CN111738856A - Stock public opinion investment decision analysis method and device - Google Patents

Stock public opinion investment decision analysis method and device Download PDF

Info

Publication number
CN111738856A
CN111738856A CN202010587191.1A CN202010587191A CN111738856A CN 111738856 A CN111738856 A CN 111738856A CN 202010587191 A CN202010587191 A CN 202010587191A CN 111738856 A CN111738856 A CN 111738856A
Authority
CN
China
Prior art keywords
stock
public opinion
information
investment decision
profitability
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.)
Pending
Application number
CN202010587191.1A
Other languages
Chinese (zh)
Inventor
唐阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan Changhong Electric Co Ltd
Original Assignee
Sichuan Changhong Electric Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sichuan Changhong Electric Co Ltd filed Critical Sichuan Changhong Electric Co Ltd
Priority to CN202010587191.1A priority Critical patent/CN111738856A/en
Publication of CN111738856A publication Critical patent/CN111738856A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • 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
    • G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
    • 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
    • G06F16/35—Clustering; Classification
    • G06F16/353—Clustering; Classification into predefined classes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/951—Indexing; Web crawling techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • 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
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Human Resources & Organizations (AREA)
  • General Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Educational Administration (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Evolutionary Computation (AREA)
  • Pure & Applied Mathematics (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Game Theory and Decision Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Mathematical Physics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Technology Law (AREA)
  • Algebra (AREA)

Abstract

The invention discloses a stock public opinion investment decision analysis method, and belongs to the technical field of decision analysis. The method comprises the following steps: determining a candidate stock pool to be evaluated; extracting historical profitability information of the candidate stock pool and historical assessment factor set information containing public opinion information; inputting the historical profitability information and the historical assessment factor set information into a public opinion investment decision assessment model; evaluating an expected excess earnings rate of the candidate stock pool using the stock public opinion investment decision evaluation model; and obtaining the scores of the candidate stock pools based on the expected excess earnings, and selecting a certain number of stocks as a target stock pool based on the scores. The public opinion model factor of the invention intensively introduces unique interactive public opinion data, thereby obtaining public opinion text emotional attitude polarity indexes and public opinion heat statistical variables and finally forming a brand new public opinion training model system; helping it to improve investment analysis efficiency, accuracy and income.

Description

Stock public opinion investment decision analysis method and device
Technical Field
The invention belongs to the field of decision analysis, and particularly relates to a stock public opinion investment decision analysis method and device.
Background
With the rapid development of internet technology, the world has rapidly advanced into the data age, and the data volume is increasing with the increasing data acquisition and data processing capability. The current market scoring and trend prediction is commonly used by stock historical base class data, such as market data, trading data, and financial data.
However, it has been demonstrated that the value of these data has been essentially mined out, and most of the factors generated based on these data have become essentially ineffective as they become widely used in the industry, gradually losing the value of the investor's decision. And the stock investment decision recommendation mode based on the static historical data has strong hysteresis, and more buying and selling decisions are still dependent on the personal experience of investors.
China has a large base number of scattered households, is not as professional as financial institutions, and lacks relevant technical means for assisting decision making of the scattered households. In addition, when buying and selling stocks, the scattered user can see the K-line graph fluctuation and the so-called lane information by means of own subjective intention, forecast the future price of the stocks by analyzing the cold K-line data, and the accuracy rate is low, so that the investment is at great risk. The public opinion factor of a stock can dynamically express the market potential of the stock to a certain extent, and when the public opinion factor of a stock varies (great interest/profit public opinion), the corresponding investment strategy also needs to be dynamically adjusted in time.
The invention is based on the interactive public opinion data obtained by the internet crawler, the dynamic public opinion factor obtained by certain processing and processing is input into the preset model, the stock score is calculated, the influence of the public opinion factor on the future income of the stock is visually displayed, the high-ranking stock is evaluated as an alternative target investment target, the dynamic investment opportunity is timely and conveniently captured, the investment risk is reduced, the analysis efficiency, the investment efficiency and the income are improved, and the invention has higher actual combat value.
Disclosure of Invention
The invention provides a stock public opinion investment decision analysis method, and relates to the field of financial investment. The method comprises the steps of extracting public opinion text contents by acquiring stock dynamic public opinion information, carrying out word segmentation and emotional part-of-speech tagging on corresponding text public opinions of all stocks in Shanghai and Shenshen cities, and calculating the emotional attitude polarity index of each public opinion text. And (4) counting to obtain a series of statistical variables representing the public opinion popularity of each stock for a period of time. And inputting the public opinion emotional attitude polarity index and the public opinion popularity statistical variable into a preset public opinion training model to obtain the expected excess earning rate and the stock score of each stock, and selecting a certain number of stocks as a target stock pool based on the grade. Meanwhile, the daily stock public opinion index is calculated according to the public opinion popularity statistical variable, and an investor can make prejudgment and continuous tracking on the future price of the stock according to the daily stock public opinion index, so that the efficiency and the accuracy of investment decision of the investor are improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
a stock public opinion investment decision analysis method, the method comprising:
determining a candidate stock pool to be evaluated;
extracting historical profitability information of the candidate stock pool and historical assessment factor set information containing public opinion information;
inputting the historical profitability information and the historical assessment factor set information into a public opinion investment decision assessment model;
evaluating an expected excess earnings rate of the candidate stock pool using the stock public opinion investment decision evaluation model;
and obtaining the scores of the candidate stock pools based on the expected excess earnings, and selecting a certain number of stocks as a target stock pool based on the scores.
Further, the extracting historical profitability information of the candidate stock pool and historical assessment factor set information including public opinion information, wherein:
the historical earning rate information is obtained by calculation according to the closing price of the stock;
the historical assessment factor set information comprises historical public opinion indexes and traditional three factors.
Further, the historical public opinion index comprises:
the text emotion polarity of the question content, the text emotion polarity of the reply content, the stock question amount, the stock reply amount, the reply timeliness, the effective text length of the question content, the effective text length of the reply content and whether to reply;
the conventional three factors include:
market portfolio factor, market value factor, account market value ratio factor.
Further, the stock public opinion investment decision evaluation model is obtained by training a public opinion investment decision evaluation model obtained based on the improvement of a traditional three-factor model; the public opinion investment decision evaluation model comprises the following steps:
Rit-Rft＝ai+bi(Rmt-Rft)+siSMBt+hiHMLt+yiSentAt+i
Figure BDA0002555087730000031
wherein, Xi＝θi0+θi1Xi1+θi2Xi2，
Figure BDA0002555087730000032
RitProfitability information for the stock object; rftRisk-free profitability, such as profitability information of national debt; a isiFor excess profitability, which is the additional profitability beyond the profitability of the portfolio or the equity in the big volatility, α represents the excess profitability, the higher the better, the more excellent our portfolio is, RmtThe market investment portfolio profitability information is replaced by proper large-scale indexes; SMB (System management bus)tMarket value portfolio profitability information; HMLtThe information is the market value to yield rate information;iis an error term; y isiSentAtScoring public sentiment factors, wherein XiAnd YiRespectively public sentiment emotional attitude polarity index variable and public sentiment popularity statistical index variable,θi1and
Figure BDA0002555087730000033
are autoregressive parameters.
Further, the public opinion investment decision evaluation model training method comprises the following steps:
collecting a stock factor set of a stock sample in a preset period;
the stock factor set comprises the profitability information of a first stage, the public sentiment information of the first stage, the public sentiment information of a second stage and the profitability information of the second stage;
and taking the profitability information of the first stage and the public opinion information of the first stage as input data, performing model training by adopting the public opinion investment decision evaluation algorithm, and performing model evaluation by taking the profitability information of the second stage and the public opinion information of the second stage as input data to obtain the stock public opinion investment decision evaluation model.
Further, the evaluating the expected excess earnings of the candidate stock pool by using the stock public opinion investment decision evaluation model comprises:
and obtaining the excess income rate information of the candidate stock pool by using the historical evaluation factor set information of the candidate stock pool and combining regression parameters of all factors obtained in the trained stock public opinion investment decision evaluation model, wherein the excess income rate information can be used for representing future expected price rise and fall of the candidate stock pool, and based on the excess income rate information, a future expected income rate curve of the candidate stock pool can be obtained, and meanwhile, the public opinion index of the candidate stock pool can be obtained based on the public opinion factor score.
The invention also provides a stock public opinion investment decision analysis device, comprising:
the data acquisition module is used for acquiring public sentiment data of stocks, closing price data of individual stocks and traditional three-factor data;
the data processing module is used for preparing input data before model evaluation and comprises a word segmentation module, an emotion analysis module, a data statistics calculation module and a data alignment module;
the public opinion scoring module is used for generating a final public opinion score from the public opinion indexes;
the stock evaluation module is used for inputting the public opinion score and the traditional three factors into a preset stock public opinion investment decision evaluation model to obtain the excess earning rate information of the candidate object;
and the public opinion index visualization module is used for converting the excess income rate information and the public opinion factor score into the excess income rate information and the public opinion index of a candidate object, and displaying the excess income rate information and the public opinion index of the stock to a stock user so that the stock user can analyze the stock according to the excess income rate information and the public opinion index of the stock.
Further, the apparatus further comprises:
the word segmentation module is used for segmenting the stock public opinion information to obtain initial word information;
the emotion analysis module is used for carrying out positive and negative emotion part-of-speech tagging on the initial word information to obtain tagged word information, and calculating and outputting a text emotion polarity index;
and the data statistics and calculation module is used for performing statistics and calculation on various index variables of the input model, including the securities questioning amount, the securities reply amount, the reply timeliness, the effective text length of questioning content, the effective text length of reply content, whether reply is performed, market asset combination, market value factor, account and market value ratio factor, individual stock yield and market risk-free yield.
Optionally, the data alignment module is configured to align time periods of training and evaluation of the input indexes of the various models.
Compared with the prior art, the invention has the beneficial effects that:
the public opinion model factor of the invention intensively introduces unique interactive public opinion data, thereby obtaining public opinion text emotional attitude polarity indexes and public opinion heat statistical variables and finally forming a brand new public opinion training model system. Meanwhile, based on the analysis result, a stock public opinion index is provided, investors can intuitively carry out price prejudgment and continuous tracking on targets, a novel decision-making assisting tool is provided for investment decision-making of the investors, and the efficiency, the accuracy and the income of investment analysis are improved.
Drawings
Fig. 1 is a flow chart of a stock public opinion investment decision analysis method provided in an embodiment.
Fig. 2 is a flow chart of a stock public opinion text word segmentation logic wireframe provided in an embodiment.
Fig. 3 is a flow chart of the stock public opinion text sentiment analysis logic provided in the embodiment.
Fig. 4 is a block diagram of a stock public opinion investment decision analysis device provided in an embodiment.
Detailed Description
The present invention will be further described with reference to the following examples, which are intended to illustrate only some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, other embodiments used by those skilled in the art without any creative effort belong to the protection scope of the present invention.
Example 1:
the invention provides a stock public opinion investment decision analysis method, fig. 1 is a flow chart of the stock public opinion investment decision analysis method according to the embodiment of the invention, as shown in fig. 1, the flow comprises the following steps:
step S100, determining a candidate stock pool to be evaluated. In this example, the candidate stock pool is the stock that the investor wants to evaluate the future price trend, and can be determined by clicking on the stock id.
And step S101, extracting historical profitability information of the stock pool and historical assessment factor set information containing public opinion information. The historical yield information of the example is obtained by adopting a calculation formula of the fluctuation range of market mainstream, and comprises the following steps:
Figure BDA0002555087730000061
wherein R isFor the profitability of the stock at time t (current period), PtFor closing price of stock at time t (current period), Pt-1The closing price of the stock at the time t-1 (the last period).
The historical assessment factor set information in this example is relevant information that affects the profit of the stock, including: the traditional three factors relate to market asset combination factor, market value factor and account market value ratio factor. The market asset combination factor is replaced by a proof index, the market value factor and the account-market ratio factor can be directly obtained from third-party financial data providers such as wind or tune and can also be calculated by self, the market value factor is the number of stocks per stock price, and the account-market ratio is the reciprocal of the net market rate.
The cross-section stock base evaluation factor data finally formed in step S101 is shown in table 1 below:
Figure BDA0002555087730000062
public opinion factor relates to the emotional polarity of the questioning content text, the emotional polarity of the reply content text, the stock questioning amount, the stock reply amount, the reply timeliness, the effective text length of the questioning content and the effective text length of the reply content. The public opinion information can be acquired from two websites of certified e interaction (with the website being http:// sns. sselnfo. com /) and Shenzhen interaction (with the website being http:// irm. cninfo. com. cn/szse/index. html.) provided by Shanghai and Shenshen exchanges respectively. The public opinion text sentiment polarity calculation formula is as follows:
Figure BDA0002555087730000071
the reply timeliness of the public sentiment question is equal to the difference value of the reply time of the corresponding question and the question time, the effective text length of the public sentiment content is equal to the text length after the beginning and stopping words (including but not limited to the tone words, prepositions, conjunctions and the like), and optionally, a general stopping word library can be downloaded from a network. The cross-section stock public opinion factor data finally formed in step S101 is shown in table 2 below:
Figure BDA0002555087730000072
and step S102, training the historical profitability information and the assessment factor set information to obtain a stock public opinion investment decision assessment model.
In this example, the training step of the stock public opinion investment decision evaluation model comprises: collecting a stock factor set of a stock sample in a preset period, wherein the stock factor set comprises return rate information of a first stage, public opinion information of the first stage, and public opinion information of a second stage and return rate information of the second stage;
and taking the profitability information of the first stage and the public opinion information of the first stage as input data, performing model training by adopting the public opinion investment decision evaluation algorithm, and performing model evaluation by taking the profitability information of the second stage and the public opinion information of the second stage as input data to obtain the stock public opinion investment decision evaluation model.
Preferably, the public opinion score model adopts a multi-factor analysis model MPMD model based on a logistic regression model, and the model has higher model accuracy and analysis efficiency, and the formula is as follows:
Figure BDA0002555087730000081
the stock alpha expected excess yield model adopts an improved Fama three-factor model, and the formula is as follows: rit-Rft＝ai+bi(Rmt-Rft)+siSMBt+hiHMLt+yiSentAt+i
Specifically, the following description will be made by taking the same-flower rank stocks shown in tables 1 and 2 as examples:
the same flower is the first stock and its public opinion score
Figure BDA0002555087730000082
X1＝θ10+θ11X11+θ12X12.
Figure BDA0002555087730000083
X11-X12Respectively representing public opinion emotional attitude polarity index variables corresponding to the same flower sequence in a preset time period, including the emotional polarity of the question public opinion text and the emotional polarity of the reply public opinion text, which can be calculated by the calculation formula of the emotional polarity of the public opinion text
Figure BDA0002555087730000084
And calculating. Y is11-Y15And respectively representing public opinion popularity heat degree statistical variables in a preset time period, wherein the variables comprise question amount, reply timeliness, question text length and reply text length. In this example, yiSentAtThe value interval of (2) is (0). Autoregressive parameters θ and
Figure BDA0002555087730000085
can be trained by a general iterative algorithm.
Optionally, the number of positive and negative emotion words in the public opinion Text emotion polarity formula can be obtained by a word segmentation tool Jieba and an emotion mark analysis tool Text Blob analysis. The word segmentation logic is simple, the text to be analyzed is subjected to text processing according to the constructed emotion dictionary to extract emotion words, the emotion tendency of the text is calculated, and the final classification effect depends on the completeness of the emotion dictionary, as shown in fig. 2.
The emotion analysis is essentially a problem of two-classification, and can be recognized by adopting a machine learning method, emotion words in a text are selected as feature words, the text is matrixed, classification is carried out by utilizing classifier algorithms such as logistic regression, naive Bayes, support vector machines and the like, and the final classification effect depends on the selection of a training text and correct emotion marking, as shown in FIG. 3.
Whether the emotion polarity of a public opinion text is positive or negative is finally obtained according to the emotion marking analysis tool, and accordingly, the number of positive and negative emotion words of the same-flower-sequence stock in a preset time period can be obtained, and the emotion polarity of a question or a reply public opinion text is obtained respectively.
After obtaining the public opinion score, adding other index items related in the model, and evaluating the model R according to the stock public opinion investment decisionit-Rft＝αi+bi(Rmt-Rft)+siSMBt+hiHMLt+yiSentAt+iAnd the OLS algorithm is adopted to train to obtain model parameters.
And step S103, evaluating the expected excess earnings of the candidate stock pool by using the stock public opinion investment decision evaluation model. Taking the same-flower-rank stock as an example, based on the stock public opinion investment decision evaluation model obtained by training in step S102 of this example, the public opinion score factor, the asset combination factor, the market value factor, and the account-market value ratio factor of the next period of the same-flower-rank stock are input, so as to predict and obtain the expected excess earning rate and the public opinion index curve trend map of the next period of the same-flower-rank stock.
Alternatively, other stock evaluation methods are similar to those described above, and finally, a stock list sorted according to expected excess profitability, which is the most important item to be focused on by investors, and a public sentiment index curve trend chart of the corresponding stocks are obtained.
From the above embodiments, those skilled in the art can clearly know and implement the method according to the above embodiments by means of some software work plus necessary hardware platform.
An embodiment of the present invention further provides a public opinion investment decision analysis apparatus, and fig. 2 is a frame diagram of a stock public opinion investment decision analysis apparatus according to an embodiment of the present invention, as shown in fig. 4, the apparatus includes:
the data acquisition module is used for acquiring public sentiment data of stocks, closing price data of individual stocks and traditional three-factor data;
and the data processing module is used for preparing input data before model evaluation and comprises a word segmentation module, an emotion analysis module, a data statistics calculation module and a data alignment module.
Optionally, the word segmentation module is configured to segment the stock public opinion information to obtain initial word information; and the emotion analysis module is used for carrying out positive and negative emotion part of speech tagging on the initial word information to obtain tagged word information, and calculating and outputting a text emotion polarity index.
Optionally, the data statistics calculation module is configured to perform statistics calculation on various index variables of the input model, including a security question amount, a security reply amount, a reply timeliness, a valid text length of a question content, a valid text length of a reply content, whether to reply, a market asset combination, a market value factor, an account-market value ratio factor, an individual stock yield, and a market risk-free yield.
Optionally, the data alignment module is configured to align time periods of training and evaluation of the input indexes of the various models.
And the public opinion scoring module is used for generating a final public opinion score from the public opinion indexes.
And the stock evaluation module is used for inputting the public opinion score and the traditional three factors into a preset stock public opinion investment decision evaluation model to obtain the excess earning rate information of the candidate object.
And the public opinion index visualization module is used for converting the excess income rate information and the public opinion factor score into the excess income rate information and the public opinion index of a candidate object, and displaying the excess income rate information and the public opinion index of the stock to a stock user so that the stock user can analyze the stock according to the excess income rate information and the public opinion index of the stock.
It should be noted that the above modules can be directly implemented by software programming.
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 and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (8)

1. A stock public opinion investment decision analysis method is characterized by comprising the following steps:
determining a candidate stock pool to be evaluated;
extracting historical profitability information of the candidate stock pool and historical assessment factor set information containing public opinion information;
inputting the historical profitability information and the historical assessment factor set information into a public opinion investment decision assessment model;
evaluating an expected excess earnings rate of the candidate stock pool using the stock public opinion investment decision evaluation model;
and obtaining the scores of the candidate stock pools based on the expected excess earnings, and selecting a certain number of stocks as a target stock pool based on the scores.
2. The method as claimed in claim 1, wherein the historical profitability information of the candidate stock pool and the historical assessment factor set information including public opinion information are extracted, wherein:
the historical earning rate information is obtained by calculation according to the closing price of the stock;
the historical assessment factor set information comprises historical public opinion indexes and traditional three factors.
3. The method as claimed in claim 2, wherein the index of historical public sentiment comprises:
the text emotion polarity of the question content, the text emotion polarity of the reply content, the stock question amount, the stock reply amount, the reply timeliness, the effective text length of the question content, the effective text length of the reply content and whether to reply;
the conventional three factors include:
market portfolio factor, market value factor, account market value ratio factor.
4. The method for analyzing the stock public opinion investment decision as claimed in claim 1, wherein the stock public opinion investment decision evaluation model is trained by a public opinion investment decision evaluation model based on a traditional three-factor model improvement; the public opinion investment decision evaluation model comprises the following steps:
Rit-Rft＝ai+bi(Rmt-Rft)+siSMBt+hiHMLt+yiSentAt+i
Figure FDA0002555087720000011
wherein, Xi＝θi0+θi1Xi1+θi2Xi2，
Figure FDA0002555087720000012
RitProfitability information for the stock object; rftRisk-free profitability, such as profitability information of national debt; a isiFor excess profitability, which is the additional profitability beyond the profitability of the portfolio or the equity in the big volatility, α represents the excess profitability, the higher the better, the more excellent our portfolio is, RmtThe market investment portfolio profitability information is replaced by proper large-scale indexes; SMB (System management bus)tMarket value portfolio profitability information; HMLtThe information is the market value to yield rate information;iis an error term; y isiSentAtScoring public sentiment factors, wherein XiAnd YiRespectively public sentiment emotional attitude polarity index variable and public sentiment popularity statistical index variable thetai1And
Figure FDA0002555087720000013
are autoregressive parameters.
5. The method for analyzing a stock public opinion investment decision, according to claim 4, wherein the training method of the public opinion investment decision evaluation model comprises:
collecting a stock factor set of a stock sample in a preset period;
the stock factor set comprises the profitability information of a first stage, the public sentiment information of the first stage, the public sentiment information of a second stage and the profitability information of the second stage;
and taking the profitability information of the first stage and the public opinion information of the first stage as input data, performing model training by adopting the public opinion investment decision evaluation algorithm, and performing model evaluation by taking the profitability information of the second stage and the public opinion information of the second stage as input data to obtain the stock public opinion investment decision evaluation model.
6. The method as claimed in claim 4, wherein the estimating expected excess earnings of the candidate stock pool using the stock public opinion investment decision evaluation model comprises:
and obtaining the excess income rate information of the candidate stock pool by using the historical evaluation factor set information of the candidate stock pool and combining regression parameters of all factors obtained in the trained stock public opinion investment decision evaluation model, wherein the excess income rate information can be used for representing future expected price rise and fall of the candidate stock pool, and based on the excess income rate information, a future expected income rate curve of the candidate stock pool can be obtained, and meanwhile, the public opinion index of the candidate stock pool can be obtained based on the public opinion factor score.
7. A stock public opinion investment decision analysis device is characterized by comprising:
the data acquisition module is used for acquiring public sentiment data of stocks, closing price data of individual stocks and traditional three-factor data;
the data processing module is used for preparing input data before model evaluation and comprises a word segmentation module, an emotion analysis module, a data statistics calculation module and a data alignment module;
the public opinion scoring module is used for generating a final public opinion score from the public opinion indexes;
the stock evaluation module is used for inputting the public opinion score and the traditional three factors into a preset stock public opinion investment decision evaluation model to obtain the excess earning rate information of the candidate object;
and the public opinion index visualization module is used for converting the excess income rate information and the public opinion factor score into the excess income rate information and the public opinion index of a candidate object, and displaying the excess income rate information and the public opinion index of the stock to a stock user so that the stock user can analyze the stock according to the excess income rate information and the public opinion index of the stock.
8. The apparatus of claim 7, wherein the apparatus further comprises:
the word segmentation module is used for segmenting the stock public opinion information to obtain initial word information;
the emotion analysis module is used for carrying out positive and negative emotion part-of-speech tagging on the initial word information to obtain tagged word information, and calculating and outputting a text emotion polarity index;
and the data statistics and calculation module is used for performing statistics and calculation on various index variables of the input model, including the securities questioning amount, the securities reply amount, the reply timeliness, the effective text length of questioning content, the effective text length of reply content, whether reply is performed, market asset combination, market value factor, account and market value ratio factor, individual stock yield and market risk-free yield.
CN202010587191.1A 2020-06-24 2020-06-24 Stock public opinion investment decision analysis method and device Pending CN111738856A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010587191.1A CN111738856A (en) 2020-06-24 2020-06-24 Stock public opinion investment decision analysis method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010587191.1A CN111738856A (en) 2020-06-24 2020-06-24 Stock public opinion investment decision analysis method and device

Publications (1)

Publication Number Publication Date
CN111738856A true CN111738856A (en) 2020-10-02

Family

ID=72651348

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010587191.1A Pending CN111738856A (en) 2020-06-24 2020-06-24 Stock public opinion investment decision analysis method and device

Country Status (1)

Country Link
CN (1) CN111738856A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113222471A (en) * 2021-06-04 2021-08-06 西安交通大学 Asset wind control method and device based on new media data
CN113283994A (en) * 2021-05-31 2021-08-20 左虎 Investment decision analysis system
WO2022227213A1 (en) * 2021-04-25 2022-11-03 平安科技（深圳）有限公司 Industry recommendation method and apparatus, computer device and storage medium
CN116226363A (en) * 2023-05-08 2023-06-06 中信建投证券股份有限公司 Reference information generation method and device
CN116957140A (en) * 2023-06-29 2023-10-27 易方达基金管理有限公司 Stock prediction method and system based on NLP (non-linear point) factors

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022227213A1 (en) * 2021-04-25 2022-11-03 平安科技（深圳）有限公司 Industry recommendation method and apparatus, computer device and storage medium
CN113283994A (en) * 2021-05-31 2021-08-20 左虎 Investment decision analysis system
CN113222471A (en) * 2021-06-04 2021-08-06 西安交通大学 Asset wind control method and device based on new media data
CN113222471B (en) * 2021-06-04 2023-06-06 西安交通大学 Asset wind control method and device based on new media data
CN116226363A (en) * 2023-05-08 2023-06-06 中信建投证券股份有限公司 Reference information generation method and device
CN116957140A (en) * 2023-06-29 2023-10-27 易方达基金管理有限公司 Stock prediction method and system based on NLP (non-linear point) factors

Similar Documents

Publication Publication Date Title
CN111738856A (en) Stock public opinion investment decision analysis method and device
Kuzey et al. The impact of multinationality on firm value: A comparative analysis of machine learning techniques
CN105022825A (en) Financial variety price prediction method capable of combining financial news mining and financial historical data
CN106022522A (en) Method and system for predicting stocks based on big data published by internet
Wang et al. Cryptocurrency price prediction based on multiple market sentiment
CN106663302A (en) Method, system, and program for evaluating intellectual property right
CN109299252A (en) The viewpoint polarity classification method and device of stock comment based on machine learning
Karagozoglu et al. Volatility wisdom of social media crowds
Yuan et al. Mining emotions of the public from social media for enhancing corporate credit rating
Horky et al. Don't miss out on NFTs?! A sentiment-based analysis of the early NFT market
M'baye Labour Productivity and Economic Growth in the WAEMU Region: Both a Puzzle and a Key to Economic Development?
Miciuła The concept of FTS analysis in forecasting trends of exchange rate changes
Chen et al. A quantitative investment model based on random forest and sentiment analysis
Corbet et al. Investigating the academic response to cryptocurrencies: Insights from research diversification as separated by journal ranking
Huang et al. A learning-based contrarian trading strategy via a dual-classifier model
Gabhane et al. Behavioral Finance: Exploring The Influence Of Cognitive Biases On Investment Decisions
Nakano et al. Enhancing Sentiment Analysis based Investment by Large Language Models in Japanese Stock Market
Lu Research on influence from investor focus and investor sentiment in on-line social media upon stock market
Dhabe et al. Stock Market Trend Prediction Along with Twitter Sentiment Analysis
Balcarová Is competitiveness in the European Union converging
Zhang Housing price prediction using machine learning algorithm
CN113570455A (en) Stock recommendation method and device, computer equipment and storage medium
Lasek et al. Are Stock Markets Driven More by Sentiments than Efficiency?
Bakhshi et al. Developing a hybrid approach to credit priority based on accounting variables (using analytical network process (ANP) and multi-criteria decision-making)
Kammili A Study on Technical Analaysis of Indian FMCG Sector

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20201002

RJ01 Rejection of invention patent application after publication