CN110727875A - Intelligent distribution method and system for legal case agent - Google Patents
Intelligent distribution method and system for legal case agent Download PDFInfo
- Publication number
- CN110727875A CN110727875A CN201911298838.2A CN201911298838A CN110727875A CN 110727875 A CN110727875 A CN 110727875A CN 201911298838 A CN201911298838 A CN 201911298838A CN 110727875 A CN110727875 A CN 110727875A
- Authority
- CN
- China
- Prior art keywords
- lawyer
- case
- client
- layer
- service
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 79
- 238000005516 engineering process Methods 0.000 claims abstract description 51
- 230000008569 process Effects 0.000 claims abstract description 49
- 238000012544 monitoring process Methods 0.000 claims abstract description 28
- 238000007726 management method Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000005457 optimization Methods 0.000 claims abstract description 18
- 230000003068 static effect Effects 0.000 claims abstract description 17
- 238000013500 data storage Methods 0.000 claims abstract description 13
- 238000011160 research Methods 0.000 claims abstract description 8
- 230000008451 emotion Effects 0.000 claims description 54
- 238000004458 analytical method Methods 0.000 claims description 25
- 230000006870 function Effects 0.000 claims description 19
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000013135 deep learning Methods 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 13
- 238000004891 communication Methods 0.000 claims description 8
- 238000013461 design Methods 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 7
- 238000003058 natural language processing Methods 0.000 claims description 7
- 238000013468 resource allocation Methods 0.000 claims description 5
- 238000012797 qualification Methods 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 230000002452 interceptive effect Effects 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims 1
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 239000000284 extract Substances 0.000 abstract description 3
- 230000004044 response Effects 0.000 description 13
- 230000007547 defect Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 230000014509 gene expression Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000013475 authorization Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000002996 emotional effect Effects 0.000 description 1
- 230000003090 exacerbative effect Effects 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9538—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
- G06Q50/182—Alternative dispute resolution
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Technology Law (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent distribution method and system for legal case agents, which comprises a user interface layer, a feature processing layer, a matching distribution layer, a data storage layer and a monitoring management layer, and adopts an online lawyer scoring recommendation technology combining dynamic indexes and client feedback to solve the problem that the conventional lawyer feature characteristics and recommendation basically depend on static information, so that the capability of a lawyer in carrying legal cases can be reflected more objectively and timely; secondly, an artificial intelligence technology is applied to automatically extract and depict the intention and the portrait of the client when seeking lawyer agents, corresponding information is sent to a matching distribution link, and the most relevant and most suitable lawyer is matched for the client; and then, based on a task assignment technology in the operational research optimization field, the optimal matching of multiple clients and multiple lawyers in the legal case agent process is realized, and meanwhile, the problems of fatigue control of the lawyers and case source distribution balance of new lawyers are solved.
Description
Technical Field
The invention relates to the technical field of legal case agents, in particular to an intelligent distribution method and system of legal case agents.
Background
When people encounter legal disputes and further seek lawyer agents for help, the following important factors are often concerned: whether lawyers are professional, whether lawyer-skilled areas match dispute cases that i encounter, how well lawyers are served, whether lawyers have enough time and energy to help i solve problems and solve doubts, how quickly i can get the most appropriate lawyer information through, and the like, which are the core complaints most concerned by the masses as clients in legal agencies or entrusting relationships.
On the other hand, the interest of lawyer groups is closely related to the appeal of the client, and is different. The reason is that the total number of lawyers in China is less than one hundred thousand at present, and the number of lawyers in every ten thousand is only 2.17 by combining the population cardinality of China; at the same time, lawyers in the united states have up to 37 per thousand. The shortage of lawyers' resources creates a certain imbalance in supply and demand, and the case agents and consultants that lawyers receive daily often exceed their workload, while for revenue considerations, lawyers are generally biased towards higher value, i.e., higher target rate cases; meanwhile, the difference of law enforcement years, case processing experience and the like can also cause case resource allocation difference in law groups to a certain extent: some highly qualified, well-known attorneys may receive a large number of premium sources, while some new attorneys who have just entered may receive less than nominal amounts and less than nominal amounts. The above two points make attorneys pay more attention when taking over the agent cases: whether the consultation or agency requirement can be finally converted into a formal consignment contract, how large the nominal amount of the case is, whether I still need to accept the simple consultation as a qualified attorney, and how I can obtain more and better case resources as a new attorney.
In the process of reconciling different value requirements of clients and lawyers, the development of internet technology enables the channels of client searching lawyers to be gradually shifted from offline to online. In recent years, a large number of case agent matching platforms mainly based on lawyer diversion function appear, and the general method for processing legal case matching by the platforms is as follows: on the user side, the user manually filters and screens own agent lawyers according to conditions such as case types, lawyer regions, working years, places and places, but general users are difficult to distinguish the two types of cases due to the fact that the screening items are too professional, for example, "labor disputes" and "labor disputes" in dispute types;
on the lawyer side, active inquiry of users is attracted relatively passively by means of showing out personal information, areas of excellence, past cases and the like on websites. Obviously, the matching between the client and the lawyer depends on self selection of both parties, but due to the asymmetry of information and the problem of cold start, the user cannot necessarily find the most suitable lawyer with sufficient time to serve the user; the new law enforcement officers still face the problem of few cases. The platform only performs basic functions of lawyer review, exhibition position control and the like, does not perform matching, and is difficult to ensure the service quality.
By means of the platforms, partial researchers and manufacturers try to make lawyer recommendations according to the characteristics of the dispute cases. The specific method generally comprises the following steps: according to the dispute type selected by the user or the dispute type obtained by case description analysis and prediction, lawyers are screened from a law teacher database (some methods are searched from the law teacher database of the user, and other methods are used for more simply extracting lawyer information from public official document data), and the screening conditions mainly surround the relatively static characteristics of the lawyers such as the regions, the practical years, the areas with strong proficiency and the like. The recommendation has the advantages that lawyers with better matching and higher winning rates can be given in a certain range, but the defects are obvious: 1. the situation that a new lawyer is in the position behind a recommendation list due to short practice period and few agent cases cannot be solved; 2. lawyer information extracted from the referee document is inaccurate or has changed; 3. the mere use of lawyer scores for recommendation results in high scoring lawyers being continually recommended for good quality case sources, exacerbating the problem of uneven case source distribution, and failing to ensure that lawyers have sufficient time and energy to better serve the customer.
Therefore, in summary, the existing method and lawyer recommendation technology for realizing case agent matching by using a website with lawyer diversion as a main function have the following disadvantages in the process of dealing with increasing legal case agent distribution and appeal:
firstly, the disadvantages in the process of lawyer recommendation or characterization of lawyer features are:
1. the existing lawyer diversion websites and lawyer recommendation technologies adopt a mode of scoring and recommending purely according to lawyer static information (such as regions, years, adequacy fields and the like), mainly rely on lawyers to submit during authentication or actively update in the follow-up process, and inevitably meet the risks of accuracy and timeliness; meanwhile, some indexes capable of assisting in representing the service capability of the lawyer at the current moment, such as the indexes of the working load of the lawyer and the like, are not included in the characteristic representation;
2. the current lawyer recommendations are generally only part of the offline calculation, and the client feedback information in the service process is not returned online to the lawyer recommendation, and the risk points generated are as follows: if online consultation or other communication links exist, lawyers have difficulty in feeding back and influencing the recommendation ranking of the lawyers in real time if some illegal professional principles or behaviors which are not friendly to clients appear.
Secondly, for the client side, the prior art generally suffers from these drawbacks in the process of screening lawyers:
1. the existing lawyer diversion website generally needs a client to manually fill out screening conditions, wherein the expression of part of the screening conditions is difficult to understand for general public with non-legal specialties, for example, the meaning of the type of dispute, which type of dispute my case should belong to, what the territorial representation of the lawyer represents, which cases need local lawyer agents, which cases can be handled in different places, and the like, which options of over-specialized expression increase the cost of the client for using the website or the system;
2. the conventional few lawyer recommendation technologies adopt prediction of case dispute types according to client case description and by combining machine learning or deep learning technologies, so that the cost of knowing professional options of a client is reduced, but the defect that information loss to a certain extent is caused when the lawyer recommendation is carried out only according to the case description exists, such as regional information, identity information, emotion information and the like of the client, which are actually taken as factors to be considered in the lawyer matching or recommendation process exists.
Finally, in the process of matching, matching or distributing the attorney side and the client side:
1. the prior art generally only makes 1-to-N pairing, namely one client comes in and matches and recommends the most appropriate set of lawyers for selection, but as the industry develops and the website traffic increases, more than one client logs in and visits at the same time, at this time, an M-to-N optimization matching problem exists, namely how to assign M users to N lawyers and optimize the target of the whole assignment process, for example, the time and energy of the qualified lawyers are limited, if M users are assigned to the best lawyers, a situation that some users need to queue in a queue occurs, however, if some users have high instantaneity requirements on responses, whether the users can be assigned to suboptimal lawyers is a better choice.
2. The existing technology does not control the fatigue degree aiming at the service capability of a lawyer, and under the condition that the lawyer receives too many orders within a period of time, the continuous order dispatching is actually reduced, otherwise, the time and the service quality of the lawyer cannot be ensured;
3. new attorneys face the problem of cold starts in both existing attorney diversion websites and recommendation technologies: for a new attorney who has a short practice period, few past cases, relatively good ability and plenty of time, it is difficult to obtain better recommendation ranking and information exposure in the prior art, which also affects the acquisition of more and better case sources.
Disclosure of Invention
Aiming at the defects or shortcomings in the prior art, an intelligent distribution method and system of legal case agents are provided, an online lawyer scoring recommendation technology combining dynamic indexes and client feedback is adopted, the problem that the conventional lawyer characteristic characteristics and recommendation basically depend on static information is solved, and the capability of lawyers in carrying legal cases can be reflected more objectively and timely; secondly, an artificial intelligence technology is applied to automatically extract and depict the intention and the portrait of the client when seeking lawyer agents, corresponding information is sent to a matching distribution link, and the most relevant and most suitable lawyer is matched for the client; and then, based on a task assignment technology in the operational research optimization field, the optimal matching of multiple clients and multiple lawyers in the legal case agent process is realized, and meanwhile, the problems of fatigue control of the lawyers and case source distribution balance of new lawyers are solved.
The technical purpose of the invention is realized by the following technical scheme:
an intelligent distribution method and system for legal case agents comprises a user interface layer, a feature processing layer, a matching distribution layer, a data storage layer and a monitoring management layer;
the method comprises the steps that a client expresses case situations on a user interface layer and finally transfers the case situations into text display, dispute types and appeal with finer granularity are predicted by means of a multi-label text classification technology based on deep learning, prediction of the multiple appeal is supported, and meanwhile, the urgency of the case appeal is analyzed and judged by means of an emotion analysis technology in natural language processing;
the off-line characteristics refer to client portrait information processed in non-real time and lawyer static information in a data layer, and are analyzed from the data layer in a T +1 mode by adopting an autonomously developed interface which can be used for data processing by python; the real-time indexes refer to lawyer dynamic indexes and monitoring indexes of a service process;
the lawyer dynamic index is composed of a real-time load index, a real-time emotion index and a real-time service quality index 3, and specifically comprises the following steps:
real-time load class indicators, including: the total number of service cases on the day, the accumulated duration of the consultation on the day, the accumulated reply times of the consultation on the day, the total number of the service cases of the last 1 hour, the accumulated duration of the consultation on the last 1 hour, the accumulated reply times of the consultation on the last 1 hour, the maximum value of the duration of the consultation on the day, the average value of the duration of the consultation on the day, the maximum value of the reply times of the consultation on the day, the average value of the reply times of the consultation on the day, the maximum first response duration of the consultation on the last 1 hour and the average value of the first response duration of the consultation on the last 1 hour are 12 indexes in total;
the real-time emotion index is obtained by applying emotion analysis technology in the field of natural language processing and through recognition: the positive emotion score of the lawyer, the negative emotion score of the lawyer and the negative emotion category number of the lawyer are 3 indexes in total;
the real-time service quality class index comprises: the average service satisfaction score of the lawyers on the day and the average service satisfaction score of the lawyers in the latest 1 hour are 2 indexes in total;
the monitoring indexes of the service process are processed by using a Flink or Spark Streaming computing technology and assisted by a real-time algorithm service, and the specific processing mode is as follows:
the attorney real-time load indexes in the service process comprise 12 indexes including total amount of daily service cases, accumulated duration of daily consultation, accumulated reply times of daily consultation, total amount of latest 1-hour service cases, accumulated duration of latest 1-hour consultation, accumulated reply times of latest 1-hour consultation, maximum duration of daily consultation, average duration of daily consultation, maximum reply times of daily consultation, average maximum first response duration of latest 1-hour consultation and average first response duration of latest 1-hour consultation, and real-time service quality indexes including 2 indexes including average service satisfaction score of daily attorney and average service satisfaction score of latest 1-hour attorney, and are processed in real time by Flink or Spark Streaming computing technology, obtaining the real-time value in the time window of the current day or the latest 1 hour;
the lawyer real-time emotion indexes in the service process comprise 3 indexes including a lawyer positive emotion score, a lawyer negative emotion score and a lawyer negative emotion category number and client emotion change indexes including 3 indexes including a client positive emotion score, a client negative emotion score and a client negative emotion category number, and are processed by the aid of deployed real-time algorithm service.
The matching distribution layer comprises a proxy case queue, a lawyer scoring module, a service lawyer queue and an intelligent distribution module, wherein the information in the proxy case queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, and meanwhile the priority of clients in the proxy case queue is adjusted by combining the analysis result of an emotion analysis technology; the lawyer scoring module integrates the processed offline features and the real-time index features through a deep learning technology to calculate a matching score so as to update the calculation process of correcting the score of the lawyer in real time, and further update the recommendation sequence and resource allocation of the service lawyer list; after the matching score is obtained through calculation, an objective function and constraint conditions are established first when the optimization problem is solved, and the constraint conditions are established according to design ideas such as lawyer fatigue degree control, new lawyers and deep lawyers on case resources, distribution balance control and the like; the intelligent distribution module adopts a task assignment technology based on operational research optimization to optimally match a plurality of lawyers and a plurality of clients in a service lawyer queue and a proxy case queue, and finally a matching distribution layer outputs an optimal matching result to an interface for displaying;
the data storage layer comprises an authentication law database and an agent case database, and the authentication law database can persistently match partial dynamic indexes and static information generated in the distribution process;
the above dynamic indicators persisted into the database include: the real-time load index, the real-time emotion index and the real-time service quality index of a lawyer are composed of 12 indexes in total, wherein the real-time load index comprises the total amount of service cases on the day, the accumulated time length of receiving consultation on the day, the accumulated reply times of receiving consultation on the day, the total amount of service cases in the last 1 hour, the accumulated time length of receiving consultation in the last 1 hour, the accumulated reply times of receiving consultation in the last 1 hour, the maximum value of the time length of receiving consultation on the day, the average value of the time length of receiving consultation on the day, the maximum value of the reply times of receiving consultation on the day, the average value of the reply times of receiving consultation on the day, the maximum first response time length of receiving consultation in the last 1 hour and the average value of the first response time length of receiving consultation in the last 1 hour; the real-time emotion index consists of 3 indexes including a lawyer positive emotion score, a lawyer negative emotion score and a lawyer negative emotion category number; the real-time service quality index consists of 2 indexes, namely the average service satisfaction score of the lawyer on the day and the average service satisfaction score of the lawyer in the latest 1 hour;
the static information persisted to the database includes: the age range, sex, service region, practice certificate number, practice period, field of excellence, place and past case description of lawyer are all 8 pieces of information, and are all stored in the database in character type data types.
More preferably: the user interface layer is an interactive page of the system and the client and comprises a description interface module for providing case text input or voice input for the client, an online communication interface for carrying out prior consultation and communication between the client and lawyers, a lawyer display interface obtained according to matching of case description, basic information and the like of the client, and a manual screening interface for supporting the lawyer which can be manually screened to meet the requirements of the client if the matching result is not satisfactory.
More preferably: the fatigue control is statistically analyzed according to the real-time workload, the historical workload and the like of a service lawyer, and a fatigue control factor is calculated; the distribution balance control adopts the mode of adding new attorney protection factors and properly adjusting the weight distribution of attorney scoring factors so as to strengthen the assessment of attorneys on service attitude and service quality.
More preferably: the task assignment technique finally solves the result to obtain a matching matrix of M client N service attorneys, and after obtaining the matching matrix, adjusts the matching result according to some set rules, which provide tools for system management and operators to control the matching result, for example, in case of an emergency, the services of some attorneys or clients need to be terminated or limited, and emergency adjustment can be performed accordingly.
More preferably: the data storage layer adopts a relational database, the retrieval architecture adopts a distributed and extensible real-time search and analysis engine, namely an Elasticissearch to construct a search function, the Elasticissearch is a search engine established on the basis of a full-text search engine Apache Lucene, and the full-text search function of distributed multi-user capability can be provided.
More preferably: the monitoring management layer comprises a visualization module for historical agent or case receiving situation of each lawyer, an authentication management module for qualification examination of lawyers applying for admission, a distribution monitoring module for monitoring a distribution process to prevent distribution errors such as repeated distribution of cases to the same lawyer, a service process monitoring module for monitoring a service process to prevent service risks of the lawyer or both clients, and a case agent management module for performing case agent process management, schedule updating and the like.
Aiming at the intelligent distribution method and the system, another processing mode is extended from intelligent matching of a service attorney and a client on a matching distribution layer, and the method specifically comprises the following steps:
the matching distribution layer comprises a proxy case queue, a lawyer scoring module, a service type queue and an intelligent distribution module, wherein the information in the proxy case queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, and meanwhile the priority of the client in the proxy case queue is adjusted by combining the analysis result of an emotion analysis technology; a service type queue of a structured data type is constructed in advance, a lawyer scoring module integrates the processed offline characteristics and real-time index characteristics of lawyers through a deep learning technology to calculate the real-time scores of the service lawyers in different fields and update a service type list in real time, wherein the different fields can be different dispute types; establishing an objective function and constraint conditions, wherein the constraint conditions are established according to design ideas such as lawyer fatigue degree control, new lawyer and deep lawyer case resource distribution balance control and the like; the intelligent distribution module adopts a task assignment technology based on operational research optimization to optimally match multiple lawyers and multiple clients in the service type queue and the agent case queue, and finally the matching distribution layer outputs the optimal matching result to an interface for displaying.
In summary, compared with the prior art, the beneficial effects of the invention are as follows: the invention adopts the idea of operational optimization to process the case distribution and matching problem of a client in the process of seeking legal case agents, can fully meet the matching requirement of cases and lawyers in the field of adequacy through reasonable objective function and constraint condition setting, also comprehensively considers the satisfaction degree of the client to the service quality, realizes the optimal distribution of lawyer resources, and can solve the problem that a new lawyer is abundant in time at the initial stage of practice but insufficient in case source to a certain extent;
on a lawyer end, an online lawyer scoring and recommending technology combining dynamic indexes and service process feedback information is adopted to overcome the defect that the characteristic representation of a lawyer in the prior art basically depends on static information of the lawyer, so that the real-time service capability and the service quality of the lawyer can be better measured, and meanwhile, the service process is monitored to a certain degree;
at a client, the intention and portrait of a client are automatically extracted and depicted by adopting an artificial intelligence-based technology when seeking legal case agents, and not only can the case dispute type related in case description of the client be predicted through a deep learning-based multi-label text classification technology during intention identification, but also a client appeal with finer granularity can be identified; meanwhile, basic information of the client and emotional information in the description are brought into the common portrayal of the client, so that the problem that only the dimension of the case dispute type of the client is considered to be too single when the client and lawyers are matched in the prior art is solved;
finally, when the case matching problem is processed by applying the operational research optimization idea, the optimization problem of multiple clients and multiple lawyers is solved, the distributed optimal solution can be obtained in a macroscopic and global range, a lawyer fatigue degree control mechanism and a new and old lawyer case source balancing mechanism are added, reasonable balanced distribution of case sources among new and old lawyers is achieved, a more reasonable case distribution mode is provided for lawyers, and meanwhile higher-quality services can be provided for clients.
Drawings
FIG. 1 is an overall frame diagram of embodiment 1;
fig. 2 is a specific technical solution architecture of embodiment 1;
FIG. 3 is a graph showing the calculation of a matching score in example 1;
fig. 4 is a typical service flow diagram in embodiment 1.
Detailed description of the preferred embodiments
The invention is described in further detail below with reference to the accompanying drawings.
Example 1:
an intelligent distribution method and system for legal case agents is shown in fig. 2 and comprises a user interface layer, a feature processing layer, a matching distribution layer, a data storage layer and a monitoring management layer.
The user interface layer is an interactive page of the system and the client, and comprises a description interface module for providing case text input or voice input for the client, an online communication interface for carrying out prior consultation and communication between the client and lawyers, a lawyer display interface obtained according to matching of case description, basic information and the like of the client, and a manual screening interface for supporting the lawyer which can be manually screened to meet the requirements of the client if the matching result is not satisfactory;
the method comprises the steps that a client expresses a case in a description interface module and finally transfers the case into a text for display, dispute types and appeal with finer granularity are predicted by means of a multi-tag text classification technology based on deep learning, and prediction of multiple appeal is supported, wherein the appeal with finer granularity is such as a dispute type as a marital family, and the appeal with fine granularity comprises child fostering rights, gift return and the like;
meanwhile, emotion analysis technology in natural language processing is applied to analyze and judge the urgency of case appeal, judge whether the appeal of a client is very urgent or not, and further determine whether the priority of the subsequent queue sequence of cases is improved or not.
The data storage layer stores authentication law information and agent case information, and comprises an authentication law database and an agent case database, wherein the authentication law database can persistently match partial dynamic indexes and static information generated in the distribution process;
dynamic metrics persisted to the database include: the real-time load index, the real-time emotion index and the real-time service quality index of a lawyer are composed of 12 indexes in total, wherein the real-time load index comprises the total amount of service cases on the day, the accumulated time length of receiving consultation on the day, the accumulated reply times of receiving consultation on the day, the total amount of service cases in the last 1 hour, the accumulated time length of receiving consultation in the last 1 hour, the accumulated reply times of receiving consultation in the last 1 hour, the maximum value of the time length of receiving consultation on the day, the average value of the time length of receiving consultation on the day, the maximum value of the reply times of receiving consultation on the day, the average value of the reply times of receiving consultation on the day, the maximum first response time length of receiving consultation in the last 1 hour and the average value of the first response time length of receiving consultation in the last 1 hour; the real-time emotion index consists of 3 indexes including a lawyer positive emotion score, a lawyer negative emotion score and a lawyer negative emotion category number; the real-time service quality index consists of 2 indexes, namely the average service satisfaction score of the lawyer on the day and the average service satisfaction score of the lawyer in the latest 1 hour;
static information that is persisted into the database includes: the age range, sex, service region, practice certificate number, practice period, field of excellence, place and past case description of lawyer are all 8 pieces of information, and are all stored in the database in character type data types.
The data storage layer adopts a relational database to store conventional information, such as a MySQL relational database, meanwhile, for the purpose of convenient retrieval, a retrieval framework adopts a distributed and extensible real-time search and analysis engine Elasticissearch to construct a search function, the Elasticissearch is a search engine established on the basis of a full-text search engine Apache Lucene, and the full-text search function of distributed multi-user capability can be provided.
Referring to fig. 1 and 2, the feature processing layer includes a combination of off-line feature processing and real-time index processing, where the off-line feature processing mainly refers to non-real-time processed client image information and lawyer static information in the data layer, such as the region of the lawyer, the practice age, the age, sex, and the region of the client, which do not change in a short time, and the feature processing layer is analyzed from the data layer in a T +1 manner by using an autonomously developed interface, and the data layer is designed as a relational database or structured information;
real-time index processing refers to dynamic indexes of lawyers and monitoring indexes of a service process, and the real-time index processing refers to dynamic indexes of the lawyers, and is composed of a large part of real-time load indexes, real-time emotion indexes and real-time service quality indexes 3:
real-time load class indicators, including: the total number of service cases on the day, the accumulated duration of the consultation on the day, the accumulated reply times of the consultation on the day, the total number of the service cases of the last 1 hour, the accumulated duration of the consultation on the last 1 hour, the accumulated reply times of the consultation on the last 1 hour, the maximum value of the duration of the consultation on the day, the average value of the duration of the consultation on the day, the maximum value of the reply times of the consultation on the day, the average value of the reply times of the consultation on the day, the maximum first response duration of the consultation on the last 1 hour and the average value of the first response duration of the consultation on the last 1 hour are 12 indexes in total;
the real-time emotion index is obtained by applying emotion analysis technology in the field of natural language processing and through recognition: the positive emotion score of the lawyer, the negative emotion score of the lawyer and the negative emotion category number of the lawyer are 3 indexes in total;
the real-time service quality class index comprises: the average service satisfaction score of the lawyers on the day and the average service satisfaction score of the lawyers in the latest 1 hour are 2 indexes in total;
calculating monitoring indexes in a service process, applying a Flink or Spark Streaming flow computing technology and processing by assisting with real-time algorithm service, wherein the specific processing mode is as follows:
the attorney real-time load indexes in the service process comprise 12 indexes including total amount of daily service cases, accumulated duration of daily consultation, accumulated reply times of daily consultation, total amount of latest 1-hour service cases, accumulated duration of latest 1-hour consultation, accumulated reply times of latest 1-hour consultation, maximum duration of daily consultation, average duration of daily consultation, maximum reply times of daily consultation, average maximum first response duration of latest 1-hour consultation and average first response duration of latest 1-hour consultation, and real-time service quality indexes including 2 indexes including average service satisfaction score of daily attorney and average service satisfaction score of latest 1-hour attorney, and are processed in real time by Flink or Spark Streaming computing technology, obtaining the real-time value in the time window of the current day or the latest 1 hour;
the lawyer real-time emotion indexes in the service process comprise 3 indexes including a lawyer positive emotion score, a lawyer negative emotion score and a lawyer negative emotion category number and client emotion change indexes including 3 indexes including a client positive emotion score, a client negative emotion score and a client negative emotion category number, and are processed by the aid of deployed real-time algorithm service.
The matching distribution layer comprises a case agent queue, a lawyer scoring module, a service lawyer queue and an intelligent distribution module, information in the case agent queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, the portrait information is modeled by extracting client characteristics such as age, gender, regions where the portrait is located, professional portrait and other information, and meanwhile the priority of clients in the case agent queue is adjusted by combining an analysis result of an emotion analysis technology;
the lawyer scoring module integrates the processed offline features and the real-time index features through a deep learning technology to calculate the matching score so as to update the calculation process of correcting the score of the lawyer in real time, and further is used for updating the recommendation sequence and resource allocation of a service lawyer list, so that the defects of lack of real-time data feedback of a service process and service risk monitoring and early warning in the prior art are overcome, and the matching score calculation process is specifically described:
assuming that the matching score is represented as c (i, j) representing the score obtained by allocating the case agency requirement of the ith client to the jth lawyer, wherein the score calculation of c (i, j) comprises not only the matching degree of the case situation of the client and the lawyer good field, but also the fit degree of the urgency degree of the client requirement and the free degree of the lawyer, and the like Whether there is sufficient time and effort to provide services, etc.; a deep neural network is used to learn the historical samples and the probability of the final result is used as the matching score of the lawyer and the client.
After the matching score is obtained through calculation, an objective function and a constraint condition are established when an optimization problem is solved, wherein the design of the objective function is related to the calculation of the matching score, and the matching score comprehensively considers whether the matching binary provides suggestions and services meeting the case and service attitudes and the like in the calculation process, so that the objective function is designed as follows:
whereinIndicating the matching score obtained by assigning the case agent requirements of the ith client to the jth attorney,the case agent requirement representing the ith client is 1 when assigned to the jth attorney, otherwise it is 0.
The establishment of the constraint condition needs to be carried out according to design ideas such as lawyer fatigue degree control, appropriate balance on new lawyers and deep lawyers case resources and the like. Taking lawyer fatigue control as an example, one of the constraints can be simply expressed as:
indicating that the workload of the jth attorney cannot exceed its upper bound Tj.
After the objective function and the constraint condition are established, the intelligent distribution module in the invention adopts a task assignment technology based on operational optimization, namely a related Algorithm of task assignment, such as Hungarian Algorithm (Hungary Method) or stable marital matching Algorithm (Gale-Shapley Algorithm) and the like to solve the optimization problem of case and attorney matching, so that the optimal matching of multiple attorneys and multiple clients in a service attorney queue and an agent case queue is realized;
specifically, the cases of M clients are used as tasks, N lawyers are used as service resources to assign and distribute the tasks, and a stable matching result is obtained; the specific implementation process comprises the following steps: when the number of cases is too large and the number of lawyers is insufficient, part of clients are inevitably required to wait, and the sequencing of the waiting queue at the moment is calculated according to the case severity, the client emotion and other portrait characteristics; when the number of lawyers is too large and the number of cases is insufficient, proper balance in the aspect of case resource allocation needs to be considered; when the number of cases and the number of lawyers are large, distributed processing is needed, and time consumption of solving the optimization problem is reduced. Finally, solving a result to obtain an M N matrix, wherein M represents the case number of the client, and N represents the lawyer number;
after the matching matrix is obtained, the matching result needs to be adjusted according to some set rules, and the adjustment rules provide tools for controlling the matching result for system management and operators, for example, in case of an emergency, services of part of lawyers or clients need to be terminated or limited, and emergency adjustment can be performed according to the tools; and finally, the matching distribution layer outputs the matched result to an interface for displaying.
The monitoring management layer comprises a visualization module for historical agent or case receiving situation of each lawyer, an authentication management module for qualification examination of lawyers applying for admission, a distribution monitoring module for monitoring a distribution process to prevent distribution errors such as repeated distribution of cases to the same lawyer, a service process monitoring module for monitoring a service process to prevent service risks of the lawyer or both clients, and a case agent management module for performing case agent process management, progress updating and the like.
In conclusion, the case distribution system intelligently matching the client case agent requirements and lawyer resources is realized through the technical design of the data storage layer, the feature processing layer, the matching distribution layer, the user interface layer and the monitoring management layer.
In view of the above, a typical business process in the implementation and operation is as follows, and reference is made to fig. 4:
1. lawyer application and admission, wherein the lawyer authentication management module is used for verifying and admitting qualification, including identity verification, real person authentication, authorization information verification and the like;
2. lawyer information passing the authentication is stored in a relational database for standby, and meanwhile, an Elasticissearch-based full-text search engine is also constructed to support manual screening and searching of lawyers by a user;
3. processing the characteristics of the static information and the dynamic indexes of the lawyers for the matching score calculation module to use, wherein the static characteristics can be calculated in a relatively offline mode, and the dynamic characteristics are processed based on a streaming calculation framework;
4. the customer inputs the case description in a case description module in a text or voice mode, the case description text and the voice can be relatively spoken expressions, and the description voice needs to be transcribed into a text display;
5. the system extracts case description texts of the clients and basic information of the clients, such as age, region, gender and the like, wherein the texts need to be preprocessed, and other information is extracted and transmitted through a user interface or a small program end;
6. the method comprises the steps of analyzing intention of case description texts and client information, wherein the intention analysis comprises dispute types and fine-grained appeal, simultaneously depicting a basic portrait of a user, such as whether legal assistance is needed, what the current emotion is, and the like, adopting a multi-label text classification technology based on deep learning for analyzing the intention of the client, and adopting a method based on machine learning and deep learning for depicting the basic portrait, such as adopting an emotion analysis technology in natural language processing to depict the current emotion of the client;
7. the client selects whether to perform automatic matching, if so, the case intelligent matching distribution algorithm is entered for calculation, and if the user is not satisfied with the intelligent matching result, the user can also select not, and the user goes to the step 10 to perform manual lawyer screening;
8. training a deep neural network according to the characteristics of the lawyers and the clients obtained in the steps 3 and 6 to calculate a matching score when a client is allocated to a lawyer, wherein the matching score not only considers the case and lawyer adequacy field and the matching degree of the client region and the lawyer region, but also comprehensively considers the service attitude of the lawyer, the urgency degree of the client demand (including whether the client emotion is excited or not) and the vacancy degree of the lawyer or not and the like;
9. solving the case matching optimization problem by adopting a task assignment problem solving method, wherein the problem solving result is finally presented as an M-N matrix, wherein M is the case number of the client; n is the number of lawyers; the solution result also needs to be post-processed later, for example, in case of an emergency, the service of part of lawyers or clients needs to be terminated or limited, and the post-processing can be directly operated in the solution result;
10. as described in step 7, if the client is not satisfied with the intelligent matching result, manual screening can be selected, and screening can be performed according to regions, practice years, adequacy fields and the like of lawyers in a traditional manner;
11. whether manual screening or automatic matching is adopted, after the final matching is successful, the two parties can carry out problem consultation and demand communication on line, and meanwhile, the system provides management on the condition of the agent case, such as the progress of case query and the like;
12. after the agent or consultation is finished, the client evaluates the service quality, the service attitude and the like of the lawyer, saves the evaluation result and updates the evaluation result into the lawyer information database;
13. and finally, performing library falling on the related information of the agent case for subsequent query and analysis.
Example 2
For the above intelligent distribution method and system, another processing mode is extended from the intelligent matching between the service attorney and the client on the matching distribution layer, the other contents are the same as those in embodiment 1, and specifically, the following are different in the matching distribution layer:
the system is positioned on a matching distribution layer and comprises a proxy case queue, a lawyer scoring module, a service type queue and an intelligent distribution module, information in the proxy case queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, and meanwhile the priority of a client in the proxy case queue is adjusted by combining an analysis result of an emotion analysis technology.
A service type queue of a structured data type is constructed in advance, and a lawyer scoring module integrates the processed offline features and real-time index features of lawyers through a deep learning technology to calculate the real-time scores of the service lawyers in different fields (dispute types) and update a service type list in real time; establishing an objective function and constraint conditions, wherein the constraint conditions are established according to design ideas such as lawyer fatigue degree control, new lawyer and deep lawyer case resource distribution balance control and the like; and the intelligent distribution module optimally matches multiple lawyers and multiple clients in the service type queue and the agent case queue by adopting a task assignment technology based on operational research optimization, and finally the matching distribution layer outputs the well-matched optimal results to an interface for display.
The above description is intended to be illustrative of the present invention and not to limit the scope of the invention, which is defined by the claims appended hereto.
Claims (7)
1. An intelligent distribution method and system for legal case agents is characterized in that: the system comprises a user interface layer, a characteristic processing layer, a matching distribution layer, a data storage layer and a monitoring management layer;
the method comprises the steps that a client expresses case situations on a user interface layer and finally transfers the case situations into text display, dispute types and appeal with finer granularity are predicted by means of a multi-label text classification technology based on deep learning, prediction of the multiple appeal is supported, and meanwhile, the urgency of the case appeal is analyzed and judged by means of an emotion analysis technology in natural language processing;
the off-line characteristics refer to client portrait information processed in non-real time and lawyer static information in the data layer, and are analyzed from the data layer in a T +1 mode by adopting an autonomously developed interface; the real-time indexes refer to lawyer dynamic indexes and monitoring indexes of a service process;
the matching distribution layer comprises a proxy case queue, a lawyer scoring module, a service lawyer queue and an intelligent distribution module, wherein the information in the proxy case queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, and meanwhile the priority of clients in the proxy case queue is adjusted by combining the analysis result of an emotion analysis technology; the lawyer scoring module integrates the processed offline features and the real-time index features through a deep learning technology to calculate a matching score so as to update the calculation process of correcting the score of the lawyer in real time, and further update the recommendation sequence and resource allocation of the service lawyer list; after the matching score is obtained through calculation, firstly establishing an objective function and constraint conditions when solving an optimization problem, wherein the constraint conditions are established according to a design idea of balanced distribution control on lawyer fatigue degree control and new lawyer and seniority lawyer case resources; the intelligent distribution module adopts a task assignment technology based on operational research optimization to optimally match a plurality of lawyers and a plurality of clients in a service lawyer queue and a proxy case queue, and finally a matching distribution layer outputs an optimal matching result to an interface for displaying;
the data storage layer comprises an authentication law database and an agent case database, and the authentication law database can persistently match partial dynamic indexes and static information generated in the distribution process.
2. An intelligent distribution method and system for legal case agents is characterized in that: the system comprises a user interface layer, a characteristic processing layer, a matching distribution layer, a data storage layer and a monitoring management layer;
the method comprises the steps that a client expresses case situations on a user interface layer and finally transfers the case situations into text display, dispute types and appeal with finer granularity are predicted by means of a multi-label text classification technology based on deep learning, prediction of the multiple appeal is supported, and meanwhile, the urgency of the case appeal is analyzed and judged by means of an emotion analysis technology in natural language processing;
the off-line characteristics refer to client portrait information processed in non-real time and lawyer static information in the data layer, and are analyzed from the data layer in a T +1 mode by adopting an autonomously developed interface; the real-time index refers to the monitoring index of the lawyer real-time workload and service process;
the matching distribution layer comprises a proxy case queue, a lawyer scoring module, a service type queue and an intelligent distribution module, wherein the information in the proxy case queue covers dispute types and specific appeal predicted by a user interface layer and client portrait information, and meanwhile the priority of the client in the proxy case queue is adjusted by combining the analysis result of an emotion analysis technology; a service type queue of a structured data type is constructed in advance, and a lawyer scoring module integrates the processed offline characteristics and real-time index characteristics of lawyers through a deep learning technology to calculate the real-time scores of the service lawyers in different fields and update a service type list in real time; establishing an objective function and constraint conditions, wherein the constraint conditions are established according to the design ideas of lawyer fatigue degree control and distribution balance control on new lawyers and seniority lawyers case resources; the intelligent distribution module adopts a task assignment technology based on operational research optimization to optimally match multiple lawyers and multiple clients in the service type queue and the agent case queue, and finally the matching distribution layer outputs an optimal matching result to an interface for displaying;
the data storage layer comprises an authentication law database and an agent case database, and the authentication law database can persistently match partial dynamic indexes and static information generated in the distribution process.
3. The intelligent distribution method and system of legal case agents as claimed in claim 1 or 2, wherein: the user interface layer is an interactive page of the system and the client and comprises a description interface module for providing case text input or voice input for the client, an online communication interface for carrying out prior consultation and communication between the client and lawyers, a lawyer display interface obtained according to the case description and basic information of the client in a matching mode, and a manual screening interface for supporting the lawyer which can be manually screened to meet the requirements of the client if the matching result is not satisfactory.
4. The intelligent distribution method and system of legal case agents as claimed in claim 1 or 2, wherein: the fatigue control is carried out statistical analysis according to the real-time workload and the historical workload of the service lawyer, and a fatigue control factor is calculated; the distribution balance control adopts the mode of adding new attorney protection factors and properly adjusting the weight distribution of attorney scoring factors so as to strengthen the assessment of attorneys on service attitude and service quality.
5. The intelligent distribution method and system of legal case agents as claimed in claim 1 or 2, wherein: and finally solving the result by the task assignment technology to obtain a matching matrix of the M clients N service lawyers, and after the matching matrix is obtained, adjusting the matching result according to some set rules, wherein the adjustment rules provide a tool for controlling the matching result for system management and operators.
6. The intelligent distribution method and system of legal case agents as claimed in claim 1 or 2, wherein: the data storage layer adopts a relational database, the retrieval architecture adopts a distributed and extensible real-time search and analysis engine, namely an Elasticissearch to construct a search function, the Elasticissearch is a search engine established on the basis of a full-text search engine Apache Lucene, and the full-text search function of distributed multi-user capability can be provided.
7. The intelligent distribution method and system of legal case agents as claimed in claim 1 or 2, wherein: the monitoring management layer comprises a visualization module for historical agent or case receiving situation of each lawyer, an authentication management module for qualification examination of lawyers applying for admission, a distribution monitoring module for monitoring a distribution process to prevent distribution errors such as repeated distribution of cases to the same lawyer, a service process monitoring module for monitoring a service process to prevent service risks of the lawyer or both clients, and a case agent management module for performing case agent process management and progress updating.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911298838.2A CN110727875B (en) | 2019-12-17 | 2019-12-17 | Intelligent distribution method and system for legal case agent |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911298838.2A CN110727875B (en) | 2019-12-17 | 2019-12-17 | Intelligent distribution method and system for legal case agent |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110727875A true CN110727875A (en) | 2020-01-24 |
CN110727875B CN110727875B (en) | 2020-05-08 |
Family
ID=69226029
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911298838.2A Active CN110727875B (en) | 2019-12-17 | 2019-12-17 | Intelligent distribution method and system for legal case agent |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110727875B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112527831A (en) * | 2021-02-05 | 2021-03-19 | 武汉听讼信息技术有限公司 | Resource processing method and device |
CN113360657A (en) * | 2021-06-30 | 2021-09-07 | 安徽商信政通信息技术股份有限公司 | Intelligent document distribution and handling method and device and computer equipment |
CN113704601A (en) * | 2021-06-18 | 2021-11-26 | 诺正集团股份有限公司 | Information interaction method, device, equipment and storage medium |
WO2022262752A1 (en) * | 2021-06-18 | 2022-12-22 | 诺正集团股份有限公司 | Information recommendation method and apparatus based on data interaction, and device and storage medium |
CN115759692A (en) * | 2022-12-01 | 2023-03-07 | 徐玉峰 | Moderator allocation method, system, medium and equipment based on two-dimensional code |
CN115809778A (en) * | 2022-12-06 | 2023-03-17 | 南北联合信息科技有限公司 | Intelligent legal case distribution system based on Internet of things |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110457357A (en) * | 2019-07-05 | 2019-11-15 | 中国平安财产保险股份有限公司 | Lawyer's recommended method neural network based and relevant device |
CN110502694A (en) * | 2019-07-23 | 2019-11-26 | 平安科技(深圳)有限公司 | Lawyer's recommended method and relevant device based on big data analysis |
-
2019
- 2019-12-17 CN CN201911298838.2A patent/CN110727875B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110457357A (en) * | 2019-07-05 | 2019-11-15 | 中国平安财产保险股份有限公司 | Lawyer's recommended method neural network based and relevant device |
CN110502694A (en) * | 2019-07-23 | 2019-11-26 | 平安科技(深圳)有限公司 | Lawyer's recommended method and relevant device based on big data analysis |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112527831A (en) * | 2021-02-05 | 2021-03-19 | 武汉听讼信息技术有限公司 | Resource processing method and device |
CN113704601A (en) * | 2021-06-18 | 2021-11-26 | 诺正集团股份有限公司 | Information interaction method, device, equipment and storage medium |
WO2022262752A1 (en) * | 2021-06-18 | 2022-12-22 | 诺正集团股份有限公司 | Information recommendation method and apparatus based on data interaction, and device and storage medium |
CN113360657A (en) * | 2021-06-30 | 2021-09-07 | 安徽商信政通信息技术股份有限公司 | Intelligent document distribution and handling method and device and computer equipment |
CN113360657B (en) * | 2021-06-30 | 2023-10-24 | 安徽商信政通信息技术股份有限公司 | Intelligent document distribution handling method and device and computer equipment |
CN115759692A (en) * | 2022-12-01 | 2023-03-07 | 徐玉峰 | Moderator allocation method, system, medium and equipment based on two-dimensional code |
CN115809778A (en) * | 2022-12-06 | 2023-03-17 | 南北联合信息科技有限公司 | Intelligent legal case distribution system based on Internet of things |
Also Published As
Publication number | Publication date |
---|---|
CN110727875B (en) | 2020-05-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110727875B (en) | Intelligent distribution method and system for legal case agent | |
US11853935B2 (en) | Automated recommendations for task automation | |
US20040039624A1 (en) | Sales activity support system, sales activity support information providing system, sales activity evaluation system, recording medium, and program | |
US20130262175A1 (en) | Ranking of jobs and job applicants | |
US11068758B1 (en) | Polarity semantics engine analytics platform | |
JP2009534736A (en) | Network-based information technology solution development management system and method | |
CN107993019A (en) | A kind of resume appraisal procedure and device | |
Kampf et al. | The identification and development of talents in the environment of logistics companies | |
US11461343B1 (en) | Prescriptive analytics platform and polarity analysis engine | |
JP7470971B2 (en) | How to support setting accountability for results | |
CN113672797A (en) | Content recommendation method and device | |
Holt et al. | A computer aided approach to employment service placement and counseling | |
Baltazar et al. | Spanish airports performance and efficiency benchmark. A PESA-AGB study | |
Ramanathan | Successful transfer of environmentally sound technologies for greenhouse gas mitigation: a framework for matching the needs of developing countries | |
CN108182512A (en) | A kind of resume appraisal procedure and device | |
Ball et al. | Service level expectation setting for air traffic flow management: Practical challenges and benefits assessment | |
CN108009735B (en) | Resume evaluation method and device | |
Nierling et al. | The landscape of crowd work in Germany: An overview of the scientific and public discourse | |
KR20040107971A (en) | Method and system for technology assessment | |
CN108182513A (en) | A kind of resume appraisal procedure and device | |
CN108153829A (en) | A kind of resume appraisal procedure and device | |
US20230101734A1 (en) | Machine learning model to fill gaps in adaptive rate shifting | |
KR20190140492A (en) | Proposal management system and method for change order request | |
CN117472300B (en) | Order dispatch method and system for distributed 3D printing manufacturing platform | |
Belova et al. | Modern methods of resume processing in recruiting information systems |
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 |