CN107507052B - A method and device for obtaining quotation information - Google Patents
A method and device for obtaining quotation information Download PDFInfo
- Publication number
- CN107507052B CN107507052B CN201710581774.1A CN201710581774A CN107507052B CN 107507052 B CN107507052 B CN 107507052B CN 201710581774 A CN201710581774 A CN 201710581774A CN 107507052 B CN107507052 B CN 107507052B
- Authority
- CN
- China
- Prior art keywords
- information
- neural network
- logistic regression
- preset
- network model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0611—Request for offers or quotes
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Technology Law (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明公开了一种报价信息获取方法及装置，属于商业信息技术领域。所述方法包括：获取用于描述报价信息的文字信息；通过预设人工智能算法对所述文字信息进行处理，获得报价信息；所述文字信息至少包括文本格式的文字信息和图片格式的文字信息，所述预设人工智能算法至少包括逻辑回归算法和神经网络算法。本发明通过预设人工智能算法对用于描述报价信息的文字信息进行数据处理，最终获得满足金融交易人员要求的统一格式报价信息，从而便于交易人员进行查看、快速反应以及利用该信息指导其交易行为，提高了本行业相关人员获取报价信息的效率，从而提高交易效率，可广泛应用于需要从众多报价信息中获取精准报价信息的金融等商业领域中。
The invention discloses a method and a device for obtaining quotation information, which belong to the technical field of commercial information. The method includes: acquiring text information for describing quotation information; processing the text information through a preset artificial intelligence algorithm to obtain quotation information; the text information at least includes text information in a text format and text information in a picture format , the preset artificial intelligence algorithm at least includes a logistic regression algorithm and a neural network algorithm. The present invention performs data processing on the text information used to describe the quotation information through preset artificial intelligence algorithms, and finally obtains quotation information in a unified format that meets the requirements of financial traders, thereby facilitating the traders to view, respond quickly and use the information to guide their transactions. It improves the efficiency of relevant personnel in the industry to obtain quotation information, thereby improving transaction efficiency, and can be widely used in financial and other commercial fields that need to obtain accurate quotation information from many quotation information.
Description
Technical Field
The invention relates to the technical field of business information, in particular to a method and a device for acquiring quotation information.
Background
The trade market of the financial institution is an OTC (Over-the-counter market) market, which is a market for inquiring quotes, and the general trade mode is that one party of the trade issues quote information to the whole market or fixed objects. In addition, the advertisement type text quotation information is mainly distributed and sent through a transparent RM, a Tencent QQ and a WeChat, so that the following problems are caused:
(1) the readability of text information is poor. The transaction quotation relates to a variety, a code, a deadline, a transaction direction, a transaction price, a transaction amount and other multiple elements, the elements are simply listed in a text form in the current quotation presentation mode, the content of the transaction quotation information is difficult to reflect visually, a transactor needs to spend a long time to read, understand and react, the transaction efficiency is obviously reduced, and the transaction opportunity is easy to miss;
(2) and (4) screen swiping phenomenon, which causes difficulty in tracking information. After the market is opened in the trade market, the price bees of the opponents are gathered, and the screen refreshing phenomenon is generated. When the local trader gets an indication of today's trading, it must roll back several or even hundreds of pages to see the opponent's advertisement he wants to trade and then trade with it. This creates an inconvenience in information lookup.
(3) New and old information is mixed. Financial markets are changing constantly, so that a quotation party withdraws quotations, revises the quotations and sends the quotations again for many times, the information which is refreshed is probably old, and the information which is seen by the quotation party is old and new and old because IM software cannot distinguish which information is covered by the old information, and the information which is viewed by the quotation party is necessary to be distinguished one by one with strength.
(4) The quotation information is scattered and difficult to compare quickly. Some mechanisms in the quotation have volume and price, and some mechanisms do not have price, so when a trader wants to find a better price, the trader needs to continuously turn pages back and forth to compare the prices.
Because of the above problems, after the quoted party obtains the information, the quoted party needs to digest the information one by one and then see whether the party has corresponding commodities or funds to trade with the corresponding commodities or funds. Because one organization faces hundreds of opponents, several advertisement type quotations appear every minute, the analysis of contents and the search of contents are both troubled, namely, the quotation information of each opponent is compared, and a proper counterparty is screened.
The investment of financial transaction, especially fixed income commodity transaction, is a competitive and highly time-efficient work, and traders need to pay close attention to market quotation information, find the optimal price among a plurality of quotation information according to own trading strategies, and reach a trading agreement with a trading opponent. However, based on the aforementioned limitations, the current quotation forms and tools are difficult to meet the business requirements, directly affect the trading efficiency, and further restrict the profitability of the trading organization. Therefore, it is desirable to provide a quote information acquiring technology that is simple, efficient and meets the current market demand.
Disclosure of Invention
In order to solve the problems in the prior art, embodiments of the present invention provide a method and an apparatus for acquiring quotation information. The technical scheme is as follows:
in a first aspect, a method for acquiring offer information is provided, where the method includes:
acquiring text information for describing quotation information; processing the text information through a preset artificial intelligence algorithm to obtain quotation information; the text information at least comprises text information in a text format and text information in a picture format, and the preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm.
With reference to the first aspect, in a first possible implementation manner, the obtaining text information describing offer information includes: and acquiring text information for describing the text format of the quotation information.
With reference to the first aspect, in a second possible implementation manner, the processing the text information through a preset artificial intelligence algorithm to obtain offer information includes: carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the target neural network model or the logistic regression model to obtain the quotation information.
With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner, the performing, by using a preset neural network model or a preset logistic regression model, training of advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening, and/or charting processing on the text information to obtain a target neural network model or logistic regression model includes: marking, defining or modeling the character information to generate a training set; and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
With reference to any one of the second and third possible implementation manners of the first aspect, in a fourth possible implementation manner, before the text information is processed through a preset neural network model or a preset logistic regression model, the method further includes: classifying the character information according to a preset weight rule to obtain a classification result; the classification result at least comprises training data meeting the first weight probability, test data meeting the second weight probability and evaluation data meeting the third weight probability.
With reference to the fourth possible implementation manner of the first aspect, the processing the text information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model includes: carrying out label definition or label setting on training data meeting the first weight probability obtained by classifying the character information to generate a training set; training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and after training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or logistic regression model, the method further comprises: the target neural network model or the logistic regression model is checked and/or evaluated through the checking data meeting the second weight probability and/or the evaluation data meeting the third weight probability, model parameters of the target neural network model or the logistic regression model are adjusted according to the checking result and/or the evaluation result, a final neural network model or the logistic regression model is obtained, and quotation information is obtained through the neural network model or the logistic regression model; and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the final neural network model or the logistic regression model to obtain the quotation information.
In a second aspect, there is provided an offer information acquiring apparatus, the apparatus including: the acquisition module is used for acquiring text information for describing quotation information; the data processing module is used for processing the text information through a preset artificial intelligence algorithm to obtain quotation information; the character information at least comprises Chinese characters, English words and numbers, and the preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm.
With reference to the second aspect, in a first possible implementation manner, the obtaining module is configured to obtain text information describing the offer information.
With reference to the second aspect, in a second possible implementation manner, the data processing module is configured to: carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the target neural network model or the logistic regression model to obtain the quotation information.
With reference to the second possible implementation manner of the second aspect, in a third possible implementation manner, the data processing module is configured to: performing label definition or label setting on the character information to generate a training set; and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
The technical scheme provided by the embodiment of the invention has the following beneficial effects:
according to the quotation information acquisition method and the device provided by the embodiment of the invention, the character information used for describing the quotation information is acquired, the character information is processed through the preset artificial intelligence algorithm to acquire the quotation information, and the quotation information in a unified format meeting the requirements of traders is acquired by adopting the artificial intelligence algorithm to perform data processing on the complicated quotation information sent by a plurality of traders on a communication tool, so that the traders can acquire effective quotation information quickly, the trading efficiency is improved by guiding subsequent trading activities, and the quotation acquisition method has the advantages of simple flow, strong applicability and the like, and can be widely applied to the business fields of finance and the like needing to acquire accurate quotation information from a plurality of quotation information.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart of a method for acquiring quotation information according to embodiment 1 of the present invention;
fig. 2 is a flowchart of a method for acquiring quotation information according to embodiment 2 of the present invention;
FIG. 3 is a diagram of a bid information advertisement as is common in the prior art;
fig. 4 is a schematic diagram of quotation information acquired by the quotation information acquisition method according to the embodiment of the present invention;
fig. 5 is a schematic structural diagram of a quotation information acquisition device provided in embodiment 3 of the present invention;
fig. 6 is a schematic structural diagram of a quote information obtaining apparatus provided in embodiment 4 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious 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.
It should be noted that the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
According to the quotation information acquisition method and device provided by the embodiment of the invention, the text information for describing the quotation information is subjected to data processing through the preset artificial intelligence algorithm, and finally the quotation information in a uniform format meeting the requirements of financial traders is obtained, so that the traders can conveniently check, quickly react and guide trading behaviors thereof by using the information, the efficiency of acquiring the quotation information by related personnel in the industry is improved, the trading efficiency is improved, and the method and device can be widely applied to the business fields such as finance and the like needing to acquire accurate quotation information from a plurality of quotation information.
The method and the apparatus for obtaining quotation information according to the embodiments of the present invention will be further described with reference to the following embodiments and accompanying drawings.
Example 1
Fig. 1 is a flowchart of a method for acquiring quotation information according to embodiment 1 of the present invention, and as shown in fig. 1, the method for acquiring quotation information according to the embodiment of the present invention includes the following steps:
101. and acquiring text information for describing the quotation information. The text information at least comprises text information in a text format and text information in a picture format.
Specifically, text information in a text format for describing the offer information is acquired.
102. And processing the character information through a preset artificial intelligence algorithm to obtain quotation information. The preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm. In addition, the preset artificial intelligence algorithm may also adopt any other artificial intelligence algorithm possible in the prior art, such as a natural language processing algorithm, and the invention is not limited thereto.
Specifically, this step can be performed in two ways:
the method comprises the steps that firstly, advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing are carried out on text format character information used for describing quotation information directly through a preset artificial algorithm or specifically through a preset neural network model, and quotation information is obtained;
and secondly, carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model so as to obtain quotation information by using the neural network model or the logistic regression model. Further, the above process is: marking, defining or modeling the character information to generate a training set; and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
And according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or the transaction element data processing including the transaction element classification processing, sorting, screening and/or charting processing by using a target neural network model or a logistic regression model to obtain the quotation information.
Further, before the text information is processed through a preset neural network model or a preset logistic regression model, the offer acquisition method provided by the embodiment of the invention further comprises the following steps:
and classifying the text information according to a preset weight rule to obtain a classification result, wherein the classification result at least comprises training data meeting the first weight probability, inspection data meeting the second weight probability and evaluation data meeting the third weight probability.
Further, the processing of the text information by the preset neural network model or the preset logistic regression model to obtain the target neural network model or the logistic regression model may be performed according to the following processes:
carrying out label definition or label setting on training data meeting the first weight probability obtained by classifying the character information to generate a training set;
training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and
after the training set is trained through a preset neural network model or a preset logistic regression model to obtain a target neural network model or logistic regression model, the method for obtaining the quoted price provided by the embodiment of the invention further comprises the following steps:
and checking and/or evaluating the target neural network model or the logistic regression model through the checking data meeting the second weight probability and/or the evaluation data meeting the third weight probability, adjusting model parameters of the target neural network model or the logistic regression model according to the checking result and/or the evaluation result, obtaining a final neural network model or the logistic regression model, and obtaining quotation information by using the neural network model or the logistic regression model.
The quotation information acquisition method provided by the embodiment of the invention obtains the quotation information by obtaining the text information for describing the quotation information and processing the text information through the preset artificial intelligence algorithm, and obtains the quotation information in a unified format meeting the requirements of traders by adopting the artificial intelligence algorithm to perform data processing on the complicated quotation information sent by a plurality of traders on a communication tool, thereby facilitating the traders to quickly obtain effective quotation information, guiding subsequent trading activities, improving trading efficiency, having the advantages of simple flow, strong applicability and the like, and being widely applied to the business fields of finance and the like needing to obtain accurate quotation information from a plurality of quotation information.
Example 2
Fig. 2 is a flowchart of a method for acquiring quotation information according to embodiment 2 of the present invention, and as shown in fig. 2, the method for acquiring quotation information according to the embodiment of the present invention includes the following steps:
201. and acquiring text information for describing the text format of the quotation information.
Specifically, the text information may include contents of chinese characters, numbers, english, and the like. Because the quoted information advertisement of financial transaction of goods like fixed income is generally issued by instant communication tools, the information in the quoted information advertisement exists in text format.
The offer information generally includes a financial commodity category name, a financial item term, a financial item price, a financial item quantity, a transaction request, etc., and the offer information includes offer information that differs according to the type of the offer information advertisement.
To further illustrate the effect of the embodiment of the present invention, fig. 3 is a schematic diagram of a bid information advertisement, which is common in the prior art, and as shown in fig. 3, the text information content includes both names or acronyms of different transaction categories and corresponding amount of data, transaction requirements, and the like. In addition, some quotation information advertisements also contain financial commodity recommendation terms for attracting other interests, and it is seen that the formats and habits of different quotation information advertisements issued by different counterparties are also different. In the face of such diverse and complex pricing information, the inefficiency is evident if it is merely through logging, screening, comparison or some other conventional means of processing. Since the text information containing the quotation information is mostly expressed in a text format, in addition to the text format, other formats such as pictures (such as screenshots and the like) may exist in some cases, but the unusual formats are few, so that the text information in the text format can be directly acquired, and the efficiency of the step of acquiring the quotation information is higher.
It should be noted that, step 201 obtains text information in a text format for describing the offer information, and the process may be implemented in other ways besides the ways described in the above steps, and the specific ways are not limited in the embodiment of the present invention.
202. Classifying the character information according to a preset weight rule to obtain a classification result; the classification result includes training data satisfying a first weight probability, test data satisfying a second weight probability, and evaluation data satisfying a third weight probability. Specifically, the classification method comprises the following steps:
after text information used for describing quotation information is obtained and before the quotation information is processed, the text information is classified according to a preset weight rule and divided into training data meeting a first weight probability, inspection data meeting a second weight probability and evaluation data meeting a third weight probability. The training data is used for carrying out subsequent learning or training on the quotation information data, the inspection data is used for inspecting the model obtained by the training, and the evaluation data is used for evaluating the effect of processing by using the obtained model data. The preset weighting rule can be set according to specific situations, and for example, can be set as follows: a first weighted probability of 60%, a second weighted probability of 20%, and a third weighted probability of 20%. Through the operation, data with different purposes can be prepared according to needs, early-stage preparation is made for a later-stage data training model, and classified training data, inspection data and evaluation data can better provide guarantee for training of the later-stage model when preset weight rules are better distributed or set.
It should be noted that, in the step 202, the text information is classified according to the preset weight rule to obtain the classification result, besides the manner described in the above step, the process may also be implemented in other manners, and the specific manner is not limited in the embodiment of the present invention.
203. And (4) performing label definition and modeling on the character information to generate a training set.
Specifically, the text information in the general quoted advertisement includes the financial commodity category, the specific different transaction price or amount, and the like. Certainly, a plurality of offer advertisements need to be distinguished according to the types of commodities, different text paragraphs belonging to different types need to be distinguished, at this time, names of commodities of different types need to be respectively labeled, one commodity type corresponds to one label, and variables corresponding to the labels are defined and modeled to generate a training set. Illustratively, the labeling process of the classification of the commodity categories may be referred to as "commodity separation labeling", and the process of commodity separation is as follows:
(1) defining variables: defining a plurality of variables y1, y2 and … yn, wherein n represents a variety expressed by characters in n rows;
(2) labeling is respectively carried out according to different situations: in a Training Set (Training Set), a line belonging to a line of characters of one variety is marked with Yes (y1 ═ 1), and the rest is No (y2 ═ y3 ═ … ═ yn ═ 0); in a Training Set, a row belonging to two lines of characters of one variety is marked with Yes (y2 ═ 1), and the rest is No (y1 ═ y3 ═ … ═ yn ═ 0); (4) and so on.
In addition, the annotation definition and modeling can also be performed on the transaction elements in the quotation information, and exemplarily, the following processes can be performed:
(1) tagging of transaction elements: labeling the digital content in the transaction element, wherein the label can be a period, a price and a quantity and is used for more conveniently identifying the learning result;
(2) defining attributes for performance indicators under different labels: defining attributes according to the financial commodity of the same industry, such as y1 being the same industry deposit, y2 being the agreement deposit, … y11 being the purchase, y12 being the sale, … y13 being the price, y14 being the quantity, y15 being the term …
(3) Defining the attribute value: and analyzing each variety advertisement, marking the corresponding attribute with 1 and marking the rest with 0.
Because the quotation text information comprises a plurality of transaction elements such as commodity types, time limit, price, quantity and the like and a plurality of commodity attribute indexes such as the same-industry deposit, protocol deposit, purchase and the like, different elements are identified, defined and modeled for the purpose of meeting machine learning language or serving for later data processing, and the training of a later model is facilitated.
It should be noted that, in step 203, the text information is labeled, defined and modeled to generate a training set, and besides the above-described manner in the step, the process may be implemented in other manners, and the specific manner is not limited in the embodiment of the present invention.
204. And carrying out advertisement category classification processing and transaction element data processing training including transaction element classification processing, sorting, screening and charting processing on the training set through a preset neural network model to obtain a target neural network model.
After step 203 is completed, a training set is obtained, and based on the training set, the following two aspects are trained through a preset neural network model:
(1) and (3) advertisement category classification processing: based on the commodity separation labeling, defining and modeling completed in step 203, training is performed according to preset training rules through a preset neural network model, and finally classification processing of quoted advertisement categories is achieved.
(2) Transaction element data processing including transaction element classification processing, sorting, screening and charting processing:
a. the transaction element classification processing refers to classifying various transaction elements in the offer information of the offer advertisement, and can be trained through a preset neural network model based on the transaction element labeling, definition and modeling completed in the step 203, so that the transaction element classification processing is finally realized;
b. the transaction element data including sequencing, screening and charting processing can be carried out on the transaction element information, so that the finally presented quotation information is more visual, diversified and clear, and better serves transaction personnel; in fact, the process may be performed not in the step of training the neural network, but after the training process is completed, and the invention is not limited thereto.
The steps a and b may be performed in a sequential manner, a parallel manner or an alternative manner, and the present invention is not limited thereto. In practice of training and acquiring the offer information, it is preferable that the advertisement classification process is performed on advertisement information of a plurality of different categories, and then the transaction element classification process of the commodity category is performed on each single category after the classification process. The operation makes the training process more orderly, and the quotation information data meeting the requirements can be more effectively obtained.
205. According to the new text information for describing the quotation information, the target neural network model is utilized to carry out advertisement category classification processing and transaction element data processing including transaction element classification processing, sorting, screening and charting processing, so as to obtain the quotation information.
After the target neural network model is obtained, when new text information used for describing quotation information is obtained, the target neural network model is utilized to carry out advertisement category classification processing and transaction element data processing including transaction element classification processing, sorting, screening and charting processing on the quotation information, and the quotation information is obtained. The specific processing steps are similar to the advertisement category classification processing and the transaction element data processing steps including the transaction element classification processing, sorting, screening and charting processing in the foregoing steps, and the related contents of the foregoing steps can be referred to, and are not described again here.
The quotation information of the quotation information is finally obtained, namely, the quotation information is obtained, the showing mode of the quotation information can be various according to different data processing processes, exemplarily, fig. 4 is a schematic diagram of the quotation information obtained by the quotation information obtaining method provided by the embodiment of the invention, as shown in fig. 4, effective information is obtained from complicated quotation advertisement text information through the series of processing processes of the quotation information obtaining method, and the effective information is converted into a standard expression form, so that the method is clear and greatly improves the efficiency of traders for obtaining the effective quotation information.
206. And checking and evaluating the target neural network model through the checking data meeting the second weight probability and the evaluation data meeting the third weight probability, adjusting model parameters of the target neural network model according to the checking result and the evaluation result, and obtaining a final neural network model so as to obtain new quotation information by using the neural network model. In order to obtain the optimal neural network model, i.e. the optimal model parameters, the target neural network model is checked and evaluated based on the check data meeting the second weight probability and the evaluation data meeting the third weight probability obtained in step 202, so as to obtain the final neural network model, and then new quotation information is obtained by using the neural network model.
207. According to the new character information for describing the quotation information, the final neural network model is utilized to carry out advertisement category classification processing and transaction element data processing including transaction element classification processing, sorting, screening and charting processing, so as to obtain the quotation information.
Specifically, after the final neural network model is obtained, when new text information for describing offer information is obtained, the neural network model is used for carrying out advertisement category classification processing and transaction element data processing including transaction element classification processing, sorting, screening and charting processing on the offer information to obtain the offer information. The specific processing steps are similar to the advertisement category classification processing and the transaction element data processing steps including the transaction element classification processing, sorting, screening and charting processing in the foregoing steps, and the related contents of the foregoing steps can be referred to, and are not described again here.
The quotation information of the quotation information is finally obtained through the further optimized neural network model, namely, the quotation information is obtained, the showing mode of the quotation information can be various according to different data processing processes, and the effective information is obtained from the complicated quotation advertisement character information through the series of processing processes of the quotation information obtaining method and is converted into a standard showing mode, so that the efficiency of obtaining the effective quotation information by traders is clear and greatly improved.
In addition, in order to obtain a better training neural network model, training set data for training can be added by various methods, for example, label identification or definition can be performed by multiple synonyms or synonyms, and by setting various label setting or definition rules, samples are artificially added so as to increase the learning effect, for example, "out" can be replaced by selling, and a new batch of samples is generated.
The quotation information acquisition method provided by the embodiment of the invention obtains the quotation information by obtaining the text information for describing the quotation information and processing the text information through the preset artificial intelligence algorithm, and obtains the quotation information in a unified format meeting the requirements of traders by adopting the artificial intelligence algorithm to perform data processing on the complicated quotation information sent by a plurality of traders on a communication tool, thereby facilitating the traders to quickly obtain effective quotation information, guiding subsequent trading activities, improving trading efficiency, having the advantages of simple flow, strong applicability and the like, and being widely applied to the business fields of finance and the like needing to obtain accurate quotation information from a plurality of quotation information.
Example 3
Fig. 5 is a quotation information acquiring apparatus provided in embodiment 3 of the present invention, and as shown in fig. 5, the quotation information acquiring apparatus 3 includes:
an obtaining module 31, configured to obtain text information for describing quotation information;
the data processing module 32 is used for processing the text information through a preset artificial intelligence algorithm to obtain quotation information;
the character information at least comprises Chinese characters, English words and numbers, and the preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm.
Specifically, the obtaining module 31 is configured to obtain text information for describing quotation information; the data processing module 32 is configured to: carrying out advertisement classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the target neural network model or the logistic regression model to obtain the quotation information. Further, the data processing module 32 is configured to: carrying out label definition or label setting on the character information to generate a training set; and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
Further, the data processing module 32 is further configured to: before the character information is processed through a preset neural network model or a preset logistic regression model, classifying the character information according to a preset weight rule to obtain a classification result; the classification result includes at least training data satisfying the first weight probability, test data satisfying the second weight probability, and evaluation data satisfying the third weight probability.
Further, the data processing module 32 is configured to perform label definition or label setting on training data meeting the first weight probability obtained by classifying the text information, and generate a training set; training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and after the training set is trained through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model, the target neural network model or the logistic regression model is checked and/or evaluated through the checking data meeting the second weight probability and/or the evaluation data meeting the third weight probability, model parameters of the target neural network model or the logistic regression model are adjusted according to the checking result and/or the evaluation result to obtain a final neural network model or the logistic regression model, and quotation information is obtained through the neural network model or the logistic regression model.
Example 4
Fig. 6 is a schematic structural diagram of a quotation information acquisition device provided in embodiment 4 of the present invention, and as shown in fig. 6, the quotation information acquisition device 4 provided in the embodiment of the present invention includes:
a memory 41 and a processor 42 connected to the memory 41, wherein the memory 41 is used for storing a set of program codes, and the processor 42 calls the program codes stored in the memory 41 to execute the following operations:
and acquiring text information for describing the quotation information. The text information at least comprises text information in a text format and text information in a picture format. Specifically, text information in a text format for describing the offer information is acquired.
And processing the text information through a preset artificial intelligence algorithm to obtain quotation information. The preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm. In addition, the preset artificial intelligence algorithm may also adopt any other artificial intelligence algorithm possible in the prior art, such as a natural language processing algorithm, and the invention is not limited thereto.
Specifically, this step can be performed in two ways:
the method comprises the steps that firstly, advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing are carried out on text format character information used for describing quotation information directly through a preset artificial algorithm or specifically through a preset neural network model, and quotation information is obtained;
and secondly, carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model. Further, the above process is: marking, defining or modeling the character information to generate a training set; and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
And according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or the transaction element data processing including the transaction element classification processing, sorting, screening and/or charting processing by using a target neural network model or a logistic regression model to obtain the quotation information.
Further, before the text information is processed through a preset neural network model or a preset logistic regression model, the offer acquisition method provided by the embodiment of the invention further comprises the following steps:
and classifying the text information according to a preset weight rule to obtain a classification result, wherein the classification result at least comprises training data meeting the first weight probability, inspection data meeting the second weight probability and evaluation data meeting the third weight probability.
Further, the processing of the text information by the preset neural network model or the preset logistic regression model to obtain the target neural network model or the logistic regression model may be performed according to the following processes:
carrying out label definition or label setting on training data meeting the first weight probability obtained by classifying the character information to generate a training set;
training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and
after the training set is trained through a preset neural network model or a preset logistic regression model to obtain a target neural network model or logistic regression model, the method for obtaining the quoted price provided by the embodiment of the invention further comprises the following steps:
the target neural network model or the logistic regression model is checked and/or evaluated through the check data meeting the second weight probability and/or the evaluation data meeting the third weight probability, model parameters of the target neural network model or the logistic regression model are adjusted according to the check result and/or the evaluation result, and a final neural network model or the logistic regression model is obtained so as to obtain quotation information by utilizing the neural network model or the logistic regression model;
and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the final neural network model or the logistic regression model to obtain the quotation information.
In summary, the quotation information obtaining method and device provided by the embodiment of the invention obtain the quotation information by obtaining the text information for describing the quotation information and processing the text information through the preset artificial intelligence algorithm, and obtain the quotation information in a uniform format meeting the requirements of traders by adopting the artificial intelligence algorithm to perform data processing on the complicated quotation information sent by a plurality of traders on a communication tool, so that the traders can quickly obtain effective quotation information, guide subsequent trading activities, improve trading efficiency, have the advantages of simple flow, strong applicability and the like, and can be widely applied to the business fields such as finance and the like needing to obtain accurate quotation information from a plurality of quotation information.
It should be noted that: in the quotation information acquiring device provided in the above embodiment, when the quotation information acquiring service is performed, only the division of the above functional modules is taken as an example, and in practical application, the function distribution may be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules, so as to complete all or part of the above described functions. In addition, the quotation information acquisition device and the quotation information acquisition method provided by the above embodiments belong to the same concept, and the specific implementation process is detailed in the method embodiments and will not be described herein again.
All the above-mentioned optional technical solutions can be combined arbitrarily to form the optional embodiments of the present invention, and are not described herein again.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (6)
1. A method for acquiring quotation information is characterized by comprising the following steps:
acquiring text information for describing quotation information;
processing the text information through a preset artificial intelligence algorithm to obtain quotation information;
the text information at least comprises text information in a text format and text information in a picture format, and the preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm;
the processing of the text information through a preset artificial intelligence algorithm to obtain the quotation information comprises the following steps:
carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model;
according to the new text information used for describing quotation information, the target neural network model or the logistic regression model is used for carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing, so as to obtain quotation information;
before processing the text information through a preset neural network model or a preset logistic regression model, the method further comprises:
classifying the character information according to a preset weight rule to obtain a classification result;
the classification result at least comprises training data meeting a first weight probability, test data meeting a second weight probability and evaluation data meeting a third weight probability;
the processing the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model comprises:
carrying out label definition or label setting on training data meeting the first weight probability obtained by classifying the character information to generate a training set;
training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and
after the training set is trained through a preset neural network model or a preset logistic regression model to obtain a target neural network model or logistic regression model, the method further includes:
the target neural network model or the logistic regression model is checked and/or evaluated through the checking data meeting the second weight probability and/or the evaluation data meeting the third weight probability, model parameters of the target neural network model or the logistic regression model are adjusted according to the checking result and/or the evaluation result, a final neural network model or the logistic regression model is obtained, and quotation information is obtained through the neural network model or the logistic regression model;
and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the final neural network model or the logistic regression model to obtain the quotation information.
2. The method of claim 1, wherein obtaining text information describing an offer message comprises:
and acquiring text information for describing the text format of the quotation information.
3. The method according to claim 1, wherein the training of the text information by performing an advertisement classification process and/or a transaction element data process including a transaction element classification process, a ranking process, a screening process and/or a charting process through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model comprises:
marking, defining or modeling the character information to generate a training set;
and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
4. An offer information acquisition apparatus, characterized in that the apparatus comprises:
the acquisition module is used for acquiring text information for describing quotation information;
the data processing module is used for processing the text information through a preset artificial intelligence algorithm to obtain quotation information, wherein the text information at least comprises Chinese characters, English characters and numbers, and the preset artificial intelligence algorithm at least comprises a logistic regression algorithm and a neural network algorithm;
the data processing module is used for:
carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; according to the new text information used for describing quotation information, the target neural network model or the logistic regression model is used for carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing, so as to obtain quotation information;
before the character information is processed through a preset neural network model or a preset logistic regression model, classifying the character information according to a preset weight rule to obtain a classification result; the classification result at least comprises training data meeting a first weight probability, test data meeting a second weight probability and evaluation data meeting a third weight probability; the processing the character information through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model comprises: carrying out label definition or label setting on training data meeting the first weight probability obtained by classifying the character information to generate a training set; training the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model; and
after a target neural network model or a logistic regression model is obtained by training the training set through a preset neural network model or a preset logistic regression model, the target neural network model or the logistic regression model is checked and/or evaluated through the checking data meeting the second weight probability and/or the evaluation data meeting the third weight probability, model parameters of the target neural network model or the logistic regression model are adjusted according to the checking result and/or the evaluation result, a final neural network model or the logistic regression model is obtained, and quotation information is obtained through the neural network model or the logistic regression model; and according to the new text information for describing the quotation information, carrying out the advertisement category classification processing and/or transaction element data processing including transaction element classification processing, sorting, screening and/or charting processing by using the final neural network model or the logistic regression model to obtain the quotation information.
5. The apparatus of claim 4, wherein the obtaining module is configured to obtain text information describing the offer information.
6. The apparatus of claim 4, wherein the data processing module is configured to:
performing label definition or label setting on the character information to generate a training set;
and carrying out advertisement category classification processing and/or transaction element data processing training including transaction element classification processing, sorting, screening and/or charting processing on the training set through a preset neural network model or a preset logistic regression model to obtain a target neural network model or a logistic regression model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710581774.1A CN107507052B (en) | 2017-07-17 | 2017-07-17 | A method and device for obtaining quotation information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710581774.1A CN107507052B (en) | 2017-07-17 | 2017-07-17 | A method and device for obtaining quotation information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107507052A CN107507052A (en) | 2017-12-22 |
| CN107507052B true CN107507052B (en) | 2021-04-09 |
Family
ID=60679946
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710581774.1A Active CN107507052B (en) | 2017-07-17 | 2017-07-17 | A method and device for obtaining quotation information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107507052B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889289B (en) * | 2018-08-17 | 2022-05-06 | 北大方正集团有限公司 | Information accuracy evaluation method, device, equipment and computer readable storage medium |
| CN111199409A (en) * | 2018-11-16 | 2020-05-26 | 浙江舜宇智能光学技术有限公司 | Cost control method and system for specific product and electronic device |
| CN116188091A (en) * | 2023-05-04 | 2023-05-30 | 品茗科技股份有限公司 | Method, device, equipment and medium for automatic matching unit price reference of cost list |
| CN116433383B (en) * | 2023-06-12 | 2023-10-31 | 宁波森浦融讯科技有限公司 | Data processing method, device, electronic equipment and computer readable storage medium |
| CN117764627B (en) * | 2023-12-27 | 2025-04-18 | 北京新联财通咨询有限公司 | Method and device for determining activity communication effect, storage medium, and terminal |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105117468A (en) * | 2015-08-28 | 2015-12-02 | 广州酷狗计算机科技有限公司 | Network data processing method and apparatus |
| CN105955952A (en) * | 2016-05-03 | 2016-09-21 | 成都数联铭品科技有限公司 | Information extraction method based on bidirectional recurrent neural network |
| CN106021442A (en) * | 2016-05-16 | 2016-10-12 | 江苏大学 | Network news outline extraction method |
-
2017
- 2017-07-17 CN CN201710581774.1A patent/CN107507052B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105117468A (en) * | 2015-08-28 | 2015-12-02 | 广州酷狗计算机科技有限公司 | Network data processing method and apparatus |
| CN105955952A (en) * | 2016-05-03 | 2016-09-21 | 成都数联铭品科技有限公司 | Information extraction method based on bidirectional recurrent neural network |
| CN106021442A (en) * | 2016-05-16 | 2016-10-12 | 江苏大学 | Network news outline extraction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107507052A (en) | 2017-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107507052B (en) | A method and device for obtaining quotation information | |
| US20210350426A1 (en) | Architecture for data processing and user experience to provide decision support | |
| US11823026B2 (en) | Artificial intelligence system employing graph convolutional networks for analyzing multi-entity-type multi-relational data | |
| Lorca et al. | Impact of e-commerce sales on profitability and revenue. The case of the manufacturing industry | |
| EP1960961A1 (en) | Data independent relevance evaluation utilizing cognitive concept relationship | |
| Cho et al. | Non-fungible token transactions: Data and challenges | |
| Hossain et al. | Evaluating the utilization of technological factors to promote e-commerce adoption in small and medium enterprises | |
| Ratna Komala et al. | Improving the quality of financial statements and the survival of MSMEs through digital economy: The case of Indonesia and Malaysia | |
| Lee et al. | A cause of oversupply and failure in the shipping market: measuring herding behavior effects | |
| CN115456745A (en) | Small and micro enterprise portrait construction method and device | |
| Azcoitia et al. | Understanding the price of data in commercial data marketplaces | |
| CN112950017A (en) | Contract risk identification method and device and electronic equipment | |
| CN117876083A (en) | Intelligent analysis method and device for business opportunity, computer equipment and storage medium | |
| Kumar et al. | Beyond Likes and Shares: Unveiling the Sequential Mediation of Brand Equity, Loyalty, Image, and Awareness in Social Media Marketing's Influence on Repurchase Intentions for High-Tech Products. | |
| US20230100685A1 (en) | Financial product information collecting platform system, financial product information collecting method, and computer program for the same | |
| Suder et al. | Examining moderating effect of environmental dynamism and hostility on entrepreneurial orientation/performance relationship | |
| CN119807331A (en) | A method and system for generating customer service information based on AI | |
| Cao et al. | NFTracer: tracing NFT impact dynamics in transaction-flow substitutive systems with visual analytics | |
| Khan et al. | Elevating Consumer Purchase Intentions in Pakistan: The Power of Digital Marketing | |
| Pakaya et al. | Decisions at Marketplace Shopee on Islamic Business Ethic Perspective: Determination of Online Consumer Reviews and Rating | |
| Gautschi et al. | A methodology for specification and aggregation in product concept testing | |
| Liao et al. | Mining customer knowledge for channel and product segmentation | |
| KR20240084602A (en) | Professional report writing system and report providing method using the same | |
| Hao et al. | Do platform recommendations in the fund market work? Evidence from a quasi-experimental study | |
| KR101596319B1 (en) | Device and method for providing FTA business model |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20250423 Address after: Room 1902, Building 1, Moon Bay International Business Center, No. 9 Cuiwei Street, Suzhou Industrial Park, Jiangsu Province, China 215123 Patentee after: KMERIT (SUZHOU) INFORMATION SCIENCE & TECHNOLOGY Co.,Ltd. Country or region after: China Address before: Room 4b05, Fuhua science and technology building, 199 Chongwen Road, Suzhou Industrial Park, Suzhou, Jiangsu 215028 Patentee before: SUZHOU KAILIAN INFORMATION TECHNOLOGY CO.,LTD. Country or region before: China |
|
| TR01 | Transfer of patent right |