News information analysis device, news information analysis program, and news information analysis method

Abstract

To prevent news information on the Internet from being browsed redundantly.SOLUTION: A news information analysis apparatus includes: a news information acquisition unit which acquires, from a communication network, multiple pieces of news information including text information and image information and stores them; a similarity determination unit which compares the text information in each news information and determines degrees of similarity between the news information; an identical news information determination unit which compares the image information in the multiple pieces of news information having degrees of similarity equal to or higher than a predetermined value, and determines, as identical news, news information in which the image information match each other; and an output unit which outputs one piece of news information out of the news information determined to be the identical news.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a news information analysis apparatus, a news information analysis program, and a news information analysis method for determining whether a plurality of pieces of news information distributed on the Internet are the same news based on text information and image information of the news information.

Background Art

[0002] Since the amount of news information distributed on the Internet is enormous, technologies for organizing news information have been developed. For example, Patent Document 1 discloses, as a news information analysis apparatus for determining the identity of a plurality of distributed pieces of news information, a technology for determining whether two pieces of news are the same based on whether the similarity calculated from independent words and numerical information in the morphological analysis result of the titles of the two pieces of news is higher than a specified value. By doing so, it is possible to avoid the inconvenience of repeatedly displaying the same news information distributed from a plurality of news sources. However, although the technology disclosed in Patent Document 1 can determine the identity of news with a certain degree of accuracy, news such as announcements of new products by companies often have similar titles, so the accuracy of identity determination decreases. In addition, in recent years, websites, web applications, and the like that purchase articles written by news organizations such as newspaper companies and post them on the company's own news portal site have become mainstream. In such cases, an article from the same news organization is posted on multiple sites, but various display methods are used, such as different ways of describing the distribution source, or posting only part of the article to encourage users to click, making it difficult to determine that the articles are the same.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0004] There is a need for a news information analysis device, news information analysis program, and news information analysis method that can determine the similarity between multiple news pieces distributed on the internet based on the text information of the news pieces, compare the image information of news pieces with a similarity score of a predetermined value or higher, and determine that they are the same news piece based on the same event if the image information matches.

[0005] This invention has been made in view of the above problems, and aims to provide a news information analysis device, a news information analysis program, and a news information analysis method that can avoid the situation of viewing news information on the internet twice by determining the similarity between news information based on the text information of multiple news information distributed on the internet, comparing the image information of news information with a similarity of a predetermined value or higher, and determining that if the image information matches, they are the same news based on the same event. [Means for solving the problem]

[0006] (1) A news information analysis device according to one aspect of the present invention (for example, the news information analysis device 1 described later) comprises: a news information acquisition unit (for example, the news information acquisition unit 110 described later) that acquires and stores a plurality of news information including text information and image information from a communication network (for example, the communication network 60 described later); a similarity determination unit (for example, the similarity determination unit 111 described later) that compares the text information of the news information and determines the similarity between the news information; a same news information determination unit (for example, the same news information determination unit 112 described later) that compares the image information of a plurality of news information whose similarity is equal to or greater than a predetermined value and determines that news information whose image information matches is the same news; and an output unit (for example, the output unit 113 described later) that outputs one of the news information determined to be the same news.

[0007] According to (1) above, it becomes possible to avoid the situation of viewing news information on the internet multiple times.

[0008] (2) In the news information analysis device described in (1) above (for example, news information analysis device 1 described later), the similarity determination unit (for example, similarity determination unit 111 described later) may determine the similarity by comparing a predetermined number of characters from the beginning of the main text portion of the text information of the news information.

[0009] According to (2) above, even news websites that display the full text by requiring users to click will be able to accurately determine duplicates.

[0010] (3) In the news information analysis device described in (1) or (2) above (for example, the news information analysis device 1 described later), the identical news information determination unit (for example, the identical news information determination unit 112 described later) may determine whether the image information matches without taking into account the number of pixels included in the image information.

[0011] According to (3) above, it is possible to accurately determine duplicates even if news sites independently resize images.

[0012] (4) In the news information analysis device described in (1) or (2) above (for example, the news information analysis device 1 described later), the identical news information determination unit (for example, the identical news information determination unit 112 described later) may determine whether the image information matches based on the recognition result of the subject in the image information.

[0013] According to (4) above, it is possible to accurately determine duplication even when news sites use their own images.

[0014] (5) In the news information analysis device described in (1) or (2) above (for example, the news information analysis device 1 described later), the output unit (for example, the output unit 113 described later) may output the news information that has the most characters in the main body portion among the news information that has been determined to be the same news.

[0015] According to (5) above, users will be able to access news information that contains the most information.

[0016] (6) In the news information analysis device described in (1) or (2) above (for example, the news information analysis device 1 described later), the output unit (for example, the output unit 113 described later) may output the news information that has the largest number of pixels included in the image information among the news information that has been determined to be the same news.

[0017] According to (6) above, users will be able to access news information that contains the most information.

[0018] (7) A news information analysis program according to one aspect of the present invention is a program that operates a computer device connected to a communication network (for example, a communication network 60 described later), and comprises: a news information acquisition function that acquires and stores a plurality of news information including text information and image information from the communication network; a similarity determination function that compares the text information of the news information and determines the similarity between the news information; a same news information determination function that compares the image information of the plurality of news information whose similarity is equal to or greater than a predetermined value and determines that each of the news information whose image information matches is the same news; and an output function that outputs one of the news information determined to be the same news.

[0019] According to (7) above, it produces the same effect as the news information analysis device described in (1) above.

[0020] (8) A news information analysis method according to an aspect of the present invention is a news information analysis method executed by a computer, comprising: a news information acquisition step of acquiring and storing a plurality of pieces of news information including text information and image information from a communication network (e.g., the communication network 60 described later); a similarity determination step of comparing text information between the pieces of news information to determine a similarity between the pieces of news information; an identical news information determination step of comparing the image information among the plurality of pieces of news information whose similarity is equal to or higher than a predetermined value, and determining each piece of news information with matching image information as identical news; and an output step of outputting one piece of news information among the pieces of news information determined as identical news.

[0021] According to the above (8), the same effect as that of the news information analysis apparatus of (1) described above is achieved. [Advantageous Effects of Invention]

[0022] According to the present invention, similarity indicating similarity between pieces of news information is determined based on text information of news information for a plurality of pieces of news information distributed on the Internet. When image information is compared for pieces of news information whose similarity is equal to or higher than a predetermined value and the image information matches, the news information is determined to be identical news based on the same event, thereby making it possible to avoid a situation where news information on the Internet is browsed repeatedly. [Brief Description of Drawings]

[0023] [Figure 1] It is a schematic diagram schematically showing a news information analysis system according to an embodiment of the present invention. [Figure 2] It is a block diagram showing a basic configuration of a news information analysis apparatus according to an embodiment of the present invention. [Figure 3] It is a flowchart showing the operation of the news information analysis apparatus in an embodiment of the present invention. [Mode for Carrying Out the Invention]

[0024] A preferred embodiment of the news information analysis device of the present invention will be described below with reference to the figures. Figure 1 shows the basic configuration of the news information analysis system 100. As shown in Figure 1, the news information analysis system 100 comprises an external system 5, a news information analysis device 1, and a communication network 60. It may also include client terminals 2 (information processing devices such as mobile devices, tablet devices, and PCs). Examples of external systems 5 include primary media sites, secondary media sites, technology and business information media sites, financial information media sites, social media sites, government agency media sites, corporate media sites, and research institute media sites. The news information analysis device 1 analyzes multiple news pieces of information distributed over the internet by an external system 5, determines the similarity between the texts of the news pieces, and for each news piece with a similarity score of a predetermined value or higher, it determines whether or not they are the same news piece based on the image information contained in the news piece, and then outputs one of the news pieces that has been determined to be the same news piece. Client terminal 2 displays news information output by news information analysis device 1 on its display unit. Client terminal 2 can, for example, prevent duplicate output of news information for the same news story. Specifically, for example, based on user instructions, news stories with high text similarity and matching images may be treated as duplicate news, and only the news information with the largest amount of text information (e.g., number of characters) or the largest image size (e.g., number of pixels) may be output. By doing so, it is possible to prevent duplicate output of news information for the same news story, and users can avoid viewing duplicate news information on the internet. The above describes the system configuration in this first embodiment. Next, the news information analysis device 1 will be described.

[0025] As shown in Figure 2, the news information analysis device 1 is composed of a control unit 11, a storage unit 12, a communication unit 13, a display unit 14, and an input unit 15.

[0026] The control unit 11 consists of a processing unit such as a microprocessor and controls each part that makes up the news information analysis device 1. Details of the control unit 11 will be described later.

[0027] The memory unit 12 is composed of semiconductor memory, SSD (Solid State Drive), HDD (Hard Disk Drive), etc., and stores various information such as control programs called firmware and operating systems, programs that make the control unit 11 function as a news information acquisition function, similarity determination function, identical news information determination function, and output function, as well as connection information necessary for connecting to the external system 5 which is the source of news information via the communication network 60, news information acquired from the external system 5, and identical news classification information for each acquired news item. Therefore, the storage unit 12 includes, in addition to the program, an external system connection information storage unit 120 containing connection information necessary for connecting to an external system 5 that will be the source of news information, and a news information storage unit 121 that stores news information acquired from each external system 5. Furthermore, the news information analysis device 1 may also include a news information management database 122 in the storage unit 12 that stores news information classified as the same news item. Note that the news information management database 122 may be configured separately from the news information analysis device 1 as a server system (news information management database system) (not shown) that is connected to the news information analysis device 1 via a communication network 6 in a manner that allows communication with the news information analysis device 1.

[0028] The external system connection information storage unit 120 is a storage unit that stores information necessary for connecting with an external system 5 that acquires news information via the communication network 60 (for example, the connection URL, authentication information, timing information for acquiring newly created and / or updated news information from the external system 5, etc.). As mentioned above, examples of external systems 5 include mass media sites, technology and business information media sites, financial information media sites, company websites, research institute websites, government agency websites, social media sites, etc. In addition to Japanese news information, news information in other languages ​​such as English, Chinese, and Korean may also be acquired. In this way, the news information acquisition unit 110, which will be described later, may refer to the external system connection information storage unit 120 to connect to each external system 5 at a predetermined timing (for example, time information, receipt of information creation / update notifications from the external system 5, etc.) and acquire new and / or updated news information provided by the external system 5.

[0029] The news information storage unit 121 is a storage unit that stores new and / or updated news information acquired by the news information acquisition unit 110 from each external system 5. The news information acquisition unit 110 may assign a news information ID to the new and / or updated news information acquired from each external system 5. This allows news information IDs representing the same news to be associated with each other in the news information management database 122, which will be described later.

[0030] The news information management database 122 is a database that registers and manages news information that has been associated as the same news by the same news information determination unit 112. The output unit 113, which will be described later, can output one piece of news information from among the news information that has been determined to be the same news based on the news information management database 122 when outputting the news information.

[0031] The memory unit 12 has been described above, but for example, the news information storage unit 121 and / or the news information management database 122 may be provided by a server device (not shown) connected to the communication network 60.

[0032] The communication unit 13 has a DSP, etc., and complies with standards such as LTE (Long Term Evolution), 4G (4th Generation), 5G (5th Generation), and Wi-Fi (registered trademark), enabling wireless and wired communication with the external system 5 via the communication network 60. The communication network 60 is realized by networks such as the Internet or a mobile phone network, or a network combining these. A LAN (Local Area Network) may also be included as part of the network.

[0033] The display unit 14 is composed of a display device such as a liquid crystal display or an organic electroluminescent panel. The display unit 14 displays an image in response to instructions from the control unit 11. The input unit 15 consists of input devices such as physical switches called a numeric keypad and a touch panel (not shown) mounted on top of the display surface of the display unit 14. As mentioned above, in addition to the configuration in which the news information analysis device 1 has an input unit 15 and a display unit 14, an information processing terminal 2 equipped with an input unit and a display unit may also be used as a client terminal for the news information analysis device 1. In this case, news information and display screens output from the news information analysis device 1 may be displayed on the information processing terminal 2 (client terminal) via the communication network 60.

[0034] Next, the details of the control unit 11 will be explained. The control unit 11 is composed of a microprocessor having a CPU, RAM, ROM, and I / O. The CPU executes each program read from the ROM or storage unit 12, and during execution, it reads information from the RAM, ROM, and storage unit 12, writes information to the RAM and storage unit 12, and exchanges signals with the communication unit 13. In this way, the processing in this embodiment is realized through the cooperation of hardware and software (programs).

[0035] The control unit 11 includes, as functional blocks, a news information acquisition unit 110, a similarity determination unit 111, a identical news information determination unit 112, and an output unit 113.

[0036] The news information acquisition unit 110 connects to each external system 5 that distributes news information (title, body text, photos, drawings, etc.) via the communication network 60 at predetermined timings based on the connection URL, authentication information, etc., and acquires newly created or updated news information distributed from each external system 5. Here, predetermined timings include, for example, a pre-set time, a pre-set time interval, or the receipt of distribution notifications from each external system 5 (including the receipt of, for example, emergency news, extra editions, etc., in addition to regularly distributed news information). Note that the news information acquisition unit 110 is not limited to automatically connecting to each external system 5 and acquiring news information. In addition to the automatic acquisition form, a user may, for example, connect to multiple external systems 5 ad-hoc via the input unit 15 or the information processing terminal 2 (client terminal) to acquire newly created or updated news information. As mentioned above, the news information distributed from the external system 5 is not limited to Japanese; for example, it may be configured to acquire news information in English, Chinese, Korean, or other languages ​​as needed by the user. In this case, the news information acquisition unit 110 may generate a Japanese translation in addition to the original language. The news information acquisition unit 110 adds information such as the site name of the external system 5 and the distribution time information (date and time) to the news information acquired from each external system 5, and stores it in the news information storage unit 121.

[0037] Before explaining the details of the function of the similarity determination unit 111, let's explain the news information. Generally, news information is written by news organizations such as newspapers and consists of concise text and images such as photographs. News information is published on the website of the news organization that wrote it, and is also sold to other news organizations and news portals, resulting in its publication on multiple websites, including those of other news organizations and news portals. Furthermore, when news is published on other news organizations' or news portal sites, the display format of the same news story may change depending on the site. For example, to display on smartphones, news information may be displayed with a smaller image size, split across multiple pages, or only the beginning of the news story may be displayed with a button prompting the user to click to read more. In this way, even for the same news story, the display format differs from site to site, so it is likely that users will often view the same news story twice. Also, it is likely that the distribution of news information for the same news story will be concentrated in close proximity (time zone). In this way, the similarity determination unit 111 can determine the similarity between news information by estimating the overlapping portion between any two news pieces of information delivered at close intervals (time zones), and based on the amount of overlap. The identical news information determination unit 112, described later, can, for example, compare the image information of news information A and news information B, whose similarity is equal to or greater than a predetermined value. If the image information matches, it can estimate that news information A and news information B are the same news. This concludes the explanation of news information. Based on the above, the function of the similarity determination unit 111 will now be explained.

[0038] As mentioned above, the same news is considered to be text with a high degree of similarity (referred to as "overlapping portion") regardless of the external system 5 from which the news is distributed. For this reason, the similarity determination unit 111 first targets news information (text) obtained from each external system 5 whose news information distribution date and time fall within a predetermined time range (referred to as the "set of similarity determination targets"), and can determine the overlapping portion between the text information of any two different news pieces included in the set of similarity determination targets (referred to as "news information A" and "news information B" for simplicity) based on the degree of similarity between the text documents. Here, for example, as a method for determining the degree of similarity between news information A (text A), where the entire text is displayed on site A, and news information B (text B), where only the first part of the text is displayed on site B, a method known to those skilled in the art may be applied, such as setting thresholds for indicators such as the number of matching characters, edit distance, or the matching rate of n-gram segmented words. For example, the similarity determination unit 111 may determine the degree of match from the matching rate of strings obtained by dividing text A and text B into n-grams in sentence order or paragraph order for a predetermined number of characters from the beginning of the main body, based on the text range of text B where only a predetermined number of characters from the start of the main body are displayed, and determine that they are a match if the degree of match exceeds or is equal to a threshold. Furthermore, the similarity determination unit 111 may determine that two texts are a match if the degree of similarity exceeds or is equal to a threshold, using the Jaccard coefficient or similar as an indicator to determine the degree of similarity of words that constitute the content of a text (e.g., nouns and verbs) from the words obtained by morphological analysis of the texts A and B in sentence order or paragraph order for a predetermined number of characters from the beginning of the main text. Alternatively, the similarity determination unit 111 may compare text information as a method for determining the degree of similarity. In this way, the similarity determination unit 111 can estimate the overlapping portion of each news piece with other news pieces based on the degree of similarity.

[0039] Next, the similarity determination unit 111 determines the similarity between the news information according to the amount of overlap. Here, the amount of overlap may be, for example, the number of characters in the overlapping text, or the proportion of the overlapping text to the total text of the news information. In this way, the similarity determination unit 111 can determine the similarity for each pair of news information included in the set of news information to be determined for similarity. The similarity determination unit 111 has been explained above. Next, the identical news information determination unit 112, which determines whether two pieces of news information are the same event news, will be explained.

[0040] The identical news information determination unit 112 compares the image information (images) of two news pieces of news included in the set of news pieces to be determined for similarity, the similarity determination unit 111 determines to be equal to or greater than a predetermined value, and determines that they are the same news if the image information (images) match. Specifically, the identical news information determination unit 112, for all combinations (pairs) of two news pieces included in the similarity determination target set, recognizes the subject matter (e.g., national flags, people, company logos, products, etc.) in the image information (images) of the two news pieces using a similarity image retrieval method known to those skilled in the art (e.g., Content-Based Image Retrieval (CBIR), etc.). Based on the recognition results, the identical news information determination unit 112 determines whether the image information (images) of the two news pieces match and whether they are the same news. Furthermore, when the identical news information determination unit 112 compares image information (images), it is preferable to compare them without considering the number of pixels, as the size of the images of news information displayed on each site differs, as mentioned above. Furthermore, if, for example, news information A or news information B is determined to be the same news as news information C, the identical news information determination unit 112 may also determine that news information A, news information B, and news information C are the same news. In this way, the identical news information determination unit 112 can classify (associate) all news information included in the set of news information to be judged for similarity into identical news information groups. In reality, this is considered to be almost impossible (given that news is often cited by web media, etc.), but if there is news information that is not similar to any other news information, that news information may be classified as having no identical news information.

[0041] The identical news information determination unit 112 may, for example, register a news information group ID for the news information that is the same news in the news information management database 122 for the news information that has been registered and stored, so that it can identify news information that has been determined to be the same news. By doing so, when a user searches the news information management database 122, it is possible to identify news information for the same news story, and for example, prevent the same news information from being searched repeatedly. The same news information determination unit 112 has now been explained.

[0042] The output unit 113 outputs one piece of news information that has been determined to be the same news by the identical news information determination unit 112 to the display unit 14 or the information processing terminal 2 (client terminal). Furthermore, when outputting news information for the same news item to the display unit 14 or the information processing terminal 2 (client terminal), the output unit 113 may output the news information that has the most characters in its main body among the news information that has been determined to be the same news item. Alternatively, when outputting news information for the same news item to the display unit 14 or the information processing terminal 2 (client terminal), the output unit 113 may output the news information that has the most pixels in its image information (image) among the news information that has been determined to be the same news item. This allows users to view news information that is most informative and from the source or closest to the source. It also prevents the display of multiple news articles covering the same topic. The configuration of each functional unit of the news information analysis device 1, which is illustrated as an example of this embodiment, has been described above.

[0043] Next, the operation of this embodiment will be described with reference to the flowchart shown in Figure 3. Figure 3 is a flowchart showing the operation of the news information analysis device 1. The process shown in Figure 3 is executed each time a processing instruction is received from the input unit 15 or the information processing terminal 2 (client terminal).

[0044] In step S10, the news information analysis device 1 (news information acquisition unit 110) connects with each external system 5 that distributes news information via the communication network 60 at a predetermined timing based on the connection URL, authentication information, etc., and acquires newly created or updated news information distributed from each external system 5.

[0045] In step S11, the news information analysis device 1 (similarity determination unit 111) targets news information (text) acquired from each external system 5 whose news information distribution date and time fall within a predetermined time range, and determines the degree of similarity between the text information of any two different news pieces included in the set.

[0046] In step S12, the news information analysis device 1 (similarity determination unit 111) determines the similarity between the two news pieces of news information according to the amount of overlap between the text information of the two pieces of news information.

[0047] In step S13, the news information analysis device 1 (identical news information determination unit 112) compares image information (images) between two news pieces of information whose similarity is equal to or greater than a predetermined value.

[0048] In step S14, the news information analysis device 1 (identical news information determination unit 112) determines (classifies or associates) two pieces of news information as the same news if the image information (images) between the two pieces of news information match.

[0049] In step S15, the news information analysis device 1 (output unit 113) outputs one news item from among the news information determined to be the same news in step S14 to the display unit 14 or the information processing terminal 2 (client terminal).

[0050] In summary, the news information analysis device 1 performs a similarity determination based on the text information of multiple news pieces distributed on the internet, compares the image information of news pieces that have a similarity of a predetermined value or higher, and if the image information matches, determines that they are the same news piece based on the same event and outputs one of the news pieces for the same event. This makes it possible for the news information analysis device 1 to avoid the situation of viewing news information on the internet twice.

[0051] Each of the devices can be implemented using hardware, software, or a combination thereof. Furthermore, the navigation method performed by the cooperation of each device included in the above-described navigation system can also be implemented using hardware, software, or a combination thereof. Here, implementation by software means implementation by a computer loading and executing a program.

[0052] Programs can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memory (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (random access memory)). Programs may also be supplied to a computer using various types of transient computer-readable media. Examples of transient computer-readable media include electrical signals, optical signals, and electromagnetic waves. Transitory computer-readable media can be supplied to a computer via wired communication channels such as electric wires and optical fibers, or via wireless communication channels.

[0053] While the embodiments described above are preferred embodiments of the present invention, the scope of the present invention is not limited to these embodiments, and various modifications can be made to implement the invention without departing from the spirit of the invention.

[0054] <Example 1> In the first embodiment described above, the news information acquisition unit 110 automatically connects to each external system 5 that distributes news information (title, body text, photographs, drawings, etc.) via the communication network 60 at a predetermined timing based on the connection URL, authentication information, etc., and acquires newly created or updated news information distributed from each external system 5. However, the embodiment is not limited to this. For example, the news information acquisition unit 110 may connect to each external system 5 that distributes news information (title, body text, photos, drawings, etc.) via the communication network 60 based on instructions from the user (for example, specifying at least one external system 5) input via the input unit 15 or the information processing terminal 2 (client terminal), based on the connection URL, authentication information, etc., and acquire newly created or updated news information distributed from each external system 5. In this way, the similarity determination unit 111, the identical news information determination unit 112, and the output unit 113 may perform the aforementioned processing on the news information from the external system 5 acquired ad hoc by the user's instructions.

[0055] <Modification 2> In the above-described embodiment, the news information analysis device 1 is implemented using a single server device, but it may also be implemented as a distributed processing system, where each function of the news information analysis device 1 is distributed among multiple server devices as appropriate. Specifically, as mentioned above, for example, the news information management database 122 may be a different server device (news information management database system) from the news information analysis device 1. Alternatively, the functions of the news information analysis device 1 may be implemented using virtual server functionality or similar features on the cloud. [Explanation of Symbols]

[0056] 100 News Information Analysis System 1. News Information Analysis Device 11 Control Unit 110 News Information Acquisition Department 111 Similarity determination unit 112 Identical News Information Determination Unit 113 Output section 12 Storage section 120 External system connection information storage unit 121 News Information Storage Unit 122 News Information Management Database 13 Communications Department 14 Display section 15 Input section 2. Information processing terminal (client terminal) 5. External Systems 60 Communication Networks

Claims

1. A news information acquisition unit that acquires and stores multiple news pieces of information, including text and image information, from a communication network. A similarity determination unit that compares the text information of the news information and determines the degree of similarity between the news information, A news information determination unit that compares the image information of multiple news pieces of news with a similarity of a predetermined value or higher and determines that news pieces of news with matching image information are the same news, An output unit that outputs one piece of news information from among the news information that has been determined to be the same news item, A news information analysis device characterized by being equipped with the following features.

2. The news information analysis device according to claim 1, characterized in that the similarity determination unit determines the similarity by comparing a predetermined number of characters from the beginning of the main text portion of the text information of the news information.

3. The news information analysis apparatus according to claim 1 or 2, characterized in that the same news information determination unit determines whether the image information matches without taking into account the number of pixels included in the image information.

4. The news information analysis apparatus according to claim 3, characterized in that the same news information determination unit determines whether the image information matches based on the recognition result of the subject in the image information.

5. The news information analysis apparatus according to claim 1 or 2, characterized in that the output unit outputs the news information with the largest number of characters in the main body portion among the news information determined to be the same news.

6. The news information analysis device according to claim 1 or 2, characterized in that the output unit outputs the news information with the largest number of pixels included in the image information among the news information determined to be the same news.

7. A program that enables the operation of computer devices connected to a communication network, A news information acquisition function that acquires and stores multiple news items, including text and image information, from a communication network. A similarity determination function that compares the text information of the aforementioned news information to determine the degree of similarity between the news information, A news information determination function that compares the image information of multiple news pieces of information with a similarity of a predetermined value or higher and determines that news pieces of information whose image information matches are the same news item, An output function that outputs one piece of news information from among the news information that has been determined to be the same news item, A news information analysis program characterized by having the following features.

8. A method for analyzing news information performed by a computer, A news information acquisition step that acquires and stores multiple news pieces of information, including text information and image information, from a communication network. A similarity determination step involves comparing the text information of the news information to determine the degree of similarity between the news information, A news information determination step in which the image information is compared among multiple news pieces of information with a similarity of a predetermined value or higher and each news piece of information whose image information matches is determined to be the same news item, An output step which outputs one news item from among the news items that have been determined to be the same news item, A news information analysis method characterized by having the following features.

Images