CN110955769B - Method for constructing processing stream and electronic equipment - Google Patents

Method for constructing processing stream and electronic equipment Download PDF

Info

Publication number
CN110955769B
CN110955769B CN201911301153.9A CN201911301153A CN110955769B CN 110955769 B CN110955769 B CN 110955769B CN 201911301153 A CN201911301153 A CN 201911301153A CN 110955769 B CN110955769 B CN 110955769B
Authority
CN
China
Prior art keywords
processing
processing state
state
pair
target problem
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201911301153.9A
Other languages
Chinese (zh)
Other versions
CN110955769A (en
Inventor
赵国光
闫晓芳
胡长建
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Lenovo Beijing Ltd
Original Assignee
Lenovo Beijing Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Lenovo Beijing Ltd filed Critical Lenovo Beijing Ltd
Priority to CN201911301153.9A priority Critical patent/CN110955769B/en
Publication of CN110955769A publication Critical patent/CN110955769A/en
Application granted granted Critical
Publication of CN110955769B publication Critical patent/CN110955769B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • G06F16/3329—Natural language query formulation or dialogue systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
    • G06F2216/03—Data mining
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Artificial Intelligence (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The application provides a construction method and electronic equipment of a processing flow, which are used for constructing the processing flow aiming at a target problem based on processing tracks by acquiring a dialogue log aiming at the target problem, determining processing states from the dialogue log and mining the processing tracks from the dialogue log based on the processing states, wherein the processing flow is a processing flow comprising all or part of processing states for processing the target problem, each processing state is used for representing one processing node of the target problem, each processing state in the processing flow can generate output information based on input information related to the target problem, the target problem can be processed through sequential execution of the processing states in the processing flow, and the application can construct the processing flow through the acquired dialogue log aiming at the target problem, so that the intelligent dialogue system is assisted in processing the target problem, and the efficiency of processing the target problem is improved.

Description

Method for constructing processing stream and electronic equipment
Technical Field
The invention relates to the technical field of intelligent conversations, in particular to a method for constructing a processing flow and electronic equipment.
Background
At present, a method for providing a user with a solution to a problem is a large application scenario of an intelligent dialogue system, and the main form is that the user presents the problem, and the intelligent dialogue system processes the problem presented by the user through multiple rounds of interaction based on a preset processing flow.
When the intelligent dialogue system cannot solve the problem of the user, the user generally resorts to a manual customer service or other systems. When a manual customer service or other systems solve the target problem proposed by the user, different solutions exist due to different situations encountered by different users.
How to assist the intelligent dialogue system based on the process flow of solving the problem for the user by the manual customer service or other systems is a technical problem to be solved by those skilled in the art.
Disclosure of Invention
In view of the above, the present invention provides a method for constructing a processing flow and an electronic device, so as to solve the above technical problems.
In order to achieve the above purpose, the present invention provides the following technical solutions:
a method of constructing a process flow, comprising:
acquiring a dialogue log aiming at a target problem;
determining a processing state from the dialogue log, and mining a processing track from the dialogue log based on the processing state;
and constructing a processing flow aiming at the target problem based on the processing track, wherein the processing flow is a processing flow comprising all or part of processing states for processing the target problem, each processing state is used for representing one processing node of the target problem, each processing state in the processing flow can generate output information based on input information related to the target problem, and the target problem can be processed through sequential execution of the processing states in the processing flow.
Preferably, the constructing a processing flow for the target problem based on the processing track includes:
constructing a processing state pair set meeting a preset condition based on the processing track; wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states;
determining a maximum set of processing state pairs in the set of processing state pairs;
determining a start processing state and an end processing state of the processing states based on the processing trajectory;
and starting with the initial processing state and ending with the ending processing state, establishing a flow direction between processing state pairs in the maximum processing state pair set, and generating a processing flow aiming at the target problem.
Preferably, any processing state in any processing state pair in the processing state pair set may include a plurality of processing states, and the plurality of processing states satisfy a selection relationship.
Preferably, the determining the largest processing state pair set in the processing state pair set includes:
determining any first processing state pair and second processing state pair in the processing state pair set;
determining that the first processing state pair comprises the second processing state pair, and reserving the first processing state pair;
when the first processing state pair is contained by the second processing state pair, reserving the second processing state pair;
and when the first processing state pair is determined not to contain the second processing state pair and is not contained by the second processing state pair, the first processing state pair is reserved.
Preferably, the causal relationship is used to characterize that in the dialog log, a first processing state occurs first in a second processing state, but the second processing state does not occur first in the first processing state;
the selection relationship is used for representing that the first processing state is not the direct precedence relationship of the second processing state in the dialogue log, and the second processing state is not the direct precedence relationship of the first processing state;
the direct precedence relationship refers to the first processing state occurring in the dialog log first in the second processing state and being directly adjacent to the second processing state.
Preferably, the method further comprises:
generating an input-output state pair, wherein the input-output state pair comprises an input state and an output state;
the step of starting with the initial processing state and ending with the ending processing state, establishing a flow direction between the processing state pairs in the maximum processing state pair set, and generating a processing flow aiming at the target problem, including:
establishing a flow direction between the input state and the initial processing state, and taking the input state as an initial state;
establishing a flow direction between the output state and the ending processing state, wherein the output state is taken as the end;
and establishing the flow direction between the processing state pairs in the maximum processing state pair set, and generating the processing flow aiming at the target problem.
Preferably, the method further comprises:
acquiring a target problem in an intelligent dialogue system; the intelligent dialogue system can complete the processing of the target problem through multiple rounds of interaction based on a preset processing flow;
the target problem is processed based on the sequential execution of the processing states in the processing flow.
An electronic device, comprising:
a memory for storing a program;
and a processor for running the program, executing to acquire a dialogue log for a target problem, determining processing states from the dialogue log, mining a processing track from the dialogue log based on the processing states, and constructing a processing flow for the target problem based on the processing track, wherein the processing flow is a processing flow including all or part of the processing states for processing the target problem, each processing state in the processing flow is used for representing one processing node of the target problem, each processing state in the processing flow is capable of generating output information based on input information related to the target problem, and the target problem is capable of being processed by sequential execution of the processing states in the processing flow.
Preferably, the processor constructs a processing flow for the target problem based on the processing trajectory, including: constructing a processing state pair set meeting a preset condition based on the processing track, determining a maximum processing state pair set in the processing state pair set, determining a starting processing state and an ending processing state in the processing states based on the processing track, starting with the starting processing state and ending with the ending processing state, and establishing a flow direction between the processing state pairs in the maximum processing state pair set to generate a processing flow aiming at the target problem;
wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states.
Preferably, the processor determines a maximum set of processing state pairs from the set of processing state pairs, including: determining any first processing state pair and second processing state pair in the processing state pair set, determining that the first processing state pair comprises the second processing state pair, reserving the first processing state pair, reserving the second processing state pair when the first processing state pair is determined to be contained by the second processing state pair, determining that the first processing state pair does not comprise the second processing state pair, and reserving the first processing state pair when the first processing state pair is determined not to be contained by the second processing state pair.
As can be seen from the above technical solution, compared with the prior art, the present application provides a method for constructing a processing flow, in which a processing track is mined from a dialogue log for a target problem based on the processing state by obtaining the dialogue log for the target problem, and a processing flow for the target problem is constructed based on the processing track.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for constructing a processing flow according to an embodiment of the method;
FIG. 2 is a flow chart of one implementation of the process flow steps provided in accordance with another embodiment of the method of the present application with respect to building a process flow for a target issue based on a process trajectory;
FIG. 3 is a flow chart of another implementation of the process flow steps provided in accordance with yet another method embodiment of the present application for building a process flow for a target issue based on a process trajectory;
FIG. 4 is a schematic flow chart of a method for constructing a processing flow according to another embodiment of the present application;
FIG. 5 is a schematic diagram of a constructed process flow provided by an example of the present application;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
An embodiment of a method of the present application provides a method for constructing a processing flow, as shown in fig. 1, including the following steps:
step 101: acquiring a dialogue log aiming at a target problem;
the dialogue log may be a dialogue log of manual service for the target problem, or may be a dialogue log recorded in other systems for the target problem than the intelligent dialogue system of the current application. Wherein the currently applied intelligent dialog system refers to an intelligent dialog system that is able to use the subsequently built process flow.
The dialogue log of the manual customer service records dialogue contents containing processing states between the customer service personnel and the user for solving the target problem. The other system records the dialogue content including the processing state between the other system and the user for solving the target problem.
Step 102: determining a processing state from the dialogue log, and mining a processing track from the dialogue log based on the processing state;
wherein the processing states are used to characterize the processing nodes of the target problem, one processing state represents one processing node, and the processing trace refers to a set of processing states having a sequence of processing states used to solve the target problem.
It should be noted that, since different solutions may be available for different users who present the same target problem, a processing track or tracks may be mined from the dialog log for a target problem.
For example, the objective question is how to connect wifi. The dialog log includes case1: a dialogue between customer service and user a; case2: a dialogue between customer service and user B; case3: a dialogue between customer service and user C. Then the processing trajectory σ1= (a, B, C, F), σ2= (a, B, C, D, E, F, G), σ3= (a, B, C, D, E, F, H, G) can be mined based on the dialog log. Wherein A, B, C, D, E, F, H and G are all in a processing state.
Step 103: and constructing a processing flow aiming at the target problem based on the processing track.
In this application, a processing flow is a processing flow including all or part of processing states for processing the target problem, each processing state is used for characterizing one processing node of the target problem, each processing state in the processing flow can generate output information based on input information related to the target problem, and the target problem can be processed by sequentially executing the processing states in the processing flow.
Taking the target question as an example of how wifi is connected. The dialog log includes case1: a dialogue between customer service and user a; case2: a dialogue between customer service and user B; case3: a dialogue between customer service and user C. Then the processing trajectory σ1= (a, B, C, F), σ2= (a, B, C, D, E, F, G), σ3= (a, B, C, D, E, F, H, G) can be mined based on the dialog log. Wherein A, B, C, D, E, F, H and G are all in a processing state.
The constructed process flow includes: a, B, C, D, E, F, H, G, A, B, C, D, F, A, B, C, D, E, F and G. It will be appreciated that this is by way of example only, and that in actual practice other processing flows not represented in the processing trace may be mined.
Each processing state in the processing flow is capable of generating output information based on input information related to the target issue, the target issue being capable of being handled by sequential execution of the processing states in the processing flow. For example, in the processing flows a→b→c→d→e→f→h→g, the target problem can be handled by the sequential execution of a→b→c→d→e→f→h→g, for example, the processing state a acquires the input information "how to connect wifi" and generates the output information "the current state of your mobile phone? The processing state B obtains the input information as "the mobile phone is in a power-on state" and generates output information "your mobile phone model", and it is to be noted that the input information may be information provided by a user or information automatically obtained by a system, for example, the input information is "how wifi is connected to the input information provided by the user, and the" the mobile phone is in a power-on state "may be information provided by the user based on the output information" your current state of the mobile phone ", or may be automatically obtained by the system.
Therefore, in this embodiment, the processing flow can be constructed through the obtained dialogue log about the target problem, so as to assist the intelligent dialogue system in processing the target problem, and improve the efficiency of processing the target problem.
Another method embodiment of the present application is directed to describing one implementation of constructing a process flow for a target problem based on a process trajectory. Specifically, as shown in fig. 2, the method may include the following steps:
step 201: constructing a processing state pair set meeting a preset condition based on the processing track;
wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states. Taking the example of the processing state pair (a, B) in the processing state pair set, the (a, B) needs to satisfy the causal relationship.
It should be noted that any processing state in any processing state pair in the processing state pair set may include a plurality of processing states, and a selection relationship is satisfied between the plurality of processing states. Taking a processing state pair (C, (D, F)) in the processing state pair set as an example, wherein C and D satisfy a causal relationship, C and F satisfy a causal relationship, D and F are a plurality of processing states included in the processing state (D, F) in the processing state pair (C, (D, F)), and D and F satisfy a selection relationship.
In this application, causal relationships are used to characterize that in the dialog log a first processing state occurs first in a second processing state, but the second processing state does not occur first in the first processing state.
The selection relationship is used to characterize the direct precedence relationship of the first processing state that is not the direct precedence relationship of the second processing state that is not the direct precedence relationship of the first processing state in the dialog log.
The direct precedence relationship refers to the first processing state occurring in the dialog log first in the second processing state and being directly adjacent to the second processing state.
The causal, selection and direct precedence relationships described above may be determined specifically based on the processing trajectories constructed from the dialog logs, as the processing trajectories can be characterized as a collection of processing states with a precedence of processing states used to solve the target problem.
The specific definition is as follows:
charge condition of direct precedence a > W b: σ=t1, t2, …, tn, σ e W, ti=a, ti+1=b;
the causal relationship a→wb: in W, a occurs first in b and b does not occur first in a;
the charge condition of the relation a#wb is selected:and->
Where W refers to a dialog log, a, b ε T, T includes T1, T1, T2, …, tn, and σ refers to a set of processing states in a processing trace.
In constructing the set of processing state pairs, the set of processing state pairs may be obtained specifically using the following first calculation formula.
A first calculation formula:
wherein X is W Refer to a set of processing state pairs, A, B e T, T refers to a set of processing states in a processing track, a1, a2 e A, B1, B2 e B, a→Wb refer to a and B satisfying a causal relationship, a1# Wa2 refers to a1 and a2 satisfying a selection relationship, and b1# Wb2 refers to B1 and B2 satisfying a selection relationship.
Step 202: determining a maximum set of processing state pairs in the set of processing state pairs;
specifically, the determining the maximum processing state pair set in the processing state pair set may include the following steps:
(1) Determining any first processing state pair and second processing state pair in the processing state pair set;
(2) Determining that the first processing state pair comprises the second processing state pair, and reserving the first processing state pair;
(3) When the first processing state pair is contained by the second processing state pair, reserving the second processing state pair;
(4) And when the first processing state pair is determined not to contain the second processing state pair and is not contained by the second processing state pair, the first processing state pair is reserved.
Assume that the pair of processing states xw= { (a, B), (B, C), (C, D), (D, E), (C, F), (E, F), (F, G), (F, H), (H, G), (C, (D, F)), ((C, E), F) }
Since the processing state pair (a, B) does not contain other processing state pairs and is not contained by other processing state pairs, the processing state pair (a, B) is reserved.
Since the processing state pair (C, D) is (C, (D, F), the (C, (D, F) processing state pair is reserved.
Since the pair of processing states ((C, E), F) contains (E, F), the pair of processing states ((C, E, F) is reserved.
The maximum processing state pair set yw= { (a, B), (B, C), (C, (D, F)), (D, E), ((C, E), F), (F, G), (F, H), (H, G) } can be obtained.
In determining the set of maximum processing state pairs, the following second calculation formula may be specifically adopted.
The second calculation formula:
wherein Y is W Referred to as the maximum set of processing state pairs, X W Refer to the set of processing state pairs, A, B, A ', B' εT, T refer to the set of processing states in the processing trace.
Step 203: determining a start processing state and an end processing state of the processing states based on the processing trajectory;
in the processing track, the processing track is always in the first processing state and is always in the last processing state. In the processing trajectory σ1= (a, B, C, F), σ2= (a, B, C, D, E, F, G), σ3= (a, B, C, D, E, F, H, G), a is the start processing state, and G is the end processing state.
Step 204: and starting with the initial processing state and ending with the ending processing state, establishing a flow direction between processing state pairs in the maximum processing state pair set, and generating a processing flow aiming at the target problem.
In generating the processing flow, the following third calculation formula may be employed.
The third calculation formula:
F W ＝{(a,p (A,B) )|(A,B)∈Y W ∧a∈A}
∪{p (A,B) ,b}|(A,B)∈Y W ∧b∈B}
wherein F is W Refer to the processing flow, A, B epsilon T, T refers to the processing state set in the processing track, Y W Referred to as the maximum set of processing state pairs, p (A,B) An intermediate node for characterizing the flow direction.
Yet another method embodiment of the present application is mainly used to describe another implementation manner of constructing a processing flow for a target problem based on a processing track, and specifically, as shown in fig. 3, the method may include the following steps:
step 301: constructing a processing state pair set meeting a preset condition based on the processing track;
wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states.
Step 302: determining a maximum set of processing state pairs in the set of processing state pairs;
step 303: determining a start processing state and an end processing state of the processing states based on the processing trajectory;
step 304: generating an input-output state pair, wherein the input-output state pair comprises an input state and an output state;
wherein the input state is used to characterize a start node of the process flow and the output state is used to characterize an end node of the process flow.
Specifically, the input/output state pairs may be combined in a maximum set of processing state pairs to generate a set of valid state pairs. Can be obtained by using the following fourth calculation formula.
A fourth calculation formula:
p W ＝{(p (A,B) )|(A,B)∈Y W }∪{i w ,o w }
wherein Y is W Referred to as the maximum set of processing state pairs, i W Referred to as input state, o W Referred to as the output state.
Step 305: establishing a flow direction between the input state and the initial processing state, and taking the input state as an initial state;
step 306: establishing a flow direction between the output state and the ending processing state, wherein the output state is taken as the end;
step 307: and establishing the flow direction between the processing state pairs in the maximum processing state pair set, and generating the processing flow aiming at the target problem.
In generating the processing flow, it can be obtained by using the following fifth calculation formula.
Fifth calculation formula:
F W ＝{(a,p (A,B) )|(A,B)∈Y W ∧a∈A}
∪{p (A,B) ,b}|(A,B)∈Y W ∧b∈B}
∪{(i W ,t)|t∈T I }∪{(t,o w )|t∈T O }
wherein F is W Refer to the processing flow, A, B epsilon T, T refers to the processing state set in the processing track, Y W Referred to as the maximum set of processing state pairs, p (A,B) Intermediate nodes for characterizing flow direction, i W Referred to as input state, o W Refer to the output state, T I Refer to the initial processing state, T O Referred to as an end processing state.
Yet another embodiment of the present application provides a method for constructing a processing flow, as shown in fig. 4, including the following steps:
step 401: acquiring a dialogue log aiming at a target problem;
step 402: determining a processing state from the dialogue log, and mining a processing track from the dialogue log based on the processing state;
step 403: constructing a processing flow for the target problem based on the processing track;
wherein the processing flow is a processing flow comprising all or part of processing states for processing the target issue, each processing state being for a processing node characterizing the target issue, each processing state in the processing flow being capable of generating output information based on input information related to the target issue, the target issue being capable of being processed by sequential execution of the processing states in the processing flow.
Step 404: acquiring a target problem in an intelligent dialogue system;
the intelligent dialogue system can complete processing of the target problem through multiple rounds of interaction based on a preset processing flow.
Step 405: the target problem is processed based on the sequential execution of the processing states in the processing flow.
It can be seen that, in this embodiment, a processing flow can be constructed by using the obtained dialogue log about the target problem, and the processing flow can be applied to the processing of the target problem in the intelligent dialogue system, so that the efficiency of processing the target problem is improved.
For ease of understanding, the embodiments of the present application will be briefly described with a specific example.
Suppose that the data in the dialog log is as shown in table 1 below:
TABLE 1
Column of table Processing state Column of table Processing state Column of table Processing state
Case1 A Case2 C Case2 F
Case2 A Case3 C Case3 F
Case3 A Case1 F Case2 G
Case1 B Case2 D Case3 H
Case2 B Case3 D Case3 G
Case3 B Case2 E
Case1 C Case3 E
(1) The dialog log contains 8 processing states, a processing trace σ1= (a, B, C, F) for Case1, a processing trace σ2= (a, B, C, D, E, F, G) for Case2, and a processing trace σ3= (a, B, C, D, E, F, H, G) for Case 3.
(2) The relationship between pairs of processing states can be derived based on the processing trajectory as follows:
direct relationship: a > WB, B > WC, C > WF, C > WD, D > WE, E > WF, F > WG, H > WG, F > WH;
causal relationship: A→WB, B→WC, C→WF, C→WD, D→WE, E→WF, F→WG, H→WG, F→WH;
(3) Building a set of processing state pairs
Xw＝{(A,B),(B,C),(C,D),(D,E),(C,F),(E,F),(F,G),(F,H),(H,G),(C,(D,F)),((C,E),F)}
Wherein D and F satisfy a selection relationship in the processing state pair (C, (D, F)), and C and E satisfy a selection relationship in the processing state pair (((C, E)).
(4) Constructing a set of maximum processing state pairs
Yw＝{(A,B),(B,C),(C,(D,F)),(D,E),((C,E),F),(F,H),(H,G),(H,G)}
(5) Constructing valid processing state pair sets
PW＝{p(A,B),p(B,C),p({C},{D,F}),p(D,E),p({C,E},F),p(F,H),p(H,G),
p(H,G),P(iw,ow)}
(6) Generating a processing stream
TI＝A
T o ＝G
FW＝{(A,p(A,B)),(p(A,B),B),(C,p({C},{D,F})),(p({C},{D,F}),D)…(G,oW)}
A schematic diagram of a specific process flow is shown in fig. 5, where fig. 5 shows process flow a→b→c→d→e→f→h→g, a→b→c→d→f, a→b→c→d→e→f→g.
Corresponding to the above method for constructing a processing flow, the embodiment of the apparatus of the present application further provides an electronic device, and the detailed description is given below through several embodiments of the apparatus.
An embodiment of an apparatus of the present application provides an electronic device person, as shown in fig. 6, including: a memory 110 and a processor 120; wherein:
the memory 110 is used for storing programs;
the processor 120 is configured to run the program, perform obtaining a dialogue log for a target problem, determine a processing state from the dialogue log, mine a processing track from the dialogue log based on the processing state, and construct a processing flow for the target problem based on the processing track.
The dialogue log may be a dialogue log of manual service for the target problem, or may be a dialogue log recorded in other systems for the target problem than the intelligent dialogue system of the current application. Wherein the currently applied intelligent dialog system refers to an intelligent dialog system that is able to use the subsequently built process flow.
The dialogue log of the manual customer service records dialogue contents containing processing states between the customer service personnel and the user for solving the target problem. The other system records the dialogue content including the processing state between the other system and the user for solving the target problem.
Wherein the processing states are used to characterize the processing nodes of the target problem, one processing state represents one processing node, and the processing trace refers to a set of processing states having a sequence of processing states used to solve the target problem.
It should be noted that, since different solutions may be available for different users who present the same target problem, a processing track or tracks may be mined from the dialog log for a target problem.
In this application, a processing flow is a processing flow including all or part of processing states for processing the target problem, each processing state is used for characterizing one processing node of the target problem, each processing state in the processing flow can generate output information based on input information related to the target problem, and the target problem can be processed by sequentially executing the processing states in the processing flow.
Therefore, in this embodiment, the processing flow can be constructed through the obtained dialogue log about the target problem, so as to assist the intelligent dialogue system in processing the target problem, and improve the efficiency of processing the target problem.
Another apparatus embodiment of the present application primarily describes one implementation in which a processor constructs a process flow for a target problem based on a process trajectory. Specifically, the processor constructs a processing flow for the target problem based on the processing track, including: and constructing a processing state pair set meeting a preset condition based on the processing track, determining a maximum processing state pair set in the processing state pair set, determining a starting processing state and an ending processing state in the processing states based on the processing track, starting with the starting processing state and ending with the ending processing state, and establishing a flow direction between the processing state pairs in the maximum processing state pair set to generate a processing flow aiming at the target problem.
Wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states.
It should be noted that any processing state in any processing state pair in the processing state pair set may include a plurality of processing states, and a selection relationship is satisfied between the plurality of processing states.
In this application, causal relationships are used to characterize that in the dialog log a first processing state occurs first in a second processing state, but the second processing state does not occur first in the first processing state.
The selection relationship is used to characterize the direct precedence relationship of the first processing state that is not the direct precedence relationship of the second processing state that is not the direct precedence relationship of the first processing state in the dialog log.
The direct precedence relationship refers to the first processing state occurring in the dialog log first in the second processing state and being directly adjacent to the second processing state.
The causal, selection and direct precedence relationships described above may be determined specifically based on the processing trajectories constructed from the dialog logs, as the processing trajectories can be characterized as a collection of processing states with a precedence of processing states used to solve the target problem.
The specific definition is as follows:
charge condition of direct precedence a > W b: σ=t1, t2, …, tn, σ e W, ti=a, ti+1=b;
the causal relationship a→wb: in W, a occurs first in b and b does not occur first in a;
the charge condition of the relation a#wb is selected:and->
Where W refers to a dialog log, a, b ε T, T includes T1, T1, T2, …, tn, and σ refers to a set of processing states in a processing trace.
In constructing the set of processing state pairs, the set of processing state pairs may be obtained specifically using the following first calculation formula.
A first calculation formula:
wherein X is W Refer to a set of processing state pairs, A, B e T, T refers to a set of processing states in a processing track, a1, a2 e A, B1, B2 e B, a→Wb refer to a and B satisfying a causal relationship, a1# Wa2 refers to a1 and a2 satisfying a selection relationship, and b1# Wb2 refers to B1 and B2 satisfying a selection relationship.
In this application, determining the largest set of processing state pairs from the set of processing state pairs by the processor may include: determining any first processing state pair and second processing state pair in the processing state pair set, determining that the first processing state pair comprises the second processing state pair, reserving the first processing state pair, reserving the second processing state pair when the first processing state pair is determined to be contained by the second processing state pair, determining that the first processing state pair does not comprise the second processing state pair, and reserving the first processing state pair when the first processing state pair is determined not to be contained by the second processing state pair.
In determining the set of maximum processing state pairs, the following second calculation formula may be specifically adopted.
The second calculation formula:
wherein Y is W Referred to as the maximum set of processing state pairs, X W Refer to the set of processing state pairs, A, B, A ', B' εT, T refer to the set of processing states in the processing trace.
In generating the processing flow, the following third calculation formula may be employed.
The third calculation formula:
F W ＝{(a,p (A,B) )|(A,B)∈Y W ∧a∈A}
∪{p (A,B) ,b}|(A,B)∈Y W ∧b∈B}
wherein F is W Refer to the processing flow, A, B epsilon T, T refers to the processing state set in the processing track, Y W Referred to as the maximum set of processing state pairs, p (A,B) An intermediate node for characterizing the flow direction.
Yet another embodiment of an apparatus of the present application is mainly used for describing another implementation manner in which a processor constructs a processing flow for a target problem based on a processing track, and specifically, the processor constructs the processing flow for the target problem based on the processing track, including: and constructing a processing state pair set meeting a preset condition based on the processing track, determining a maximum processing state pair set in the processing state pair set, determining a starting processing state and an ending processing state in the processing states based on the processing track, generating an input and output state pair, wherein the input and output state pair comprises an input state and an output state, establishing a flow direction between the input state and the starting processing state, starting with the input state, establishing a flow direction between the output state and the ending processing state, ending with the output state, establishing a flow direction between the processing state pair in the maximum processing state pair set, and generating a processing flow aiming at the target problem.
Wherein the input state is used to characterize a start node of the process flow and the output state is used to characterize an end node of the process flow.
Specifically, the input/output state pairs may be combined in a maximum set of processing state pairs to generate a set of valid state pairs. Can be obtained by using the following fourth calculation formula.
A fourth calculation formula:
p W ＝{(p (A,B) )|(A,B)∈Y W }∪{i w ,o w }
wherein Y is W Referred to as the maximum set of processing state pairs, i W Referred to as input state, o W Referred to as the output state.
In generating the processing flow, it can be obtained by using the following fifth calculation formula.
Fifth calculation formula:
F W ＝{(a,p (A,B) )|(A,B)∈Y W ∧a∈A}
∪{p (A,B) ,b}|(A,B)∈Y W ∧b∈B}
∪{(i W ,t)|t∈T I }∪{(t,o w )|t∈T O }
wherein F is W Refer to the processing flow, A, B epsilon T, T refers to the processing state set in the processing track, Y W Referred to as the maximum set of processing state pairs, p (A,B) Intermediate nodes for characterizing flow direction, i W Referred to as input state, o W Refer to the output state, T I Refer to the initial processing state, T O Referred to as an end processing state.
Yet another embodiment of an apparatus of the present application provides an electronic device, including a memory and a processor; wherein:
the memory is used for storing programs;
the processor is used for acquiring a dialogue log aiming at the target problem, determining a processing state from the dialogue log, mining a processing track from the dialogue log based on the processing state, constructing a processing flow aiming at the target problem based on the processing track, acquiring the target problem in the intelligent dialogue system, and processing the target problem based on the sequential execution of the processing states in the processing flow.
Wherein the processing flow is a processing flow comprising all or part of processing states for processing the target issue, each processing state being for a processing node characterizing the target issue, each processing state in the processing flow being capable of generating output information based on input information related to the target issue, the target issue being capable of being processed by sequential execution of the processing states in the processing flow.
The intelligent dialogue system can complete the processing of the target problem through multiple rounds of interaction based on a preset processing flow.
Therefore, in the embodiment, the processing flow can be constructed through the acquired dialogue log about the target problem, and the processing flow can be applied to the intelligent dialogue system to assist the intelligent dialogue system in processing the target problem, so that the efficiency of processing the target problem is improved.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

1. A method of constructing a process flow, comprising:
acquiring a dialogue log aiming at a target problem;
determining a processing state from the dialogue log, mining a processing track from the dialogue log based on the processing state, wherein the processing track refers to a processing state set with the sequence of the processing states used for solving a target problem;
and constructing a processing flow aiming at the target problem based on the processing track, wherein the processing flow is a processing flow comprising all or part of processing states for processing the target problem, each processing state is used for representing one processing node of the target problem, each processing state in the processing flow can generate output information based on input information related to the target problem, and the target problem can be processed through sequential execution of the processing states in the processing flow.
2. The method of claim 1, the constructing a process flow for the target problem based on the process trajectory, comprising:
constructing a processing state pair set meeting a preset condition based on the processing track; wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states;
determining a maximum set of processing state pairs in the set of processing state pairs;
determining a start processing state and an end processing state of the processing states based on the processing trajectory;
and starting with the initial processing state and ending with the ending processing state, establishing a flow direction between processing state pairs in the maximum processing state pair set, and generating a processing flow aiming at the target problem.
3. The method of claim 2, wherein any of the processing states in any of the processing state pairs in the set of processing state pairs may include a plurality of processing states, the plurality of processing states satisfying a selection relationship therebetween.
4. A method according to claim 2 or 3, the determining the largest of the set of processing state pairs comprising:
determining any first processing state pair and second processing state pair in the processing state pair set;
determining that the first processing state pair comprises the second processing state pair, and reserving the first processing state pair;
when the first processing state pair is contained by the second processing state pair, reserving the second processing state pair;
and when the first processing state pair is determined not to contain the second processing state pair and is not contained by the second processing state pair, the first processing state pair is reserved.
5. A method according to claim 3, the causal relationship being used to characterize that in the dialog log a first processing state occurs first in a second processing state, but the second processing state does not occur first in the first processing state;
the selection relationship is used for representing that the first processing state is not the direct precedence relationship of the second processing state in the dialogue log, and the second processing state is not the direct precedence relationship of the first processing state;
the direct precedence relationship refers to the first processing state occurring in the dialog log first in the second processing state and being directly adjacent to the second processing state.
6. The method of claim 5, further comprising:
generating an input-output state pair, wherein the input-output state pair comprises an input state and an output state;
the step of starting with the initial processing state and ending with the ending processing state, establishing a flow direction between the processing state pairs in the maximum processing state pair set, and generating a processing flow aiming at the target problem, including:
establishing a flow direction between the input state and the initial processing state, and taking the input state as an initial state;
establishing a flow direction between the output state and the ending processing state, wherein the output state is taken as the end;
and establishing the flow direction between the processing state pairs in the maximum processing state pair set, and generating the processing flow aiming at the target problem.
7. The method of claim 1, further comprising:
acquiring a target problem in an intelligent dialogue system; the intelligent dialogue system can complete the processing of the target problem through multiple rounds of interaction based on a preset processing flow;
the target problem is processed based on the sequential execution of the processing states in the processing flow.
8. An electronic device, comprising:
a memory for storing a program;
a processor for running the program, executing a dialogue log for a target problem, determining processing states from the dialogue log, mining processing trajectories from the dialogue log based on the processing states, and constructing a processing flow for the target problem based on the processing trajectories, wherein the processing trajectories refer to a set of processing states having a sequence of processing states used for solving the target problem, the processing flow is a processing flow including all or part of the processing states for processing the target problem, each processing state is used for characterizing one processing node of the target problem, each processing state in the processing flow is capable of generating output information based on input information related to the target problem, and the target problem is capable of being processed by sequential execution of the processing states in the processing flow.
9. The electronic device of claim 8, the processor building a process flow for the target problem based on the process trajectory, comprising: constructing a processing state pair set meeting a preset condition based on the processing track, determining a maximum processing state pair set in the processing state pair set, determining a starting processing state and an ending processing state in the processing states based on the processing track, starting with the starting processing state and ending with the ending processing state, and establishing a flow direction between the processing state pairs in the maximum processing state pair set to generate a processing flow aiming at the target problem;
wherein the predetermined condition refers to that any processing state pair of the processing state pair set meets a causal relationship between two processing states.
10. The electronic device of claim 9, the processor determining a largest set of processing state pairs of the set of processing state pairs, comprising: determining any first processing state pair and second processing state pair in the processing state pair set, determining that the first processing state pair comprises the second processing state pair, reserving the first processing state pair, reserving the second processing state pair when the first processing state pair is determined to be contained by the second processing state pair, determining that the first processing state pair does not comprise the second processing state pair, and reserving the first processing state pair when the first processing state pair is determined not to be contained by the second processing state pair.
CN201911301153.9A 2019-12-17 2019-12-17 Method for constructing processing stream and electronic equipment Active CN110955769B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911301153.9A CN110955769B (en) 2019-12-17 2019-12-17 Method for constructing processing stream and electronic equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911301153.9A CN110955769B (en) 2019-12-17 2019-12-17 Method for constructing processing stream and electronic equipment

Publications (2)

Publication Number Publication Date
CN110955769A CN110955769A (en) 2020-04-03
CN110955769B true CN110955769B (en) 2023-07-21

Family

ID=69982083

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911301153.9A Active CN110955769B (en) 2019-12-17 2019-12-17 Method for constructing processing stream and electronic equipment

Country Status (1)

Country Link
CN (1) CN110955769B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106095965A (en) * 2016-06-17 2016-11-09 上海智臻智能网络科技股份有限公司 A kind of data processing method and device
CN109344242A (en) * 2018-09-28 2019-02-15 广东工业大学 A kind of dialogue answering method, device, equipment and storage medium
CN110019149A (en) * 2019-01-30 2019-07-16 阿里巴巴集团控股有限公司 A kind of method for building up of service knowledge base, device and equipment

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9348817B2 (en) * 2014-01-09 2016-05-24 International Business Machines Corporation Automatic generation of question-answer pairs from conversational text
TWI562000B (en) * 2015-12-09 2016-12-11 Ind Tech Res Inst Internet question answering system and method, and computer readable recording media
US11341422B2 (en) * 2017-12-15 2022-05-24 SHANGHAI XIAOl ROBOT TECHNOLOGY CO., LTD. Multi-round questioning and answering methods, methods for generating a multi-round questioning and answering system, and methods for modifying the system
CN110020014B (en) * 2017-12-15 2021-11-12 上海智臻智能网络科技股份有限公司 Multi-turn question and answer device
CN108804643A (en) * 2018-06-05 2018-11-13 北京玄科技有限公司 Flow dispensing unit, method and the intelligent robot interactive system of more wheel sessions
CN109308320A (en) * 2018-07-20 2019-02-05 北京智能点科技有限公司 Conversation process configuration method is taken turns more by a kind of robot
CN109189803A (en) * 2018-09-03 2019-01-11 深圳前海微众银行股份有限公司 Question and answer are to construction method, device and computer readable storage medium
CN109783510A (en) * 2019-01-22 2019-05-21 深圳追一科技有限公司 Customer service auxiliary robot, method and system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106095965A (en) * 2016-06-17 2016-11-09 上海智臻智能网络科技股份有限公司 A kind of data processing method and device
CN109344242A (en) * 2018-09-28 2019-02-15 广东工业大学 A kind of dialogue answering method, device, equipment and storage medium
CN110019149A (en) * 2019-01-30 2019-07-16 阿里巴巴集团控股有限公司 A kind of method for building up of service knowledge base, device and equipment

Also Published As

Publication number Publication date
CN110955769A (en) 2020-04-03

Similar Documents

Publication Publication Date Title
CN108961052A (en) Verification method, storage method, device, equipment and the medium of block chain data
CN109951547B (en) Transaction request parallel processing method, device, equipment and medium
US20220253710A1 (en) Human-Machine Multi-Turn Conversation Method and System for Human-Machine Interaction, and Intelligent Apparatus
CN105491078B (en) Data processing method and device, SOA system in SOA system
WO2017000761A1 (en) Method and apparatus for extracting feature information of terminal device
JP2019536122A (en) Method and apparatus for information interaction
CN106294564A (en) A kind of video recommendation method and device
CN110717647A (en) Decision flow construction method and device, computer equipment and storage medium
CN109582289A (en) The processing method of regular flow, system, storage medium and processor in regulation engine
CN108664665A (en) Data format method for transformation, device, equipment and readable storage medium storing program for executing
CN113573093A (en) Live broadcast data analysis method and device, electronic equipment and storage medium
CN110932929B (en) Method, system and medium for classifying and extracting satellite telemetry packets in CCSDS system
CN106845625A (en) A kind of interactive output intent and robot for robot
CN111581390A (en) Knowledge graph construction method and device and electronic equipment
CN110955769B (en) Method for constructing processing stream and electronic equipment
US8368698B2 (en) Calculating a webpage importance from a web browsing graph
CN104424352A (en) System and method for providing agent service to user terminal
CN111666388A (en) Dialogue data processing method, device, computer equipment and storage medium
CN112560939B (en) Model verification method and device and computer equipment
CN111062612B (en) Construction method of auxiliary processing stream and electronic equipment
CN110955591A (en) System performance evaluation method and device, computer equipment and storage medium
CN106919374A (en) A kind of scenario generation method and device
US10659304B2 (en) Method of allocating processes on node devices, apparatus, and storage medium
WO2023142349A1 (en) Behavior sequence generation method and apparatus, storage medium and electronic device
CN103617796A (en) Voice recognition method based on cloud computing

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant