CN113342844A - Industrial intelligent search system - Google Patents

Abstract

The invention discloses an industrial intelligent search system, which comprises a dispatcher, a collector, a knowledge engine, an industrial dictionary and a word segmentation device, wherein the dispatcher is used for dispatching a plurality of words to be searched; the collector is used for collecting and processing data according to task requirements; the collector is in communication connection with the scheduler, and the scheduler is used for monitoring the execution condition of the collection task in real time according to the task list; the collector transmits the collected data to a database; the database is respectively in communication connection with the indexer and the knowledge engine; the indexer is in communication connection with the index database; the index database is respectively in communication connection with the industrial dictionary and the searcher; the industrial dictionary is in communication connection with the word segmentation device; the word segmentation device is used for segmenting a text into word lists; the invention can effectively improve the efficiency and the accuracy of information search of industrial enterprises and meet the requirement of manufacturing enterprise users on personalized search of internal information; the word stock in the industrial manufacturing field covers the whole process from design to scrapping and has strong pertinence.

Description

Industrial intelligent search system

Technical Field

The invention belongs to the technical field of retrieval, and particularly relates to an industrial intelligent search system.

Background

The information search system is widely applied to various information systems, knowledge bases and the like because of good performance. The information search system can quickly retrieve the required information from a large amount of information, usually with a response on the order of seconds. The retrieval system generally comprises a searcher, an indexer and a retriever, so that the good usability of the system is ensured, and the system has obvious advantages in the aspects of application range and performance by automatically capturing tasks, establishing an index library and carrying out matching in an evaluation mode.

With the continuous development of enterprise informatization and digitization, enterprise users put forward higher requirements for information search systems, and the diversity, processing capacity, efficiency and intelligence of retrieval objects. The search system on the market at present has single type of retrieval object, low assistant capability and poor intelligence, and is difficult to meet the higher requirements of enterprise users, particularly industrial enterprise users, on the aspect of information search. Industrial enterprises, particularly large manufacturing enterprises, generate large amounts of data each day; such as PLM, CSM, CRM, ERP, OA system, DCS, PLC, SCADA and various sensors; the amount of data may need to be calculated in TB, with drawing, picture, video, audio, SQL databases. Data structures encompass structured, unstructured, and semi-structured; the process data is mostly distributed to various databases and file management systems. The data format is not uniform, and even the data is invalid or incomplete. The general search tool can only deal with one type of data, and cannot perform station-type retrieval and analysis. In this case, a plurality of sets of retrieval tools need to be deployed, so that the purchase cost and the operation and maintenance difficulty are increased; the search requires high accuracy and the required information needs to be found in a large amount of data. The user needs to input keywords, but the keywords input by the user do not necessarily represent the content that the user wants to search for; in a large part of application scenarios, the purpose of the search is not to want the results but rather to solve the problem.

Disclosure of Invention

In order to solve the problem of difficult retrieval caused by industrial data diversity, the invention provides an industrial intelligent search system. The invention can effectively improve the efficiency and the accuracy of information search of industrial enterprises and meet the requirement of manufacturing enterprise users on personalized search of internal information; the word stock in the industrial manufacturing field covers the whole process from design to scrapping and has strong pertinence.

The purpose of the invention can be realized by the following technical scheme:

an industrial intelligent search system, comprising: the system comprises a scheduler, a collector, a knowledge engine, an industrial dictionary and a word segmentation device;

the collector is used for collecting and processing data according to task requirements; the collector is in communication connection with the scheduler, and the scheduler is used for monitoring the execution condition of the collection task in real time according to the task list;

the collector transmits the collected data to a database; the database is respectively in communication connection with the indexer and the knowledge engine; the indexer is in communication connection with the index database; the index database is respectively in communication connection with the industrial dictionary and the searcher;

the industrial dictionary is in communication connection with the word segmentation device; the word segmentation device is used for segmenting a piece of text into word lists.

Further, the data processing steps of the collector include data cleaning, data conversion, data description, feature selection and feature extraction.

Further, the collector is deployed using distributed components.

Furthermore, the scheduler monitors the execution condition of the acquired tasks in real time according to the task list, and divides the acquired tasks into a plurality of task groups to be executed in a task queue mode; after the execution is finished, feeding the result back to the scheduler to judge whether the task needs to be executed again; meanwhile, the scheduler selects an execution time point according to the use condition of the resources and the strategy; collecting information with high timeliness requirements in real time according to requirements; and initiating the tasks in batches on time for the information with low timeliness requirement.

Further, the knowledge engine realizes automatic knowledge category building and alignment through knowledge modeling; and building an industrial manufacturing professional knowledge system through a knowledge graph, and completing recommendation of knowledge and solutions based on Rule-base.

Further, the searcher is used for the user to retrieve data, and the specific method comprises the following steps:

step SA 1: acquiring a search word input by a user;

step SA 2: analyzing the search word category, and acquiring a corresponding data code base according to the search word category;

step SA 3: and searching in a corresponding data code base, and recommending the MRO information, the operation instruction and the related operation instruction file of the equipment.

Further, the MRO information of the device is used to determine the root cause of the fault.

Further, the word segmentation device carries out matching through a maximum matching algorithm, and carries out optimized matching, statistics and sequencing through a hash tree and an HMM algorithm.

Furthermore, the accuracy of word segmentation is improved through training;

firstly, preparing a corpus with words as a training level;

secondly, labeling each word, and respectively counting transition probability, the state of each word, the frequency of each word in each state and the frequency of each label;

and then processed using the viterbi algorithm.

Compared with the prior art, the invention has the beneficial effects that: the invention can effectively improve the efficiency and the accuracy of information search of industrial enterprises and meet the requirement of manufacturing enterprise users on personalized search of internal information; the word stock in the industrial manufacturing field covers the whole process from design to scrapping, and has strong pertinence; meanwhile, the word stock supports Chinese and English; advanced algorithms such as machine learning and the like are adopted in the word segmentation process, and the word segmentation accuracy rate is over 99% through long-term training; a set of recommendation algorithm is built in to meet the requirements of relevant knowledge recommendation in the field of production and manufacturing; searching keyword recommendation, solving processing schemes and historical records through a user; the problem of search object diversification is solved, the problem that "intelligence" degree is not enough is solved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic block diagram of the present invention;

FIG. 2 is a search feedback diagram according to the present invention.

Detailed Description

The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1-2, an industrial intelligent search system, raw data, a scheduler, a collector, a knowledge engine, an industrial dictionary, and a word segmenter;

the original data oriented by the invention mainly comprises a database and a database file management system;

the database can be an SQL database and an NOSQL database; most of core information in enterprise management systems such as ERP, CRM, WMS and MES exists in an SQL database; the SQL database is stored through a data table and managed by a DBMS; data such as production plans, accounting information, equipment assets, production execution, material inventory and the like are managed through an SQL database; the file management system processes product models, drawings, training videos and the like;

the scheduler monitors the execution condition of the acquired tasks in real time according to the task list, and divides the acquired tasks into a plurality of task groups to be executed in a task queue mode; after the execution is finished, feeding the result back to the scheduler to judge whether the task needs to be executed again; meanwhile, the scheduler selects an execution time point according to the use condition of the resources and the strategy; collecting information with high timeliness requirements in real time according to requirements; initiating tasks in batches on time for information with low timeliness requirements;

the collector collects and processes various data according to task requirements; the collector adopts distributed components for deployment, and a container mode is recommended; different collectors are arranged in the system due to different formats; the acquired data needs to be subjected to necessary pre-processing: for example, Data cleaning (Data cleaning), Data conversion (Data Transformation), Data Description (Data Description), Feature Selection (Feature Selection), and Feature Extraction (Feature Extraction) are performed to remove dirty Data, ensure Data quality, and further save storage space;

the collector transmits the collected data to a database; the database is respectively in communication connection with the indexer and the knowledge engine;

the knowledge engine realizes automatic knowledge category building and alignment through knowledge modeling; building an industrial manufacturing professional knowledge system through a knowledge graph, and completing recommendation of knowledge and a solution based on Rule-base;

the knowledge map is called knowledge domain visualization or knowledge domain mapping map in the book intelligence world, is a series of different graphs for displaying the relationship between the knowledge development process and the structure, describes knowledge resources and carriers thereof by using visualization technology, and excavates, analyzes, constructs, draws and displays knowledge and the mutual relation between the knowledge resources and the carriers;

searching needs high accuracy, and information needed to be found in a large amount of data is needed; the user needs to input keywords, but the keywords input by the user do not necessarily represent the content that the user wants to search for; in a large part of application scenarios, the purpose of the search is not to want the results but rather to solve the problem; how to recommend valuable information by the user through the search behavior of the user is also the scope of the scheme;

the indexer is in communication connection with the index database; the index database is respectively connected with an industrial dictionary and a searcher in a communication way, the searcher is used for a user to retrieve data, and the specific method comprises the following steps:

step SA 1: acquiring a search word input by a user;

step SA 2: analyzing the category of the search word, wherein the category of the search word is the category of the data in the corresponding index library, and acquiring a corresponding data code library according to the category of the search word; for example: the fault code corresponds to the fault code library;

step SA 3: searching in a corresponding data code base, and recommending equipment MRO information, operation instructions and related operation guide files;

as shown in fig. 2, when the search term of the user is: the system firstly detects related information in a fault code library and recommends equipment MRO information, an operation description and a related operation instruction file; MRO information of the equipment is used for judging the root cause of the fault; if the accident or the general phenomenon occurs or not and the fault caused by some factors are judged by analyzing the maintenance record, the operation record, the working condition information and the inspection record;

the word bank is one of the core components of the invention, and determines the accuracy and the intelligence degree of searching; the word stock in the market is relatively weak in pertinence, and the emphasis is universality; such as the Chinese thesaurus of THUOCL, the subjects are IT, finance and economics, idioms, place names, historical celebrities, poems, medicine, catering and the like;

the industrial dictionary mainly contains words of aerospace, ship, automobile manufacturing, instruments and meters, equipment maintenance and metal product industries, and various word banks are manufactured from design to customer requirements to sales, orders, planning, research and development, design, process, manufacturing, purchasing, supply, inventory, delivery and delivery, after-sales service, operation and maintenance, scrapping or recycling; the number of Chinese and English entries which are already collected is 2000W;

the word stock in the industrial manufacturing field covers the whole process from design to scrapping and has strong pertinence. Meanwhile, the word stock supports Chinese and English, and can support German and Japanese languages in the later period;

the industrial dictionary is in communication connection with the word segmentation device; the word segmentation device is used for segmenting a text into word lists;

the word segmentation device is based on a word bank and splits a section of text into a word list through splitting; the word segmentation device carries out matching through various modes, such as forward maximum matching, reverse maximum matching and bidirectional maximum matching, and carries out optimized matching, statistics and sequencing through a Hash tree and an HMM algorithm; the accuracy of word segmentation can be effectively improved through training; firstly, preparing a corpus with words as a training level; secondly, labeling each word, and respectively counting transition probability, the state of each word, the frequency of each word in each state and the frequency of each label; then processing by adopting a Viterbi algorithm;

advanced algorithms such as machine learning and the like are adopted in the word segmentation process, and the word segmentation accuracy rate is over 99% through long-term training;

in computer science, a hash tree (or hash trie) is a persistent data structure that can be used to implement collections and mappings, intended to replace hash tables in purely functional programming; in its basic form, a hash tree stores hash values (treated as bit strings) of its keys in a trie, with the actual keys and (optional) values stored in the "final" node of the trie;

the maximum matching algorithm mainly comprises a forward maximum matching algorithm, a reverse maximum matching algorithm, a two-way matching algorithm and the like; the main principle is that single character strings are cut out and then compared with a word bank, if a word is a word, the word is recorded, otherwise, a single character is added or reduced, the comparison is continued, and a single character is remained all the time, the comparison is terminated, and if the single character strings can not be cut, the word strings are treated as unregistered;

for example: i eat alone;

the reverse maximum matching mode is adopted, and the maximum length is 5;

the person has a meal;

people have a meal;

eating = = = get one word-eat;

i a person;

a person;

person = = = get one word-person;

i, I;

one = = = get one word-one;

i = = = get a word-me;

the final result of the reverse max match is/I/personal/meal/;

and (3) a forward maximum matching algorithm, namely matching a plurality of continuous characters in the text to be segmented with the word list from left to right, and segmenting a word if the continuous characters are matched with the word list. But there is a problem that the maximum matching is not matched for the first time but can be split;

the viterbi algorithm is a dynamic programming algorithm used to find the sequence of-viterbi paths-hidden states that are most likely to produce a sequence of observed events, particularly in the context of markov information sources and hidden markov models. The terms "viterbi path" and "viterbi algorithm" are also used to find the dynamic programming algorithm for which observations are most likely to explain the correlation. Dynamic programming algorithms, such as in statistical syntactic analysis, can be used to find the most likely context-free derived (parsed) strings, sometimes referred to as "viterbi analysis".

A container management platform: the search system supports multi-environment flexible deployment, such as: deployment in a physical machine, a virtual machine and a container; each component of the search system supports distributed deployment, and different personalized requirements of users are met.

The working principle of the invention is as follows: the collector collects and processes various data according to task requirements; the collector adopts distributed components for deployment, and a container mode is recommended; different collectors are arranged in the system due to different formats; the acquired data needs to be subjected to necessary pre-processing: such as data cleaning, data conversion, data description, feature selection and feature extraction, removes dirty data, ensures data quality and further saves storage space; the scheduler monitors the execution condition of the acquired tasks in real time according to the task list, and divides the acquired tasks into a plurality of task groups to be executed in a task queue mode; after the execution is finished, feeding the result back to the scheduler to judge whether the task needs to be executed again; meanwhile, the scheduler selects an execution time point according to the use condition of the resources and the strategy; collecting information with high timeliness requirements in real time according to requirements; initiating tasks in batches on time for information with low timeliness requirements; the collector transmits the collected data to a database; the database is respectively in communication connection with the indexer and the knowledge engine;

the knowledge engine realizes automatic knowledge category building and alignment through knowledge modeling; building an industrial manufacturing professional knowledge system through a knowledge graph, and completing recommendation of knowledge and a solution based on Rule-base; the indexer is in communication connection with the index database; the index library is respectively in communication connection with the industrial dictionary and the searcher, and the searcher is used for searching data by a user and acquiring search words input by the user; analyzing the search word category, and acquiring a corresponding data code base according to the search word category; searching in a corresponding data code base, and recommending equipment MRO information, operation instructions and related operation guide files;

the industrial dictionary mainly contains words of aerospace, ship, automobile manufacturing, instruments and meters, equipment maintenance and metal product industries, and various word banks are manufactured from design to customer requirements to sales, orders, planning, research and development, design, process, manufacturing, purchasing, supply, inventory, delivery and delivery, after-sales service, operation and maintenance, scrapping or recycling; the word stock in the industrial manufacturing field covers the whole process from design to scrapping and has strong pertinence. Meanwhile, the word stock supports Chinese and English, and can support German and Japanese languages in the later period; the industrial dictionary is in communication connection with the word segmentation device; the word segmentation device is used for segmenting a text into word lists; the word segmentation device is based on a word bank and splits a section of text into a word list through splitting; the word segmentation device carries out matching through various modes, such as forward maximum matching, reverse maximum matching and bidirectional maximum matching, and carries out optimized matching, statistics and sequencing through a Hash tree and an HMM algorithm; the accuracy of word segmentation can be effectively improved through training; firstly, preparing a corpus with words as a training level; secondly, labeling each word, and respectively counting transition probability, the state of each word, the frequency of each word in each state and the frequency of each label; then processing by adopting a Viterbi algorithm; the word segmentation process adopts advanced algorithms such as machine learning and the like, and the word segmentation accuracy rate exceeds 99 percent through long-term training.

In the embodiments provided by the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and there may be other divisions when the actual implementation is performed; the modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the method of the embodiment.

It will also be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.

The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.

Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms second, etc. are used to denote names, but not any particular order.

Finally, it should be noted that the above examples are only intended to illustrate the technical process of the present invention and not to limit the same, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical process of the present invention without departing from the spirit and scope of the technical process of the present invention.

Claims (9)

1. An industrial intelligent search system, comprising: the system comprises a scheduler, a collector, a knowledge engine, an industrial dictionary and a word segmentation device;

the collector is used for collecting and processing data according to task requirements; the collector is in communication connection with the scheduler, and the scheduler is used for monitoring the execution condition of the collection task in real time according to the task list;

the collector transmits the collected data to a database; the database is respectively in communication connection with the indexer and the knowledge engine; the indexer is in communication connection with the index database; the index database is respectively in communication connection with the industrial dictionary and the searcher;

the industrial dictionary is in communication connection with the word segmentation device; the word segmentation device is used for segmenting a piece of text into word lists.

2. The industrial intelligent search system of claim 1, wherein the data processing by the collector comprises data cleaning, data conversion, data description, feature selection and feature extraction.

3. The industrial intelligent search system of claim 1, wherein the collectors are deployed using distributed components.

4. The industrial intelligent search system according to claim 1, wherein the scheduler monitors the execution of the collection tasks in real time according to the task list, and divides the collection tasks into a plurality of task groups to be executed in a task queue manner; after the execution is finished, feeding the result back to the scheduler to judge whether the task needs to be executed again; meanwhile, the scheduler selects an execution time point according to the use condition of the resources and the strategy; collecting information with high timeliness requirements in real time according to requirements; and initiating the tasks in batches on time for the information with low timeliness requirement.

5. The industrial intelligent search system according to claim 1, wherein the knowledge engine realizes automatic knowledge category building and alignment through knowledge modeling; and building an industrial manufacturing professional knowledge system through a knowledge graph, and completing recommendation of knowledge and solutions based on Rule-base.

6. The industrial intelligent search system of claim 1, wherein the searcher is configured to retrieve data from a user, and the specific method comprises:

step SA 1: acquiring a search word input by a user;

step SA 2: analyzing the search word category, and acquiring a corresponding data code base according to the search word category;

step SA 3: and searching in a corresponding data code base, and recommending the MRO information, the operation instruction and the related operation instruction file of the equipment.

7. The industrial intelligent search system of claim 6, wherein the MRO information of a device is used to determine the root cause of the fault.

8. The industrial intelligent search system of claim 1, wherein the word segmenter performs matching through a maximum matching algorithm, and performs optimized matching and statistics and sorting through a hash tree and an HMM algorithm.

9. The industrial intelligent search system of claim 8, wherein the accuracy of word segmentation is improved by training;

firstly, preparing a corpus with words as a training level;

secondly, labeling each word, and respectively counting transition probability, the state of each word, the frequency of each word in each state and the frequency of each label;

and then processed using the viterbi algorithm.

Images