WO2021012684A1

WO2021012684A1 - Method and system for establishing market sentiment monitoring system

Info

Publication number: WO2021012684A1
Application number: PCT/CN2020/078384
Authority: WO
Inventors: 郑晴晓; 于洋; 程国艮
Original assignee: 中译语通科技股份有限公司
Priority date: 2019-07-23
Filing date: 2020-03-09
Publication date: 2021-01-28
Also published as: CN110400173A

Abstract

A method and system for establishing a market sentiment monitoring system. The method comprises: acquiring a sample text set comprising a plurality of pieces of sample text (110); performing corpus labeling on any one piece of sample text, and acquiring a sentiment label and a sentiment degree label of the any one piece of sample text (120); on the basis of the sentiment label and the sentiment degree label of each piece of sample text, acquiring a sentiment monitoring degree (130); and on the basis of the sentiment monitoring degree and preset sentiment acuity, constructing a market sentiment monitoring system (140), the market sentiment monitoring system being used to monitor market sentiments. In the described method and system, corresponding sentiment labels and sentiment degree labels are obtained by performing corpus labeling on sample text, and then the sentiment monitoring degree is selected from among the sentiment labels of a large amount of sample text to construct a market sentiment detection system, which helps to implement objective and accurate market sentiment monitoring, and helps investors reduce erroneous determinations and improve investment correctness.

Description

Method and system for establishing market sentiment monitoring system

Technical field

The invention relates to the technical field of text analysis, in particular to a method and system for establishing a market sentiment monitoring system.

Background technique

The correct judgment of the market situation is an important prerequisite for investors to make effective investments. As an important factor affecting the market situation, market sentiment greatly affects the accuracy of market situation judgments. How to accurately count and judge market sentiment has always been an issue of concern.

In the past, investors mainly conducted national policy analysis and industry research by browsing news information and referencing research reports, and then comprehensive technical analysis and other methods to judge the market situation and invest. Due to the lack of information, it is extremely difficult to dig market sentiment from the limited information.

With the rapid development of the Internet, all kinds of market-related news and information emerge in an endless stream. The content covers all aspects, and the number is also exploding. The timeliness and diversity of market information not only provide investors with rich market sentiment monitoring materials, but also make it difficult for investors to accurately judge market sentiment from massive amounts of information in a short time. Moreover, the efficiency of artificial statistical judgment of market sentiment is low, and the market sentiment obtained is also highly subjective, with low accuracy and poor timeliness, and cannot meet the needs of investors.

How to establish a market sentiment monitoring system to achieve efficient, accurate and objective market sentiment monitoring has become an urgent problem for people to solve.

Summary of the invention

The embodiment of the present invention provides a method and system for establishing a market sentiment monitoring system to solve the problems of poor objectivity, low accuracy, and poor timeliness of the existing manual analysis of market sentiment.

In the first aspect, an embodiment of the present invention provides a method for establishing a market sentiment monitoring system, including:

Obtain a sample text set including multiple sample texts;

Corpus tagging any of the sample texts, and obtaining the emotion label and the emotion degree label of any sample text;

Obtaining the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text;

Based on the emotion monitoring degree and preset emotion acuity, a market emotion monitoring system is constructed; the market emotion monitoring system is used to monitor market emotions.

In the second aspect, an embodiment of the present invention provides a system for establishing a market sentiment monitoring system, including:

The sample acquisition unit is used to acquire a sample text set including multiple sample texts;

The corpus tagging unit is used to tag any of the sample texts to obtain the emotion label and the emotion degree tag of any sample text;

An emotion monitoring degree obtaining unit, configured to obtain the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text;

The system establishment unit is configured to construct a market emotion monitoring system based on the emotion monitoring degree and preset emotion acuity; the market emotion monitoring system is used to monitor market emotions.

In a third aspect, an embodiment of the present invention provides an electronic device that includes a processor, a communication interface, a memory, and a bus. The processor, the communication interface, and the memory communicate with each other through the bus, and the processor can call logic in the memory. Instructions to perform the steps of the method provided in the first aspect.

In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method provided in the first aspect are implemented.

According to an embodiment of the present invention, a method and system for establishing a market sentiment monitoring system are obtained by labeling sample texts on corpus to obtain corresponding sentiment labels and sentiment degree labels, and then selecting sentiment monitoring degrees from the sentiment labels of a large number of sample texts to construct a market Emotion detection reflects, helps to achieve objective and accurate market sentiment monitoring, helps investors reduce misjudgments, and improves investment accuracy.

Description of the drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the drawings used in the description of the embodiments or the prior art. Obviously, the drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

FIG. 1 is a schematic flowchart of a method for establishing a market sentiment monitoring system provided by an embodiment of the present invention;

2 is a schematic structural diagram of a market sentiment monitoring system establishment system provided by an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of the embodiments of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

The timeliness and diversity of market information not only provide investors with rich market sentiment monitoring materials, but also make it difficult for investors to accurately judge market sentiment from massive amounts of information in a short time. Moreover, the artificial statistical judgment of market sentiment is inefficient, low in objectivity and accuracy, and poor in timeliness, which cannot meet the needs of investors. In this regard, the embodiment of the present invention provides a method for establishing a market sentiment monitoring system. Fig. 1 is a schematic flowchart of a method for establishing a market sentiment monitoring system provided by an embodiment of the present invention. As shown in Fig. 1, the method includes:

Step 110: Obtain a sample text set including a plurality of sample texts.

Here, the sample text may be a news report, a market report or a comment on a social platform, etc., which is not specifically limited in the embodiment of the present invention. The sample text set contains a large number of sample texts.

Step 120: Perform corpus labeling on any sample text, and obtain the emotion label and emotion degree label of the sample text.

Specifically, for any sample text, emotion analysis is performed on each clause in the sample text, and an emotion label and an emotion degree label for each clause are formed. For any clause, the emotion label refers to the emotion type obtained after sentiment analysis of the clause, and the emotion degree label refers to the intensity of the emotion type obtained after the emotion analysis of the clause. From this, the emotion label and emotion degree label of any sample text are actually obtained for the emotion label and emotion degree label of the sentence in the sample text.

In addition, before step 120 is performed, a corpus labeling model for corpus labeling can also be pre-trained, which can be obtained by training in the following manner: First, a large number of training texts and emotion labels and emotional degree labels corresponding to the training texts are collected; among them, The training text is the text used for the training of the corpus labeling model, and the sentiment label and sentiment degree label of the training text are labelled by professionals. Then, the initial model is trained based on the training text and its emotional label and emotional degree label to obtain the corpus labeling model. The initial model may be a single neural network model, or a combination of multiple neural network models, and the embodiment of the present invention does not specifically limit the type and structure of the initial model.

Step 130: Obtain the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text.

Specifically, after obtaining the sentiment label and sentiment degree label of each sample text, the sentiment label and sentiment degree label of each sample text can be counted, and then sentiment monitoring degree can be selected from the sentiment label to realize the classification of market sentiment . Here, the emotion monitoring degree is a representative emotion label selected from a large number of emotion labels, and the essence of the emotion monitoring degree is a type of market sentiment. There are multiple levels of emotion monitoring, and different levels of emotion monitoring can represent different types of emotions with different degrees of positive or negative.

Step 140: Construct a market sentiment monitoring system based on the sentiment monitoring degree and preset sentiment acuity; the market sentiment monitoring system is used to monitor market sentiment.

Specifically, the preset emotional acuity is a preset indicator that reflects the intensity of the emotion corresponding to the emotional monitoring degree. For example, the preset emotional acuity can be divided into 6 levels from 0 to 5, and the 6-level preset emotional acuity Corresponding emotions increase in turn. The market sentiment monitoring system thus constructed includes the sentiment monitoring degree and the preset emotional acuity. The market sentiment detection system is applied to the market sentiment monitoring, and the emotional acuity corresponding to each emotion monitoring degree can be obtained based on the text collection collected in real time. Here, the value of the emotional acuity corresponding to any emotional monitoring degree is within the preset emotional acuity range.

The method provided by the embodiment of the present invention obtains the corresponding emotion label and emotion degree label by labeling the sample text, and then selects the emotion monitoring degree from the emotion label of a large number of sample texts, constructs the market emotion detection manifestation, and helps to achieve objectiveness. , Accurate market sentiment monitoring helps investors reduce misjudgments and improve investment accuracy.

Based on the above embodiment, in the method, step 110 specifically includes: determining the data collection method; the data collection method is determined according to the preset collection object, collection frequency, collection content, and object level; based on the data collection method, through web crawlers Collect sample texts from the collection objects to construct a sample text set.

Specifically, the collection object is a text collection object, and the collection object may be a market-related institution website, a financial news website, a social media or a market report published by a research institution, etc., which is not specifically limited in the embodiment of the present invention. Collection frequency is the frequency of text collection. Collection content includes title, text (including pictures and videos), release time and release platform, etc. The object level is used to characterize the importance of the collection object. The object level can be based on the popularity or credibility of the collection object Decided. Each collection object is set with a corresponding collection frequency, collection content and object level.

A web crawler (also known as a web spider or web robot) is a program or script that automatically crawls information on the World Wide Web in accordance with preset rules. After determining the data collection method, you can follow the data collection method and grab a sample text from each collection object through a web crawler. Further, the weight for selecting the emotion monitoring degree based on the emotion label and emotion degree label of the sample text captured at the corresponding collection object can also be set according to the object level.

Based on any of the foregoing embodiments, in the method, between step 110 and step 120, the method further includes: preprocessing the sample text set; the preprocessing includes data cleaning and data management.

Specifically, the sample text captured from each collection object, especially the sample text captured from social media, may contain a large number of typos, grammatical errors, and stop words, as well as invalid, outdated, or extremely market-related. Weak data, the existence of these problems is likely to interfere with the accuracy of emotion monitoring. Data cleaning can find and correct identifiable errors in the sample text, such as checking the consistency of the data, removing duplicates, and handling invalid and missing values. Data cleaning of the sample text can ensure the accuracy and validity of the sample text.

In addition, the sample text captured from each collection object has very different expression styles. It is necessary to unify each sample text through standardized processing to facilitate sentiment analysis and corpus labeling of each sample text. Data governance refers to the process of changing from using scattered data to using unified master data. Sample texts can be governed through six aspects: metadata, master data, data standards, data models, data quality, and data security, so as to realize the sample text Data standardization.

After preprocessing for each sample text, each sample text after preprocessing is applied to the corpus annotation to facilitate the selection of the subsequent emotion monitoring degree and the establishment of the market emotion monitoring system.

Based on any of the foregoing embodiments, in the method, between step 110 and step 120, the method further includes: performing market relevance screening on the sample text set, and deleting sample texts whose market relevance is lower than a preset relevance threshold.

Specifically, the sample texts in the sample text set involve various aspects of information content, and there may be some sample texts that are not related to the market. Before applying the text to the selection of sentiment monitoring, it is necessary to filter out those that are not related to the market. Sample text, only the sample text related to the market is retained. Here, the market relevance is used to measure the degree of relevance between the sample text and the market, and the preset relevance threshold may be a preset threshold for measuring whether the sample text is relevant to the market.

Any sample text can be input into the correlation model to obtain the market relevance of the sample text output by the correlation model. The correlation model here can be obtained based on sample text and market correlation training corresponding to the sample text.

Based on any of the foregoing embodiments, in the method, after step 120, the method further includes: constructing an emotion dictionary based on the emotion label and the emotion degree label of the sample text.

Specifically, after step 120, an initial sentiment dictionary and corpus can be constructed according to the sentiment label and sentiment degree label of the sample text, and the initial sentiment dictionary and corpus can be regarded as seed words and seed libraries, and word similarity and new words can be used. Discovery, text similarity and other technologies expand the initial sentiment dictionary and corpus to obtain the sentiment dictionary.

Based on any of the foregoing embodiments, in the method, step 130 specifically includes: counting the number of clauses of each emotion label and emotion level label based on the emotion label and emotion level label of each sample text; based on each emotion label and emotion level label The number of sentences of the degree label to obtain the emotional monitoring degree.

Specifically, after obtaining the sentiment label and sentiment degree label of each sentence in the sample text, the sentiment label and sentiment degree label of each sentence in the sample text are counted to obtain the corresponding sentiment label and sentiment degree label. The number of clauses. Furthermore, based on the number of clauses of each emotion label and emotion degree label, Pareto's law or other statistical methods are adopted to select the emotion monitoring degree from the emotion label to realize the classification of market sentiment.

Based on any of the foregoing embodiments, in the method, between step 130 and step 140, the method further includes: performing result back-testing and data correction on the emotion monitoring degree.

Specifically, after obtaining the emotion monitoring degree, the type and degree of market sentiment can be back-tested, so as to determine whether the current emotion monitoring degree can accurately reflect the emotion of the sample text.

If it is judged that the current emotion monitoring degree cannot accurately reflect the emotion of the sample text, the emotion monitoring degree is corrected to ensure that the emotion monitoring degree can be closer to the actual situation of the text emotion.

Based on any of the foregoing embodiments, in the method, the emotion monitoring degree includes worry, fear, disappointment, panic, despair, hope, optimism, peace of mind, excitement, and excitement.

Specifically, the interpretation of the above ten levels of emotional monitoring and the corresponding emotional sensitivity are shown in the following table:

Based on any of the foregoing embodiments, the method for establishing a market sentiment monitoring system specifically includes the following steps:

First, determine the collection object and its corresponding collection frequency, collection content and object level, and thus determine the data collection method. Here, the collection objects include, but are not limited to, domestic and foreign government agency websites, financial news websites, public reports from research institutions, and social media. The object level is divided according to the platform influence of the collected objects. The collected content can cover information such as title, body (including pictures and videos involved in the text), release time, and release platform.

Subsequently, the sample text is collected. Based on the data collection method, a web crawler goes to each collection object to collect sample text. Here, the sample text involves 65 languages, covering more than 200 countries and regions around the world, and the final sample text size reaches more than 50 million.

Next, perform data processing on the sample text. The data processing here includes data cleaning, data management, text screening and corpus labeling. Among them, the text screening is that professionals (including financial professionals, economics majors, psychology majors, linguistics and other professional background personnel) check and filter the sample texts after cleaning and governance one by one, and screen out the sample texts related to the market. Corpus labeling is to perform sentiment labeling on each sample text, obtain the sentiment label and sentiment degree label of the sentence in the sample text, and build a corpus labeling model through machine learning.

Based on the sentiment labels and sentiment degree labels of the clauses in the sample text, sort out related words, words, fixed collocations, and clauses in the clauses, and use techniques such as vocabulary similarity, new word discovery, and text similarity to expand existing emotions Dictionary and corpus.

Based on the corpus labeling model, a large number of sample texts are corpus labelled, and the sentiment labels and sentiment degree labels of a large number of sample texts are obtained.

Perform data statistics on the sentiment labels and sentiment degree labels of a large number of sample texts marked by the corpus, and on this basis, combined with Pareto's law or other data statistics methods to determine the sentiment monitoring degree.

After that, the results of the emotional monitoring degree were back-tested and the data revised, and the market emotion monitoring system was constructed based on the final emotional monitoring degree and the preset emotional acuity. The market sentiment monitoring system mainly includes ten levels of emotional monitoring such as worry, fear, disappointment, panic, despair, hope, optimism, peace of mind, excitement, and excitement and its six levels of emotional acuity.

Based on any of the foregoing embodiments, FIG. 2 is a schematic structural diagram of a market sentiment monitoring system establishment system provided by an embodiment of the present invention. As shown in FIG. 2, the system includes a sample acquisition unit 210, a corpus tagging unit 220, and an emotion monitoring degree acquisition unit 230 and system establishment unit 240;

Wherein, the sample obtaining unit 210 is configured to obtain a sample text set including a plurality of sample texts;

The corpus tagging unit 220 is configured to tag any of the sample texts, and obtain the emotion label and the emotion degree tag of any sample text;

The emotion monitoring degree obtaining unit 230 is configured to obtain the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text;

The system establishing unit 240 is configured to construct a market emotion monitoring system based on the emotion monitoring degree and preset emotion acuity; the market emotion monitoring system is used to monitor market emotions.

The system provided by the embodiment of the present invention obtains the corresponding emotion label and emotion degree label by labeling the sample text, and then selects the emotion monitoring degree from the emotion label of a large number of sample texts, constructs the market emotion detection manifestation, and helps to achieve objectiveness. , Accurate market sentiment monitoring helps investors reduce misjudgments and improve investment accuracy.

Based on any of the foregoing embodiments, in the system, the sample acquisition unit 210 is specifically configured to:

Determine the data collection method; the data collection method is determined according to the preset collection object, collection frequency, collection content and object level;

Based on the data collection method, the sample text is collected from the collection object through a web crawler to construct the sample text set.

Based on any of the above embodiments, the system further includes a preprocessing unit; the preprocessing unit is used to:

Preprocessing the sample text set; the preprocessing includes data cleaning and data management.

Based on any of the above embodiments, the system further includes a correlation screening unit; the correlation screening unit is used to:

Perform market relevance screening on the sample text set, and delete the sample texts whose market relevance is lower than a preset relevance threshold.

Based on any of the above embodiments, the system further includes an emotion dictionary unit; the emotion dictionary unit is specifically used for:

An emotion dictionary is constructed based on the emotion label and the emotion degree label of the sample text.

Based on any of the foregoing embodiments, the emotion monitoring degree obtaining unit 230 is specifically configured to:

Based on the sentiment label and sentiment degree label of each of the sample texts, count the number of sentences of each sentiment label and the sentiment degree label;

Obtain the emotion monitoring degree based on the number of clauses of each emotion label and emotion degree label.

Based on any of the above embodiments, the system further includes a correction unit; the correction unit is used for:

Perform result back-testing and data correction on the emotional monitoring degree.

Based on any of the foregoing embodiments, in the system, the emotion monitoring degree includes worry, fear, disappointment, panic, despair, hope, optimism, peace of mind, excitement and excitement.

FIG. 3 is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. As shown in FIG. 3, the electronic device may include: a processor 301, a communications interface 302, a memory 303, and communications The bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304. The processor 301 can call a computer program stored in the memory 303 and run on the processor 301 to execute the method for establishing a market sentiment monitoring system provided by the foregoing embodiments, for example, including: obtaining a sample text set including multiple sample texts Corpus tagging any one of the sample texts to obtain the emotion label and emotion level label of any one of the sample texts; obtaining the emotion monitoring degree based on the emotion label and emotion level label of each sample text; based on The emotion monitoring degree and the preset emotion acuity constitute a market emotion monitoring system; the market emotion monitoring system is used to monitor market emotions.

In addition, the above-mentioned logical instructions in the memory 303 can be implemented in the form of a software functional unit and when sold or used as an independent product, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention can be embodied in the form of software products in essence or parts that contribute to the prior art or parts of the technical solutions, and the computer software product is stored in a storage medium , Including several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code .

The embodiment of the present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. The computer program is implemented when executed by a processor to execute the method for establishing a market sentiment monitoring system provided by the foregoing embodiments, for example, including : Obtain a sample text set including a plurality of sample texts; perform corpus labeling on any of the sample texts, and obtain the emotion label and the emotion degree label of any sample text; the emotion label based on each sample text And the emotional degree label to obtain the emotional monitoring degree; based on the emotional monitoring degree and preset emotional acuity, a market sentiment monitoring system is constructed; the market sentiment monitoring system is used to monitor market sentiment.

The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the objectives of the solutions of the embodiments. Those of ordinary skill in the art can understand and implement it without creative work.

Through the description of the above implementation manners, those skilled in the art can clearly understand that each implementation manner can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions can be embodied in the form of software products, which can be stored in computer-readable storage media, such as ROM/RAM, magnetic A disc, an optical disc, etc., include a number of instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiment.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions recorded in the foregoing embodiments are modified, or some of the technical features are equivalently replaced; these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

A method for establishing a market sentiment monitoring system, which is characterized in that it includes:

Obtain a sample text set including multiple sample texts;

Corpus tagging any of the sample texts, and obtaining the emotion label and the emotion degree label of any sample text;

Obtaining the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text;

Based on the emotion monitoring degree and preset emotion acuity, a market emotion monitoring system is constructed; the market emotion monitoring system is used to monitor market emotions.
The method for establishing a market sentiment monitoring system according to claim 1, wherein said acquiring a sample text set including a plurality of sample texts specifically includes:

Determine the data collection method; the data collection method is determined according to the preset collection object, collection frequency, collection content and object level;

Based on the data collection method, the sample text is collected from the collection object through a web crawler to construct the sample text set.
The method for establishing a market sentiment monitoring system according to claim 1, wherein the corpus labeling of any of the sample texts to obtain the sentiment label and sentiment degree label of any of the sample texts further comprises:

Preprocessing the sample text set; the preprocessing includes data cleaning and data management.
The method for establishing a market sentiment monitoring system according to claim 1, wherein the corpus labeling of any of the sample texts to obtain the sentiment label and sentiment degree label of any of the sample texts further comprises:

Perform market relevance screening on the sample text set, and delete the sample texts whose market relevance is lower than a preset relevance threshold.
The method for establishing a market sentiment monitoring system according to claim 1, wherein the corpus labeling any one of the sample texts to obtain the sentiment label and sentiment degree label of the any sample text, and then further comprises:

An emotion dictionary is constructed based on the emotion label and the emotion degree label of the sample text.
The method for establishing a market sentiment monitoring system according to claim 1, wherein the obtaining the sentiment monitoring degree based on the sentiment label and sentiment degree label of each of the sample texts specifically comprises:

Based on the sentiment label and sentiment degree label of each of the sample texts, count the number of sentences of each sentiment label and the sentiment degree label;

The emotion monitoring degree is obtained based on the number of sentences of each emotion label and emotion degree label.
The method for establishing a market sentiment monitoring system according to claim 1, wherein said obtaining the sentiment monitoring degree based on said sentiment label and sentiment degree label of each said sample text further comprises:

Perform result back-testing and data correction on the emotional monitoring degree.
The method for establishing a market sentiment monitoring system according to any one of claims 1 to 7, wherein the sentiment monitoring degree includes worry, fear, disappointment, panic, despair, hope, optimism, peace of mind, excitement and excitement.
A system for establishing a market sentiment monitoring system, which is characterized in that it includes:

The sample acquisition unit is used to acquire a sample text set including multiple sample texts;

The corpus tagging unit is used to tag any of the sample texts to obtain the emotion label and the emotion degree tag of any sample text;

An emotion monitoring degree obtaining unit, configured to obtain the emotion monitoring degree based on the emotion label and the emotion degree label of each sample text;

The system establishment unit is configured to construct a market emotion monitoring system based on the emotion monitoring degree and preset emotion acuity; the market emotion monitoring system is used to monitor market emotions.
An electronic device, comprising a memory, a processor, and a computer program stored on the memory and running on the processor, wherein the processor executes the program as described in any one of claims 1 to 8. The steps of the method for establishing a market sentiment monitoring system are described.