CN108268451A - One B shareB affection index construction method and system - Google Patents

Abstract

The present invention provides a method and system for constructing a stock sentiment index. The method includes: S1, dividing each document into sentences according to the punctuation marks in each document related to the target stock issued in the current time period; S2, according to The emotional polarity of each word in each described sentence determines the emotional polarity of each described sentence, determines the emotional polarity of each described document according to the emotional polarity of each sentence in each described document; S3, according to each described document The emotional polarity of the target stock is used to construct the emotional index of the target stock in the current time period. The construction method of the invention is simple, more accurately reflects people's attitudes towards target stocks, and helps guide investors to avoid risks and make investment decisions.

Description

Method and system for constructing stock sentiment index

technical field

The invention belongs to the technical field of data analysis, and more specifically relates to a method and system for constructing a stock sentiment index.

Background technique

Securities analysts who have a deep understanding of the stock market predict the future development direction of the stock market and the degree of ups and downs based on the development of the stock market. Investors can refer to the forecast results of the stock market to make risk aversion and investment decisions. Therefore, the accuracy of forecasting results has an important impact on the success or failure of investment.

In the stock market, changes in stock prices are related to the country's macroeconomic development, the formulation of laws and regulations, the company's operations, investment psychology, and trading techniques, which are difficult to predict accurately. The forecasting behavior of securities analysts is only based on assumptions and preconditions, and their forecasts can only be used as reference opinions for investors. The current common stock forecasting methods are generally based on standardized financial data such as stock opening price, closing price and trading volume. In recent years, a large number of academic studies in the field of finance have found non-standardized financial data, such as investors' sentimental polarity volatility index on the market, related indices on regulatory policy uncertainty, and positive and negative investment reports published by investors on the Internet. Speech, etc., play a key role in explaining and analyzing stock market fluctuations.

Currently, there are few and imprecise methods for constructing non-standardized financial data. Therefore, there is an urgent need to construct new non-standardized financial data for stock forecasting.

Contents of the invention

In order to overcome the above-mentioned problem of few and inaccurate methods for constructing non-standardized financial data, or at least partially solve the above-mentioned problem, the present invention provides a method and system for constructing a stock sentiment index.

According to a first aspect of the present invention, a kind of stock sentiment index construction method is provided, comprising:

S1, according to the punctuation marks in each document related to the target stock released in the current time period, divide each document into sentences;

S2. Determine the emotional polarity of each of the sentences according to the emotional polarity of each word in each of the sentences, and determine the emotional polarity of each of the documents according to the emotional polarity of each of the sentences in each of the documents;

S3. According to the sentiment polarity of each document, construct the sentiment index of the target stock in the current time period.

Specifically, in the step S2, determining the emotional polarity of each of the sentences according to the emotional polarity of each word in each of the sentences specifically includes:

For any of the sentences, if the number of words with positive emotion in the sentence is greater than the number of words with negative emotion in the sentence, then the emotional polarity of the sentence is positive emotion;

For any of the sentences, if the number of words with positive emotion in the sentence is equal to the number of words with negative emotion in the sentence, then the emotional polarity of the sentence is neutral emotion;

For any sentence, if the number of words with positive emotion in the sentence is less than the number of words with negative emotion in the sentence, then the emotion polarity of the sentence is negative emotion.

Specifically, determining the emotional polarity of each of the documents according to the emotional polarity of each sentence in each of the documents in the step S2 specifically includes:

For any of the documents, if the number of positive sentiment statements in the document is greater than the number of negative sentiment statements in the document, then the sentiment polarity of the document is positive sentiment;

For any of the documents, if the number of sentences with positive sentiment in the document is equal to the number of sentences with negative sentiment in the document, then the sentiment polarity of the document is neutral sentiment;

For any document, if the number of sentences with positive sentiment in the document is less than the number of sentences with negative sentiment in the document, then the sentiment polarity of the document is negative sentiment.

Specifically, in the step S3, the sentiment index of the current time period of the target stock is constructed by the following formula:

Among them, S _t is the sentiment index of the target stock in the tth time period, is the number of positive sentiment documents published in the tth time period, The number of documents with negative sentiment published for the tth time period.

Specifically, the step S2 also includes:

For any of the words, if the emotional polarity of the word is positive emotion and the preceding word of the word is a negative word, then the word and the negative word are combined into a word, and the emotional polarity of the synthetic word is negative emotion;

For any of the words, if the emotional polarity of the word is negative emotion and the previous word of the word is a negative word, then this word and the negative word are combined into a word, and the emotional polarity of the synthetic word is positive emotion.

Specifically, after the step S3, it also includes when the current time period is Monday, adjusting the sentiment index of the target stock in the current time period by the following formula:

Wherein, S _t ' is the sentiment index after adjustment of the tth time period of the target stock, S _t is the sentiment index before the adjustment of the tth time period of the target stock, and S _t-1 is the tth time period of the target stock Sentiment index before adjustment in -1 time period, S _t-2 is the sentiment index of the target stock before adjustment in the t-2th time period, a ₁ , a ₂ and a ₃ are constants, a ₁ >a ₂ > a ₃ , λ are preset parameters.

Specifically, after the step S3, it also includes adjusting the sentiment index of the target stock in the current time period by the following formula when the target stock is closed for k consecutive days:

S _k+1 '=e ^-kλ S ₁ +e ^-(k-1)λ S ₂ +...+e ^-λ S _k +S _k+1 ;

Wherein, k represents that the first k time periods of the current time period are closed, S _k+1 ' is the adjusted sentiment index of the current time period of the target stock, and S _k+1 is before the adjustment of the current time period of the target stock Sentiment index, S ₁ is the sentiment index of the target stock in the first time period of the market break, S ₂ is the sentiment index of the target stock in the second time period of the market break, S _k is the kth of the market break The sentiment index of the target stock in the time period, λ is a preset parameter.

According to a second aspect of the present invention, a kind of stock sentiment index construction system is provided, comprising:

A division module, configured to divide each document into sentences according to the punctuation marks in each document related to the target stock issued in the current time period;

An acquisition module, configured to acquire the emotional polarity of each of the sentences according to the number of words of various emotional polarities in each of the sentences, and acquire according to the number of the sentences of various emotional polarities in each of the documents the sentimental polarity of each said document;

The construction module is used to construct the sentiment index of the target stock in the current time period according to the number of the documents with various sentiment polarities.

According to a third aspect of the present invention, a kind of stock sentiment index construction equipment is provided, comprising:

at least one processor, at least one memory, and a bus; wherein,

The processor and the memory complete mutual communication through the bus;

The memory stores program instructions executable by the processor, and the processor invokes the program instructions to execute the aforementioned method.

According to a fourth aspect of the present invention, there is provided a non-transitory computer-readable storage medium for storing the computer program of the aforementioned method.

The invention provides a method and system for constructing a stock sentiment index. The method obtains documents related to target stocks, determines the sentiment polarity of each sentence in the document according to the sentiment polarity of the words that make up each sentence in the document, and determines the sentiment polarity of each statement according to the constituent documents. The emotional polarity of each statement determines the emotional polarity of each document, and constructs the sentiment index of the target stock according to the emotional polarity of each document. The construction method is simple, and it more accurately reflects people’s attitude towards the target stock, which is helpful for guiding Investors make risk aversion and investment decisions.

Description of drawings

Fig. 1 is the overall schematic flow chart of the stock sentiment index construction method that the embodiment of the present invention provides;

Fig. 2 is a schematic diagram of the overall structure of the stock sentiment index construction system provided by the embodiment of the present invention;

Fig. 3 is a schematic diagram of the overall structure of the stock sentiment index construction device provided by the embodiment of the present invention.

Detailed ways

The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

In one embodiment of the present invention, a method for constructing a stock sentiment index is provided. Fig. 1 is a schematic diagram of the overall process flow of the method for constructing a stock sentiment index provided by an embodiment of the present invention. The punctuation mark in each relevant document divides each described document into sentences; S2, obtains the emotional polarity of each described sentence according to the number of words of various emotional polarities in each described sentence, and according to each described The number of sentences of various emotional polarities in the document obtains the emotional polarity of each document; S3, constructing the emotional index of the target stock in the current time period according to the number of the documents of various emotional polarities.

Specifically, in S1, the current time period can be one month, one day, one week, one hour or one minute, so the sentiment index of each time period of the target stock can be constructed to form a time series sentiment index. The target stock is a stock that needs to construct a sentiment index. The documents related to the target stock include the user's remarks on the target stock acquired from webpages, articles about the target stock issued by securities analysts, stock researchers or investors, and the like. Each of the documents is divided into one or more sentences according to the punctuation in each of the documents. The punctuation marks include one or more of full stop, comma, colon, semicolon and exclamation point.

In S2, word segmentation is performed on each of the sentences, and the emotional polarity of each word in each of the sentences is determined through an emotion polarity dictionary commonly used in the field of semantic analysis. The emotional polarity, also known as emotional orientation, refers to the emotional color of the text, such as positive, negative, and neutral. The type of emotional polarity is not limited in this embodiment. For any sentence, the number of words of various emotional polarities in the sentence is counted respectively, and the emotional polarity of the sentence is determined according to the number of words of various emotional polarities in the sentence. For any document, count the number of sentences with various emotional polarities in the document, and determine the emotional polarity of the document according to the number of sentences with various emotional polarities in the document.

In S3, the number of the documents of various emotional polarities is counted respectively, and the sentiment index of the current time period of the target stock is constructed according to the number of the documents of various emotional polarities. The sentiment index reflects people's attitude towards the target stock. The sentiment index constructed in this embodiment is more reasonable and intuitive, and can be used as a reference for investors to avoid risks and make investment decisions.

In this embodiment, by obtaining each document related to the target stock, the sentiment polarity of each sentence in the document is determined according to the sentiment polarity of the words that make up each sentence in the document, and the sentiment of each document is determined according to the sentiment polarity of each sentence that makes up the document Polarity, according to the emotional polarity of each document to construct the sentiment index of the target stock, the construction method is simple, more accurately reflects people's attitude towards the target stock, and helps guide investors to avoid risks and make investment decisions.

On the basis of the above-mentioned embodiments, determining the emotional polarity of each of the sentences according to the emotional polarity of each word in the step S2 in this embodiment specifically includes: for any of the sentences, if the The number of words of positive emotion in the sentence is greater than the number of words of negative emotion in the sentence, then the emotional polarity of the sentence is positive emotion; for any described sentence, if the number of words of positive emotion in the sentence Equal to the number of words of negative emotion in this sentence, then the emotion polarity of this sentence is neutral emotion; For any described sentence, if the number of words of positive emotion in this sentence is less than the words of negative emotion in this sentence , then the emotional polarity of the statement is negative emotion.

Specifically, since sentences can express more complete attitudes and opinions than words, the sentiment expressed in a document can be better determined in units of sentences. In order to determine the emotional polarity of each sentence in the document, first divide each document into sentences according to the punctuation marks in each document, and count the number of words containing positive emotions and words with negative emotions in each sentence. For any sentence i, it is assumed that the number of words in sentence i containing positive emotions is p _i , and the number of words containing negative emotions is n _i . If p _i >n _i , then determine the emotion polarity of sentence i as positive emotion; if p _i =n _i , determine the emotion polarity of sentence i as neutral emotion; if p _i <n _i , determine sentence i The emotional polarity of is negative emotion.

On the basis of the above-mentioned embodiments, determining the emotional polarity of each document according to the emotional polarity of each statement in the step S2 in this embodiment specifically includes: for any of the documents, if the If the number of sentences with positive sentiment in the document is greater than the number of sentences with negative sentiment in the document, then the sentiment polarity of the document is positive sentiment; for any document, if the number of sentences with positive sentiment in the document Equal to the number of sentences with negative sentiment in the document, the sentiment polarity of the document is neutral sentiment; for any document, if the number of sentences with positive sentiment in the document is less than the sentence with negative sentiment in the document , the sentiment polarity of the document is negative sentiment.

Specifically, in this embodiment, the emotional polarity of each document is determined according to the emotional polarity of each sentence in each document. For any document j, suppose the number of sentences in document j containing positive sentiment is p _j , and the number of sentences containing negative sentiment is n _j . If p _j >n _j , determine the sentiment polarity of document j as positive sentiment; if p _j =n _j , determine the sentiment polarity of document j as neutral sentiment; if p _j <n _j , determine document j The emotional polarity of is negative emotion.

On the basis of the foregoing embodiments, in step S3 described in this embodiment, the sentiment index of the current time period of the target stock is constructed by the following formula:

Among them, S _t is the sentiment index of the target stock in the tth time period, is the number of positive sentiment documents published in the tth time period, The number of positive sentiment documents published for the tth time period.

Specifically, multiple documents about the target stock will be released in a time period, assuming that the number of positive emotional documents released in the tth time period is The number of published documents with negative sentiment is Then the sentiment index of the target stock time series is:

It can be seen from the above formula that the value of the sentiment index S _t changes in the interval [-1,1]. If S _t >0, it means that people have a positive attitude towards the target stock; if S _t =0, it means that people have a neutral attitude towards the stock market; if S _t <0, it means that people have a positive attitude towards the stock market Negative attitude. Taking the current time period as the tth time period, using the above formula to calculate the sentiment index of the current time period.

On the basis of the above-mentioned embodiments, the step S2 in this embodiment also includes: for any of the words, if the emotional polarity of the word is a positive emotion and the previous word of the word is a negative word, then the The word and the negative word are synthesized into one word, and the emotional polarity of the synthesized word is negative emotion; for any of the words, if the emotional polarity of the word is negative emotion and the preceding word of the word is a negative word , then the word and the negative word are synthesized into one word, and the emotional polarity of the synthesized word is positive emotion.

Specifically, for any word in any sentence, if the previous word of the word is a negative word, and the negative word is a word expressing a negative meaning, then the negative word and the word are regarded as the same word, and the expression is the same as The opposite sentiment of the word.

On the basis of the above-mentioned embodiments, after the step S3 in this embodiment, it also includes when the current time period is Monday, adjusting the sentiment index of the target stock in the current time period by the following formula:

Wherein, S _t ' is the sentiment index after adjustment of the tth time period of the target stock, S _t is the sentiment index before the adjustment of the tth time period of the target stock, and S _t-1 is the tth time period of the target stock Sentiment index before adjustment in -1 time period, S _t-2 is the sentiment index of the target stock before adjustment in the t-2th time period, a ₁ , a ₂ and a ₃ are constants, a ₁ >a ₂ > a ₃ , λ are preset parameters.

Specifically, this embodiment takes into account the impact of the calendar effect on the construction of the stock sentiment index. The calendar effect is a typical manifestation of financial market anomalies, which means that the average return rate on Monday is significantly lower than the return rate on other trading days within a week. It originally included a flood of news that piled up in the market after the market closed on Friday, which had a significant impact on people's investment decisions. Therefore, when the current time period is Monday, adjust Monday's sentiment index S _t to:

Among them, S _t on the right side of the equal sign is the sentiment index before Monday’s adjustment, S _t-1 is the sentiment index before Sunday’s adjustment, and S _t-2 is the sentiment index before Saturday’s adjustment. a ₁ , a ₂ and a ₃ are constants, and can be set as a ₁ =2, a ₂ =1, and a ₃ =0. The sentiment index S _t ' of Monday after adjustment is the weighted value of the sentiment indices of Friday, Saturday and Sunday before adjustment, and the weight is expressed in the form of an exponential function. As time gets farther away from Monday, the weight decreases exponentially, so a ₁ >a ₂ >a ₃ . In this embodiment, the sentiment index of the target stock on Monday is adjusted by considering the calendar effect, so that the constructed sentiment index is more accurate.

On the basis of the above-mentioned embodiments, after the step S3 in this embodiment, it also includes when the market is closed in one or more time periods before the current time period, the sentiment of the target stock in the current time period by the following formula The index is adjusted:

S _k+1 '=e ^-kλ S ₁ +e ^-(k-1)λ S ₂ +...+e ^-λ S _k +S _k+1 ;

Wherein, k represents that the first k time periods of the current time period are closed, S _k+1 ' is the adjusted sentiment index of the current time period of the target stock, and S _k+1 is before the adjustment of the current time period of the target stock Sentiment index, S ₁ is the sentiment index of the target stock in the first time period of the market break, S ₂ is the sentiment index of the target stock in the second time period of the market break, S _k is the kth The sentiment index of the target stock in the time period, λ is a preset parameter.

In particular, this embodiment takes into account the generalization of calendar effects for holiday situations. The stock market will be closed on national statutory holidays and the following special dates. When the market is closed for k consecutive time periods, the sentiment index of the k+1th time period, that is, the sentiment index of the first time period after the market opens, can be adjusted as follows:

S _k+1 ′=e ^−kλ S ₁ +e ^−(k−1)λ S ₂ + . . . +e ^−λ S _k +S _k+1 .

This embodiment considers the situation that the stock market is closed on national statutory holidays and some special dates, and adjusts the sentiment index in the first time period after the market close, so as to make the constructed sentiment index more accurate.

In another embodiment of the present invention, a kind of stock sentiment index construction system is provided, and Fig. 2 is the stock sentiment index construction system overall structure schematic diagram that the embodiment of the present invention provides, and this system comprises division module 1, acquisition module 2 and construction module 3 ,in:

The division module 1 is used to divide each document into sentences according to the punctuation marks in each document related to the target stock released in the current time period; The number of words of polarity obtains the emotional polarity of each described sentence, obtains the emotional polarity of each described document according to the number of described sentences of various emotional polarities in each described document; the building block 3 It is used to construct the sentiment index of the target stock in the current time period according to the number of the documents with various sentiment polarities.

Specifically, the current time period may be one month, one day, one week, one hour or one minute, so the sentiment index of each time period of the target stock may be constructed to form a time series sentiment index. The target stock is a stock that needs to construct a sentiment index. The documents related to the target stock include the user's remarks on the target stock acquired from webpages, articles about the target stock issued by securities analysts, stock researchers or investors, and the like. The division module 1 divides each document into one or more sentences according to the punctuation in each document. The punctuation marks include one or more of full stop, comma, colon, semicolon and exclamation point.

The acquisition module 2 performs word segmentation on each of the sentences, and determines the emotional polarity of each word in each of the sentences through the sentimental polarity dictionary commonly used in the field of semantic analysis. The emotional polarity, also known as emotional orientation, refers to the emotional color of the text, such as positive, negative, and neutral. The type of emotional polarity is not limited in this embodiment. For any sentence, the number of words of various emotional polarities in the sentence is counted respectively, and the emotional polarity of the sentence is determined according to the number of words of various emotional polarities in the sentence. For any document, count the number of sentences with various emotional polarities in the document, and determine the emotional polarity of the document according to the number of sentences with various emotional polarities in the document.

The construction module 3 counts the number of documents of various emotional polarities respectively, and constructs the sentiment index of the target stock in the current time period according to the number of documents of various emotional polarities. The sentiment index reflects people's attitude towards the target stock. The sentiment index constructed in this embodiment is more reasonable and intuitive, and can be used as a reference for investors to avoid risks and make investment decisions.

In this embodiment, by obtaining each document related to the target stock, the sentiment polarity of each sentence in the document is determined according to the sentiment polarity of the words that make up each sentence in the document, and the sentiment of each document is determined according to the sentiment polarity of each sentence that makes up the document Polarity, according to the emotional polarity of each document to construct the sentiment index of the target stock, the construction method is simple, more accurately reflects people's attitude towards the target stock, and helps guide investors to avoid risks and make investment decisions.

On the basis of the above-mentioned embodiments, the acquisition module in this embodiment is specifically used for: for any of the sentences, if the number of words with positive emotions in the sentence is greater than the number of words with negative emotions in the sentence, Then the emotion polarity of this sentence is positive emotion; For any described sentence, if the number of words of positive emotion in this sentence is equal to the number of words of negative emotion in this sentence, then the emotion polarity of this sentence is middle Sexual emotion; For any one of the sentences, if the number of words of positive emotion in the sentence is less than the number of words of negative emotion in the sentence, then the emotional polarity of the sentence is negative emotion.

On the basis of the above embodiments, the acquisition module in this embodiment is specifically configured to: for any document, if the number of sentences with positive emotions in the document is greater than the number of sentences with negative emotions in the document, Then the sentiment polarity of the document is positive sentiment; for any document, if the number of positive sentiment sentences in the document is equal to the number of negative sentiment sentences in the document, then the sentiment polarity of the document is medium sexual sentiment; for any document, if the number of sentences with positive sentiment in the document is less than the number of sentences with negative sentiment in the document, then the sentiment polarity of the document is negative sentiment.

On the basis of the foregoing embodiments, the construction unit in this embodiment constructs the sentiment index of the current time period of the target stock through the following formula:

Among them, S _t is the sentiment index of the target stock in the tth time period, is the number of positive sentiment documents published in the tth time period, The number of positive sentiment documents published for the tth time period.

On the basis of the above-mentioned embodiments, the acquisition unit in this embodiment is further configured to: for any of the words, if the emotional polarity of the word is positive and the previous word of the word is a negative word, then This word and described negation word are synthesized into a word, and the emotional polarity of the compound word is negative emotion; For any described word, if the emotional polarity of this word is negative emotion and the preceding word of this word is negation word, then the word and the negative word are synthesized into one word, and the emotional polarity of the synthesized word is positive emotion.

On the basis of the above embodiments, this embodiment also includes a first adjustment unit, which is used to adjust the sentiment index of the target stock in the current time period by the following formula when the current time period is Monday:

Wherein, S _t ' is the sentiment index after adjustment of the tth time period of the target stock, S _t is the sentiment index before the adjustment of the tth time period of the target stock, and S _t-1 is the tth time period of the target stock Sentiment index before adjustment in -1 time period, S _t-2 is the sentiment index of the target stock before adjustment in the t-2th time period, a ₁ , a ₂ and a ₃ are constants, a ₁ >a ₂ > a ₃ , λ are preset parameters.

On the basis of the above embodiments, this embodiment also includes a second adjustment unit, which is used to adjust the current time period of the target stock by the following formula when the market is closed in one or more time periods before the current time period The sentiment index is adjusted by:

S _k+1 '=e ^-kλ S ₁ +e ^-(k-1)λ S ₂ +...+e ^-λ S _k +S _k+1 ;

Wherein, k represents that the first k time periods of the current time period are closed, S _k+1 ' is the adjusted sentiment index of the current time period of the target stock, and S _k+1 is before the adjustment of the current time period of the target stock Sentiment index, S ₁ is the sentiment index of the target stock in the first time period of the market break, S ₂ is the sentiment index of the target stock in the second time period of the market break, S _k is the kth The sentiment index of the target stock in the time period, λ is a preset parameter.

The present embodiment provides a stock sentiment index construction device, and Fig. 3 is a schematic diagram of the overall structure of the stock sentiment index construction device provided by the embodiment of the present invention, the device includes: at least one processor 31, at least one memory 32 and a bus 33; wherein,

The processor 31 and the memory 32 complete mutual communication through the bus 33;

The memory 32 stores program instructions that can be executed by the processor 31, and the processor calls the program instructions to execute the methods provided by the above-mentioned method embodiments, for example, including: S1, issued according to the current time period Punctuation marks in each document related to the target stock, each described document is divided into sentences; S2, obtain the emotional polarity of each described sentence according to the number of words of various emotional polarities in each described sentence, according to The number of sentences of various emotional polarities in each described document obtains the emotional polarity of each described document; S3, according to the number of the described documents of various emotional polarities, constructs the current time period of the target stock emotional index.

This embodiment provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the methods provided in the above method embodiments, for example, including : S1, divide each document into sentences according to the punctuation marks in each document related to the target stock released in the current time period; S2, obtain each The emotional polarity of the sentence is obtained according to the number of sentences of various emotional polarities in each of the documents; S3, according to the number of the documents of various emotional polarities, A sentiment index for the current time period of the target stock is constructed.

Those of ordinary skill in the art can understand that all or part of the steps for realizing the above-mentioned method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the It includes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

The above-described stock sentiment index construction device embodiment is only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units , which can be located in one place, or can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. It can be understood and implemented by those skilled in the art without any creative efforts.

Through the above description of the implementations, those skilled in the art can clearly understand that each implementation can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic discs, optical discs, etc., including several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments.

Finally, the method of the present application is only a preferred embodiment, and is not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (10)

1. A method for constructing a stock sentiment index is characterized by comprising the following steps:

s1, dividing each document into sentences according to punctuation marks in each document related to the target stock issued in the current time period;

s2, determining the emotion polarity of each statement according to the emotion polarity of each word in each statement, and determining the emotion polarity of each document according to the emotion polarity of each statement in each document;

and S3, constructing the emotion index of the target stock in the current time period according to the emotion polarity of each document.

2. The method according to claim 1, wherein the determining the emotion polarity of each sentence according to the emotion polarity of each word in each sentence in step S2 specifically comprises:

for any statement, if the number of words with positive emotion in the statement is greater than the number of words with negative emotion in the statement, the emotion polarity of the statement is positive emotion;

for any statement, if the number of words with positive emotion in the statement is equal to the number of words with extreme emotion in the statement, the emotion polarity of the statement is neutral emotion;

for any statement, if the number of words with positive emotion in the statement is smaller than the number of words with negative emotion in the statement, the emotion polarity of the statement is negative emotion.

3. The method according to claim 1, wherein the determining the emotion polarity of each document according to the emotion polarity of each sentence in each document in step S2 specifically comprises:

for any document, if the number of sentences with positive emotions in the document is greater than that of sentences with negative emotions in the document, the emotion polarity of the document is the positive emotions;

for any document, if the number of sentences with positive emotion in the document is equal to the number of sentences with extreme emotion in the document, the emotion polarity of the document is neutral emotion;

for any document, if the number of sentences with positive emotion in the document is less than the number of sentences with extreme emotion in the document, the emotion polarity of the document is negative emotion.

4. The method as claimed in claim 1, wherein the sentiment index of the current time period of the target stock is constructed in the step S3 by the following formula:

wherein S is_tFor the sentiment index of the target stock for the t-th time period,the number of documents for positive emotion issued for the t-th time period,the number of documents with negative emotion issued for the t-th time period.

5. The method according to any one of claims 1 to 4, wherein the step S2 further comprises:

for any word, if the emotion polarity of the word is positive emotion and the previous word of the word is negative word, combining the word and the negative word into one word, wherein the emotion polarity of the combined word is negative emotion;

for any word, if the emotion polarity of the word is negative emotion and the previous word of the word is a negative word, the word and the negative word are combined into one word, and the emotion polarity of the combined word is positive emotion.

6. The method according to any one of claims 1-4, wherein the step S3 is further followed by adjusting the sentiment index of the target stock in the current time period when the current time period is Monday by the following formula:

wherein S is_t' adjusted sentiment index for the t-th time period of the target stock, S_tAdjusting the emotion index for the t-th time period of the target stock, S_t-1The emotion index before the t-1 time period adjustment of the target stock, S_t-2The emotion index before the adjustment for the t-2 th time period of the target stock, a₁、a₂And a₃Is a constant number, a₁>a₂>a₃And λ is a preset parameter.

7. The method as claimed in any one of claims 1-4, wherein the step S3 further comprises adjusting the sentiment index of the target stock in the current time period when the target stock is in k-day break, according to the following formula:

S_k+1'＝e^-kλS₁+e^-(k-1)λS₂+…+e^-λS_k+S_k+1；

wherein k represents the first k time slots of the current time slot, S_k+1' adjusted sentiment index for current time period of the target stock, S_k+1Adjusting the emotion index of the target stock in the current time period, S₁The sentiment index of the target stock for the first time period of the holiday, S₂For the sentiment index of the target stock in the second time period of the said hou city, S_kAnd lambda is a preset parameter, and is the sentiment index of the target stock in the kth time period of the hound city.

8. A stock sentiment index construction system is characterized by comprising:

the dividing module is used for dividing each document into sentences according to punctuation marks in each document related to the target stock issued in the current time period;

the obtaining module is used for obtaining the emotional polarity of each statement according to the number of the words with various emotional polarities in each statement, and obtaining the emotional polarity of each document according to the number of the statements with various emotional polarities in each document;

and the construction module is used for constructing the emotion index of the target stock in the current time period according to the number of the documents with various emotion polarities.

9. A stock sentiment index construction device is characterized by comprising:

at least one processor, at least one memory, and a bus; wherein,

the processor and the memory complete mutual communication through the bus;

the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the method of any one of claims 1 to 7.