TWI651622B - Intelligent article summary system and method - Google Patents

Abstract

An intelligent article summary system includes a server unit and a terminal electronic device. The server unit stores a large number of financial product identification materials and a majority of articles. Each financial product identification data corresponds to an industry classification data, and each industry classification data includes at least one first industry item. The server unit semantic analysis analyzes the articles to obtain a vocabulary material that contains a plurality of component vocabularies. The server unit obtains a plurality of sentiment analysis results according to the financial product identification data, the industrial classification data and the vocabulary data, and each of the sentiment analysis results includes one of the component vocabulary and one of the first An industrial project, one of the financial product identification materials, and a public opinion sentiment, and corresponds to one of the articles.

Description

Intelligent article summary system and method

The present invention relates to an intelligent article abstraction system, and more particularly to an intelligent article abstraction system for processing unstructured text.

With the development and popularity of Internet technology, people can receive explosive growth on the Internet. Because the amount of data that can be accessed by the Internet is too large, people often want to find information about the issues and read the analysis to determine the entry point for buying and selling financial products. It often takes a lot of time, so that they miss the investment opportunity and even suffer significant loss. Therefore, how to develop a new intelligent article abstraction system can help investors to quickly clarify issues, grasp investment paradox, and make corresponding investment allocations, which will become the subject of further exploration of the present invention.

Accordingly, it is an object of the present invention to provide an intelligent article abstraction system.

Another object of the present invention is to provide a smart article abstraction method.

The intelligent article summary system of the present invention comprises a server unit and a terminal electronic device. The server unit stores a plurality of financial product identification materials, and a plurality of articles related to a topic, each of the financial product identification materials corresponding to an industry classification data, each of the industrial classification materials includes At least one first industry project. The terminal electronics can communicate with the server unit.

The server unit semantic analysis analyzes the articles to obtain a vocabulary material that contains a plurality of component vocabularies. The server unit obtains a plurality of sentiment analysis results based on the financial product identification data, the industry classification data, and the vocabulary data, and each of the sentiment analysis results includes the vocabulary data. One of the component vocabulary, one of the at least one first industry project of one of the industry classification materials, one of the financial product identification materials, and one of the public opinion emotions, and corresponding to the articles One.

The server unit transmits the sentiment analysis results to the terminal electronic device. The terminal electronic device displays at least the component vocabulary, the financial product identification data, and the public opinion sentiment of each of the sentiment analysis results.

When the terminal electronic device receives an article reading instruction related to one of the target of the sentiment analysis results, the terminal electronic device displays the article corresponding to the target sentiment analysis result.

When the terminal electronic device receives a transaction instruction related to one of the target of the sentiment analysis results, the terminal electronic device displays a transaction interface of the financial product identification data related to the target sentiment analysis result.

In some implementations, the terminal electronic device displays at least the component vocabulary, the financial item identification data, and the public opinion mood for each of the sentiment analysis results in a sunburst map.

In some implementations, the article reading instruction is a click instruction associated with the one of the plurality of blocks of the sunburst map that corresponds to the target sentiment analysis result.

In some implementations, the transaction instruction is a click instruction associated with one of the blocks of the sunburst map that corresponds to the financial item identification data of the target sentiment analysis result.

In some implementations, the public opinion mood indicates forward, reverse, or flat.

In some implementations, each of the industry classification materials further includes at least one second industry item, the at least one first industry item and the at least one second industry item of each of the industry classification materials There is a parent-child classification relationship, and each of the sentiment analysis results includes one of the at least one second industry project that further includes one of the industry classification materials.

In some implementations, the vocabulary data further includes a commodity vocabulary, each of the sentiment analysis results further including the commodity vocabulary of the vocabulary material.

The smart article abstract method of the present invention is implemented by a smart article abstracting system, which comprises a server unit and a terminal electronic device, wherein the server unit stores a plurality of financial product identification materials, and most of the pens are related to An article of the subject, each of the financial product identification materials corresponding to an industry classification data, each of the industry classification materials comprising at least one first industry item, the terminal electronic device capable of communicating with the server unit The method includes: the server unit semantically analyzing the articles to obtain a vocabulary data, the vocabulary data comprising a plurality of component vocabulary; the server unit according to the financial product identification data, the industry classification data, and the vocabulary data Emotional analysis of the articles to obtain a plurality of sentiment analysis results, each of the sentiment analysis results including one of the component vocabulary words of the vocabulary data, the at least one of the ones of the industry classification materials One of the industrial projects, one of these financial product identification materials, and a glimpse of the situation And corresponding to one of the articles; the server unit transmits the sentiment analysis results to the terminal electronic device; the terminal electronic device displays at least the component vocabulary of each of the sentiment analysis results, the financial Commodity identification data and the public opinion emotion; when the terminal electronic device receives an article reading instruction related to one of the target objects of the sentiment analysis results, the terminal electronic device displays the article corresponding to the target sentiment analysis result; and when the When the terminal electronic device receives the transaction instruction related to one of the target of the sentiment analysis results, the terminal electronic device displays a transaction interface of the financial product identification data related to the target sentiment analysis result.

The effect of the present invention is that the article is semantically analyzed by the server unit, and the server unit analyzes the articles according to the financial product identification data, the industrial classified materials, and the vocabulary data. Emotional analysis results clearly show the relationship between financial products and the supply chain of popular goods and public opinion sentiment, and facilitate users to quickly find articles related to their concerns, and facilitate users to immediately trade financial products, in addition, by The results of the sentiment analysis are displayed in the sunburst chart, so that the sentiment analysis results are visualized and displayed for the user to read and understand.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are denoted by the same reference numerals.

Referring to FIG. 1, a first embodiment of the smart article summary system 100 of the present invention includes a server unit 1 and a terminal electronic device 2.

The terminal electronics 2 can communicate with the server unit 1 via a communication network. The terminal electronic device 2 is, for example, a smart phone, a tablet computer, a desktop computer or a laptop computer, but is not limited thereto.

The server unit 1 stores a plurality of pen financial product identification materials D1, and a plurality of articles related to a topic D3. Each of the financial product identification data D1 includes a company name and a financial product code, such as "TSMC_2330". Each of the financial product identification materials D1 corresponds to an industry classification data D2. Each of the industry classification materials D2 includes at least one first industry project and at least one second industry project. The at least one first industry item of each of the industry classification materials D2 and the at least one second industry item have a child-child classification relationship. For example, one of the industry classification data D2 includes four primary industry projects and one secondary industry project, such as small and medium size panels, OLED, TFT-LCD, and touch panel. The second industry project is, for example, the panel industry.

The articles D3 are, for example, unstructured texts collected from the Internet, such as research reports, news, and the like. In this embodiment, the subject related to the article D3 is taken as an example of the "iPhone 8 supply chain", but is not limited thereto.

Referring to Figures 1 and 2, the steps of a smart article digest method performed by the smart article abstraction system 100 are described below. First, as shown in step S01, the server unit 1 semantically analyzes the articles D3 to obtain a vocabulary data, the vocabulary data including a plurality of component vocabularies, and a commodity vocabulary. For example, the product vocabulary is, for example, "iPhone 8", such as "carrier board", "electroacoustic component", "A11 processor", "control chip", "NOR Flash" and " Words like 3D touch.

Next, as shown in step S02, the server unit 1 obtains a plurality of sentiment analysis results based on the financial product identification data D1, the industrial classification data D2, and the vocabulary data, and emotionally analyzing the articles D3. Each of the sentiment analysis results includes the commodity vocabulary of the vocabulary material, one of the component vocabulary words of the vocabulary material, and the at least one first industry item of one of the industry classification materials D2 One of the at least one second industry project of one of the industry classification data D2, one of the financial product identification materials D1, and one of the public opinion emotions, and corresponding to one of the articles D3 By. This public opinion sentiment indicates positive, negative or flat.

Next, as shown in step S03, the server unit 1 transmits the sentiment analysis result to the terminal electronic device 2.

Then, as shown in step S04, the terminal electronic device 2 displays at least the component vocabulary, the financial product identification data D1 and the public opinion emotion of each of the sentiment analysis results. In this embodiment, the terminal electronic device 2 displaying the commodity vocabulary of each of the sentiment analysis results, the component vocabulary, the first industry item, the second industry item, the financial product identification data D1, and the public opinion mood.

As shown in FIG. 3, the terminal electronic device 2 displays the product vocabulary, the component vocabulary, the first industry project, and the second industry for each of the sentiment analysis results by a Sunburst Diagram 200. Project, the financial product identification data D1 and the public opinion mood. In this embodiment, the center of the sunburst chart 200 presents the product vocabulary, and the component vocabulary, the first industry item, the second industry item, the financial product identification data D1, and the The public opinion mood.

Next, as shown in step S05, the terminal electronic device 2 determines whether an article reading instruction related to one of the target persons of the sentiment analysis results is received, and if yes, executes step S06, and if not, executes step S07.

Step S06 is when the terminal electronic device 2 receives an article reading instruction related to one of the target of the sentiment analysis results, the terminal electronic device 2 displays the article D3 corresponding to the target sentiment analysis result. In this embodiment, the article reading instruction is a click instruction related to the public opinion emotion of one of the plurality of blocks of the sunday chart 200 corresponding to the target sentiment analysis result. For example, if the user of the terminal electronic device 2 wants to read "IPhone 8", "Trayboard", "Printed Circuit Board Related", "Printed Circuit Board", "Jingshuo_3189", "Forward" In the article D3 corresponding to the sentiment analysis result, the user only clicks on the "IPhone 8", "Trayboard", "Printed Circuit Board Related", "Printed Circuit Board", "Jingshuo_3189" To the block, the terminal electronic device 2 displays the article D3 corresponding to the sentiment analysis result for the user to read. In this way, the user can quickly find the article D3 related to the topic of interest.

Step S07 is that the terminal electronic device 2 determines whether a transaction instruction related to one of the target persons of the sentiment analysis results is received, and if yes, executes step S08, and if not, executes step S05 again.

Step S08 is a transaction in which the terminal electronic device 2 displays the financial item identification data D1 related to the target sentiment analysis result when the terminal electronic device 2 receives the transaction instruction related to one of the target of the sentiment analysis results. interface. In this embodiment, the transaction instruction is a click instruction related to one of the blocks of the sunday chart 200 corresponding to the financial item identification data D1 of the target sentiment analysis result. For example, if the user of the terminal electronic device 2 has read "IPhone 8", "Trayboard", "Printed Circuit Board Related", "Printed Circuit Board", "Jing Shuo_3189", "正正" After the article D3 corresponding to the sentiment analysis result, if you want to buy the financial products of Jingshuo, the user will click on the block of "Jingshuo_3189", and the terminal electronic device 2 will display the financial products of Jingshuo. The trading interface is for users to place orders. Thereby, it is convenient for the user to trade the financial products.

Referring to FIG. 4, the second embodiment of the smart article abstraction system 100 of the present invention is similar to the first embodiment except that the terminal electronic device 2 displays only the results of the sentiment analysis results when displaying the sentiment analysis results. The commodity vocabulary, the component vocabulary, the financial commodity identification data D1, and the public opinion mood. As shown in FIG. 4, the center of the sunburst chart 200 also presents the commodity vocabulary, and the financial product identification data D1, the component vocabulary and the public opinion mood are sequentially presented from the center to each ring.

Referring to FIG. 5, the third embodiment of the smart article abstraction system 100 of the present invention is similar to the first embodiment except that the terminal electronic device 2 displays only the results of the sentiment analysis results when displaying the sentiment analysis results. Financial commodity identification data D1, the component vocabulary, the public opinion mood and the primary industry project. As shown in FIG. 5, the center of the sunburst chart 200 presents the financial product identification data D1, and the component vocabulary, the public opinion mood, and the first industry item are sequentially presented from the center to each ring. It is worth noting that in the third embodiment, each of the sunburst charts 200 presents only sentiment analysis of the same financial product identification data D1 (for example, "TSMC_2330") and the same component vocabulary (for example, "A11 processor"). As a result, the user sentiment analysis result is provided for the single financial item identification data D1 and the single component vocabulary.

In summary, the smart article abstraction system 100 of the present invention analyzes the articles D3 by the server unit 1 in a semantic manner, and the server unit 1 uses the financial product identification data D1 and the industrial classified data D2 and The lexical data sentiment analysis of the articles D3 makes the sentiment analysis results clearly show the relationship between the financial products and the hot commodity supply chain and the public opinion sentiment, and facilitates the user to quickly find the article D3 related to the topic of interest, and is convenient The user immediately conducts the transaction of the financial product. In addition, by displaying the sentiment analysis results in the sunburst chart 200, the sentiment analysis results are visually presented, which is convenient for the user to read and understand, so that the present invention can be achieved. purpose.

However, the above is only the embodiment of the present invention, and the scope of the invention is not limited thereto, and all the simple equivalent changes and modifications according to the scope of the patent application and the patent specification of the present invention are still Within the scope of the invention patent.

100‧‧‧Intelligent Article Summary System

1‧‧‧Server unit

D1‧‧‧ Financial product identification information

D2‧‧‧Industry Classification Information

D3‧‧‧ article

2‧‧‧Terminal electronic devices

200‧‧‧Rising Sun

S01~S08‧‧‧Steps

Other features and effects of the present invention will be apparent from the following description of the drawings, wherein: FIG. 1 is a schematic diagram of a hardware connection relationship of a first embodiment of the intelligent article abstraction system of the present invention; FIG. 3 is a sunburst view of the first embodiment; FIG. 4 is a sunburst view of a second embodiment of the smart article abstract system of the present invention; and FIG. 5 is an intelligent view of the present invention A sunburst chart of a third embodiment of the article summary system.

Claims (14)

An intelligent article summary system, comprising: a server unit storing a plurality of financial product identification materials, and an article related to a topic of a plurality of pens, each of the financial product identification materials corresponding to an industry classification data, Each of the industry classification materials includes at least one first industry item; and a terminal electronic device capable of communicating with the server unit; the server unit semantically analyzing the articles to obtain a vocabulary data, the vocabulary data comprising a plurality of a component vocabulary; the server unit obtains a plurality of sentiment analysis results according to the financial product identification data, the industrial classification data, and the vocabulary data, and each of the sentiment analysis results includes the vocabulary One of the component vocabularies of the data, one of the at least one first industry project of one of the industry classification materials, one of the financial product identification materials, and one of the public opinion emotions, and corresponding to One of the articles; the server unit transmits the sentiment analysis results to the terminal electronic device The terminal electronic device displays at least the component vocabulary, the financial product identification data, and the public opinion emotion of each of the sentiment analysis results; when the terminal electronic device receives one of the target persons related to the sentiment analysis results When the article reads the instruction, the terminal electronic device displays the article corresponding to the target sentiment analysis result; when the terminal electronic device receives the transaction instruction related to one of the target of the sentiment analysis results, the terminal electronic device displays A trading interface of the financial product identification data of the target sentiment analysis result. The smart article summary system of claim 1, wherein the terminal electronic device displays at least the component vocabulary, the financial product identification data, and the public opinion mood of each of the sentiment analysis results in a sunburst map. The smart article abstraction system of claim 2, wherein the article reading instruction is a click instruction related to the public opinion emotion of one of the plurality of blocks of the sunburst map corresponding to the target sentiment analysis result. The smart article summary system of claim 2, wherein the transaction instruction is a click instruction related to the financial item identification material corresponding to the target sentiment analysis result of one of the plurality of blocks of the sunburst chart . The smart article abstraction system of claim 1, wherein the public opinion mood indicates forward, reverse or flat. The smart article summary system of claim 1, wherein each of the industry classification materials further comprises at least one second industry item, the at least one first industry item of each of the industry classification materials The at least one second industry item has a child-child classification relationship, and each of the sentiment analysis results includes one of the at least one second industry items further including one of the industry classification materials. The smart article abstraction system of claim 1, wherein the vocabulary data further comprises a commodity vocabulary, each of the sentiment analysis results further comprising the commodity vocabulary of the vocabulary data. An intelligent article abstraction method is implemented by an intelligent article abstraction system, which comprises a server unit and a terminal electronic device, wherein the server unit stores a plurality of financial product identification data, and a plurality of pens are associated with one In the subject matter, each of the financial product identification materials corresponds to an industry classification data, each of the industry classification materials includes at least one first industry item, and the terminal electronic device can communicate with the server unit. The method includes: the server unit semantically analyzing the articles to obtain a vocabulary data, the vocabulary data comprising a plurality of component vocabulary; the server unit according to the financial product identification data, the industry classification data and the vocabulary data, Emotional analysis of the articles obtains a plurality of sentiment analysis results, each of the sentiment analysis results including one of the component vocabulary words of the vocabulary data, the at least one of the ones of the industry classification materials One of the industrial projects, one of the financial product identification materials, and a glimpse of the situation And corresponding to one of the articles; the server unit transmits the sentiment analysis results to the terminal electronic device; the terminal electronic device displays at least the component vocabulary of each of the sentiment analysis results, the financial Commodity identification data and the public opinion emotion; when the terminal electronic device receives an article reading instruction related to one of the target persons of the sentiment analysis results, the terminal electronic device displays the article corresponding to the target sentiment analysis result; and when the When the terminal electronic device receives the transaction instruction related to one of the target of the sentiment analysis results, the terminal electronic device displays a transaction interface of the financial product identification data related to the target sentiment analysis result. The smart article summary method of claim 8, wherein the terminal electronic device displays at least the component vocabulary, the financial product identification data, and the public opinion mood of each of the sentiment analysis results in a sunburst map. The smart article abstraction method according to claim 9, wherein the article reading instruction is a click instruction related to the public opinion emotion of one of the plurality of blocks of the sunburst map corresponding to the target sentiment analysis result. The smart article summary method of claim 9, wherein the transaction instruction is a click instruction related to the financial item identification material corresponding to the target sentiment analysis result of the plurality of blocks of the sunburst chart . The smart article summary method of claim 8, wherein the public opinion mood indicates forward, reverse or flat. The smart article summary method of claim 8, wherein each of the industry classification materials further comprises at least one second industry item, the at least one first industry item of each of the industry classification materials The at least one second industry item has a child-child classification relationship, and each of the sentiment analysis results includes one of the at least one second industry items further including one of the industry classification materials. The smart article summary method of claim 8, wherein the vocabulary data further comprises a commodity vocabulary, each of the sentiment analysis results further comprising the commodity vocabulary of the vocabulary data.