterogeneous network based on a preset reasoning strategy to obtain a target reasoning result;

the heterogeneous network comprises a basic network layer and a high-level relation network layer, wherein the basic network layer takes people, affairs, places and objects as network nodes, and network layers are established by taking human-to-human, human-to-affairs, human-to-ground and human-to-object as network relations; the high-level relation network layer is a network layer with probability weight established by mining generated relations.

7. The terminal according to claim 6, wherein when the preset inference policy comprises the event inference policy, the inference module comprises:

the analysis unit is used for carrying out semantic analysis on the query information to obtain a query statement consisting of a principal and predicate object oriented Service (SPO) triple;

the matching unit is used for matching in the basic network layer according to the query statement to obtain a matching result;

wherein the target inference result comprises the matching result.

8. The terminal according to claim 6, wherein when the preset inference policy comprises the rule inference policy, the inference module comprises:

the identification unit is used for identifying the query information to obtain a first fact, wherein the first fact comprises an entity corresponding to the network node, the relationship information of the entity and the attribute information of the entity;

the control unit is used for screening the first facts according to attribute constraint conditions to obtain second facts meeting the attribute constraint conditions in the first facts;

the inference unit is used for transmitting the first fact in a rule network based on the basic network layer to obtain a rule inference result;

and the target inference result comprises the rule inference result, and the rule network is established according to a target rule matched with the query information in a preset rule base.

9. The terminal according to claim 8, wherein the control unit is further configured to, during the transferring of the first fact in the regular network, perform attribute constraint on an intermediate fact generated during the transferring of the first fact in the regular network according to the attribute constraint condition.

10. The terminal of claim 6, wherein when the preset inference policy comprises the conditional inference policy, the inference module is configured to: a conditional inference engine of Bayesian inference is utilized to infer from the basic network layer and the advanced relationship network layer to obtain a conditional inference result corresponding to the query information;

wherein the target inference result comprises the rule inference result, and the conditional inference engine comprises at least one of a multi-tree propagation inference engine, a clique-tree propagation inference engine, a combinatorial optimization-based method engine, a search-based method engine, and a Monte Carlo algorithm engine.

11. A terminal, characterized in that it comprises a processor, a memory and a computer program stored on said memory and executable on said processor, said computer program, when executed by said processor, implementing the steps of the method of relational inference of heterogeneous networks according to any of claims 1-5.

12. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for relational inference of heterogeneous networks according to any one of claims 1 to 5.

Images