CN111797213A - Method for mining financial risk clues from unstructured network information - Google Patents
Method for mining financial risk clues from unstructured network information Download PDFInfo
- Publication number
- CN111797213A CN111797213A CN202010580841.XA CN202010580841A CN111797213A CN 111797213 A CN111797213 A CN 111797213A CN 202010580841 A CN202010580841 A CN 202010580841A CN 111797213 A CN111797213 A CN 111797213A
- Authority
- CN
- China
- Prior art keywords
- financial
- information
- feature word
- model
- entity
- 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.)
- Withdrawn
Links
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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
Abstract
The invention relates to the technical field of financial prevention and control, in particular to a method for mining financial risk clues from unstructured network information, which comprises the following steps of S1, constructing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list; s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature lexicon and the list are selected in S1, the staff can use the feature lexicon to construct co-occurrence feature word combinations. The method has the advantages that the accuracy of mining the financial risk prevention and control clues on the Internet can be improved, the screening of the financial risk prevention and control clues can be accelerated when the financial risk prevention and control clues are screened, and therefore the efficiency of screening the financial risk prevention and control clues is improved.
Description
Technical Field
The invention relates to the technical field of financial prevention and control, in particular to a method for mining financial risk clues from unstructured network information.
Background
In massive unstructured internet data, financial risk prevention and control clues are searched for the fishing needles without difference from the fishing needles in the sea.
The Chinese patent No. CN109522416A provides a financial risk control knowledge map construction method which can fully utilize continuous conversion from data to knowledge, quickly integrate financial data from different sources by constructing a financial risk control knowledge map, construct an anti-fraud engine and quickly and efficiently identify financial fraud cases; extracting knowledge from the financial data by using a deep belief network, wherein the knowledge comprises entities and relationships and attributes among the entities; and taking the target entity as a knowledge graph node to obtain a financial risk control knowledge graph and storing the financial risk control knowledge graph in a Neo4j graph database.
Although the financial risk control knowledge graph can identify financial fraud cases, the information searched from the internet not only has valuable negative public sentiment, but also has a large amount of non-negative public sentiment including the name of the enterprise, and meanwhile, because the entities involved in illegal financial activities are uncertain, the direct search according to the monitoring list cannot be carried out, so that the development of a method for mining financial risk clues from unstructured network information is urgently needed to solve the problems.
Disclosure of Invention
The invention aims to provide a method for mining financial risk clues from unstructured network information, which solves the problems that the information searched from the internet, which is proposed in the background art, not only has valuable negative public sentiment, but also has a large amount of non-negative public sentiment containing the name of the enterprise and entities related to illegal financial activities are uncertain, and cannot be directly searched according to a monitoring list.
The technical scheme of the invention is as follows: a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the S1, a worker can use the feature word library to construct a co-occurrence feature word combination, the combination length is 1-3, meanwhile, the co-occurrence feature word combination can also be manually selected, and then the worker can select the capturing timeliness length, wherein the length is 6-36 h;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Further, the information collected by the illegal financial activity feature lexicon in S1 is of a second type, and the monitoring entity list constructed in S1 is of a first type.
Further, in S1, the words or phrases specific to the illegal financial activity are selected as high interest, rolling money, running.
Further, the co-occurrence feature words in S2 are behavior feature words appearing in the same content, and the feature words are combined into rolling money, running, no cash withdrawal, and the like.
Further, if the combination length in S2 is 1, the co-occurrence feature word combination may be represented as rolling money, running, and unable to be presented.
Further, if the combination length is 2 in S2, there are three combinations of < roll payment, running >, < roll payment, cash withdrawal >, < running, cash withdrawal impossible >.
Further, if the combination length is 3 in S2, there is < rolling, running, and withdrawal impossible > one combination.
Further, for the extraction of the second type information in S4, a financial entity extraction model is selected for extraction, and the financial entity extraction model can be used to focus on only the financial entity to shield interference of other entities, and for the first type, the step is skipped to perform the next processing in S4.
Further, in S4, a financial entity extraction model is used for extraction, and the financial entity extraction model is a general-domain NER model based on the NLP pre-training model and a specific-domain NER model based on the NLP pre-training model.
Further, the fused entity negative information recognition model in S5 is a model obtained based on deep learning and pre-training model training, and the input of the model is an entity and its text, and the output is a result of determining whether the text is the negative public opinion of the entity, and the text length is generally 128/256/512 characters according to the setting of the pre-training model.
The invention provides a method for mining financial risk clues from unstructured network information by improving, compared with the prior art, the following improvements and advantages are provided:
(1) the financial entity negative information identification model designed by the invention can judge the extracted financial entity and text when processing the duplication-removing information, thereby obtaining whether the information is negative public opinion or not, and being beneficial to improving the accuracy of mining financial risk prevention and control clues on the Internet.
(2) The feature word bank and the co-occurrence feature words adopted by the invention can accelerate the screening of the financial risk prevention and control clues when the financial risk prevention and control clues are screened, thereby improving the efficiency of screening the financial risk prevention and control clues.
(3) The financial entity extraction model adopted by the invention is a general field NER model based on an NLP pre-training model and a special field NER model based on the NLP pre-training model, and can improve the extraction rate of public sentiment entity identification in the financial field when extracting duplication-removing information.
Drawings
The invention is further explained below with reference to the figures and examples:
FIG. 1 is a flow chart of the architecture of the present invention.
Detailed Description
The present invention will be described in detail below with reference to the attached drawings, and the technical solutions in the embodiments of the present invention will be clearly and completely described. 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.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
The invention provides a method for mining financial risk clues from unstructured network information through improvement, as shown in FIG. 1, comprising the following steps:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 2, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 24 hours;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Further, the information collected by the illegal financial activity feature lexicon in the step S1 is the second type of information, and the monitoring entity list constructed in the step S1 is the first type of information.
Further, in S1, the words or phrases specific to the illegal financial activity are selected as high interest, rolling money, running.
Further, the co-occurrence feature words in S2 are behavior feature words appearing in the same content, and the feature words are combined as a rolling, running, no cash withdrawal, and the like.
Further, if the combination length in S2 is 1, the co-occurrence feature word combination may be represented as rolling, running, and cannot be presented.
Further, if the combination length is 2 in S2, there are three combinations of < roll payment, running >, < roll payment, cash withdrawal impossible >, < running, cash withdrawal impossible >.
Further, if the combination length is 3 in S2, there is < rolling, running, and withdrawal impossible > one combination.
Further, for the extraction of the second type information in S4, a financial entity extraction model is selected for extraction, and the financial entity extraction model can be used to focus on only the financial entity to shield interference of other entities, and for the first type, the step is skipped to perform the next processing in S4.
Further, in S4, a financial entity extraction model is used for extraction, and the financial entity extraction model is a general-domain NER model based on the NLP pre-training model and a specific-domain NER model based on the NLP pre-training model.
Further, the fused entity negative information recognition model in S5 is a model obtained by deep learning and pre-training model training, and the input of the model is an entity and its text, and the output is the result of determining whether the text is the negative public opinion of the entity, and the text length is generally 128/256/512 characters according to the setting of the pre-training model.
The first embodiment is as follows:
a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 1, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 24 hours;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Example two:
a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 2, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 24 hours;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Example three:
a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 3, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 24 hours;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Example four:
a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 2, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 12 h;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Example five:
a method for mining financial risk cues from unstructured network information, comprising the steps of:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the step S1, the worker may use the feature word library to construct a co-occurrence feature word combination, the combination length is 1, and at the same time, the co-occurrence feature word combination may also be manually selected, and then the worker may select a capturing aging length, the length is 12 h;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
Through comparison of the above examples, the result obtained by the second embodiment is fastest and highest in accuracy rate by calculating data through a computer, so that the second embodiment is the best embodiment of the invention.
The working principle of the invention is as follows: constructing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list; constructing a co-occurrence feature word combination and selecting capturing timeliness: after the feature word library and the list are selected in the above steps, a worker can use the feature word library to construct a co-occurrence feature word combination, the combination length is 3, meanwhile, the co-occurrence feature word combination can also be manually selected, and then the worker can select the capturing timeliness length, wherein the capturing timeliness length is 24 hours; information grabbing and de-duplication: the staff can use all the co-occurrence feature word combinations in the last step as search keywords to capture information data in a set time limit from the internet, and meanwhile, the staff can capture the information data containing name of the name list entity in the internet according to the monitored name list selected in S1, and then the data is de-duplicated after the data capture is finished; and (3) financial entity extraction: the staff can extract the data after the duplication is removed in the last step; and (3) identifying by utilizing a financial entity negative information identification model: and inputting the data received in the steps into a financial entity negative information identification model, selecting the data as negative public sentiment if the data is judged to be negative, and giving up the information if the data is judged to be non-negative.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method for mining financial risk cues from unstructured network information, characterized by: the method comprises the following steps:
s1, establishing an illegal financial activity feature word bank and an entity monitoring list: the method comprises the following steps that a worker collects information of illegal financial activities from the Internet, selects words or phrases specific to the illegal financial activities from the information to form a feature word bank, and meanwhile, the worker can list enterprise or product lists from the Internet to build a monitoring list;
s2, constructing a co-occurrence feature word combination and selecting a capturing time effect: after the feature word library and the list are selected in the S1, a worker can use the feature word library to construct a co-occurrence feature word combination, the combination length is 1-3, meanwhile, the co-occurrence feature word combination can also be manually selected, and then the worker can select the capturing timeliness length, wherein the length is 6-36 h;
s3, information grabbing and de-weighting: the staff can use all the co-occurrence feature word combinations in the S2 as search keywords to capture information data in a set time limit from the Internet, and meanwhile, the staff can capture the information data containing name of the list entity in the Internet according to the monitored list selected in the S1, and then the data is de-duplicated after the data capture is finished;
s4, financial entity extraction: the staff can extract the data after the duplication in the S3;
s5, identifying by using a financial entity negative information identification model: and inputting the data collected in the step S4 into a financial entity negative information identification model, selecting negative public sentiment if the model is judged to be negative, and abandoning the information if the model is judged to be non-negative.
2. The method of claim 1, wherein the method comprises the steps of: the information collected by the illegal financial activity feature word bank in the step S1 is the second type of information, and the monitoring entity list constructed in the step S1 is the first type of information.
3. The method of claim 1, wherein the method comprises the steps of: and selecting words or phrases specific to the illegal financial activities in the S1, wherein the words or phrases are high interest credits, rolling money and running.
4. The method of claim 1, wherein the method comprises the steps of: the co-occurrence feature words in the S2 are behavior feature words appearing in the same content, and the feature words are combined into money, running, cash withdrawal failure and the like.
5. The method of claim 1, wherein the method comprises the steps of: if the combination length in S2 is 1, the co-occurrence feature word combination may be represented as a rolling money, a running, and cannot be presented.
6. The method of claim 1, wherein the method comprises the steps of: if the combination length is 2 in S2, there are three combinations of < rolling, running >, < rolling, withdrawal impossible >, < running, withdrawal impossible >.
7. The method of claim 1, wherein the method comprises the steps of: if the combination length is 3 in S2, there is a combination of < rolling, running, and no cash-out >.
8. The method of claim 1, wherein the method comprises the steps of: in the step S4, for the extraction of the second type of information, a financial entity extraction model is selected for extraction, and the financial entity extraction model can be used to focus on only the financial entity to shield interference of other entities, and in the step S4, for the first type, the step is skipped to perform the next processing.
9. The method of claim 1, wherein the method comprises the steps of: and in the step S4, a financial entity extraction model is adopted for extraction, and the financial entity extraction model is a general domain NER model based on an NLP pre-training model and a special domain NER model based on the NLP pre-training model.
10. The method of claim 1, wherein the method comprises the steps of: the fused entity negative information recognition model in S5 is a model obtained by deep learning and pre-training model training, the input of the model is an entity and its text, and the output is the result of determining whether the text is the entity negative public opinion, the text length is generally 128/256/512 characters according to the pre-training model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010580841.XA CN111797213A (en) | 2020-06-23 | 2020-06-23 | Method for mining financial risk clues from unstructured network information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010580841.XA CN111797213A (en) | 2020-06-23 | 2020-06-23 | Method for mining financial risk clues from unstructured network information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111797213A true CN111797213A (en) | 2020-10-20 |
Family
ID=72803712
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010580841.XA Withdrawn CN111797213A (en) | 2020-06-23 | 2020-06-23 | Method for mining financial risk clues from unstructured network information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111797213A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113240241A (en) * | 2021-04-14 | 2021-08-10 | 北京蓝光讯智科技有限责任公司 | Internet financial clue analysis method, system and device based on WeChat data |
-
2020
- 2020-06-23 CN CN202010580841.XA patent/CN111797213A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113240241A (en) * | 2021-04-14 | 2021-08-10 | 北京蓝光讯智科技有限责任公司 | Internet financial clue analysis method, system and device based on WeChat data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110020422B (en) | Feature word determining method and device and server | |
| CN112699246B (en) | Domain knowledge pushing method based on knowledge graph | |
| CN104572958A (en) | Event extraction based sensitive information monitoring method | |
| CN110297988A (en) | Hot topic detection method based on weighting LDA and improvement Single-Pass clustering algorithm | |
| CN109408804A (en) | The analysis of public opinion method, system, equipment and storage medium | |
| CN111767716A (en) | Method and device for determining enterprise multilevel industry information and computer equipment | |
| CN103854064A (en) | Event occurrence risk prediction and early warning method targeted to specific zone | |
| CN110909542B (en) | Intelligent semantic serial-parallel analysis method and system | |
| Ronan et al. | Determining light verb constructions in contemporary British and Irish English | |
| CN110287329A (en) | A kind of electric business classification attribute excavation method based on commodity text classification | |
| CN112116168B (en) | User behavior prediction method and device and electronic equipment | |
| CN112883734B (en) | Block chain security event public opinion monitoring method and system | |
| CN103853700A (en) | Event forewarning method based on regions and object information discovery | |
| CN107526721A (en) | A kind of disambiguation method and device to electric business product review vocabulary | |
| CN114385775A (en) | Sensitive word recognition method based on big data | |
| CN110413998B (en) | Self-adaptive Chinese word segmentation method oriented to power industry, system and medium thereof | |
| WO2015030112A1 (en) | Document sorting system, document sorting method, and document sorting program | |
| CN110287493A (en) | Risk phrase chunking method, apparatus, electronic equipment and storage medium | |
| CN111797213A (en) | Method for mining financial risk clues from unstructured network information | |
| CN117420998A (en) | Client UI interaction component generation method, device, terminal and medium | |
| US11789983B2 (en) | Enhanced data driven intelligent cloud advisor system | |
| CN109308572A (en) | The expected performance evaluation method of project of inviting outside investment based on policy goals guiding | |
| CN114443930A (en) | News public opinion intelligent monitoring and analyzing method, system and computer storage medium | |
| CN110705597B (en) | Network early event detection method and system based on event cause and effect extraction | |
| CN113158672A (en) | Relation analysis method and device based on news events |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20201020 |
|
| WW01 | Invention patent application withdrawn after publication |