Method and system for providing news content in a digital environment

Abstract

The present invention provides a method for providing news content, comprising collecting news content, separating a plurality of sentences included in the news content according to sentence units, clustering the separated plurality of sentences to generate a plurality of news clusters, calculating a weight for each of the plurality of news clusters based on a plurality of sentences belonging to each of the plurality of news clusters, and providing the plurality of news clusters based on the calculated weight.

Description

Technology Field

[0001] The present invention relates to a method and system for providing news content in a digital environment. Background Technology

[0002] With the advancement of communication services, news content, which was traditionally provided in the form of broadcasts and newspapers, is gradually shifting to a format where it is delivered through portals or media company websites.

[0003] Accordingly, various portal sites classify and provide news content produced by diverse media outlets by category, and along with the production of news content, an increasingly large amount of news content is being produced and provided due to the characteristics of the online environment, which allows for real-time access.

[0004] However, in this process, numerous news contents containing the same information are being produced redundantly, and many news contents with biased political leanings or ideologies are also being produced.

[0005] Accordingly, consumers of news content endure the inconvenience of cross-verifying multiple news contents containing the same information or cross-verifying news contents produced by media outlets with different political leanings in order to determine the authenticity of the information included in the news content. The problem to be solved

[0006] The present invention relates to a method and system for providing news content in a digital environment.

[0007] In addition, the present invention relates to a method and system for collecting news content, dividing the collected news content into sentence units, and clustering a plurality of divided sentences to provide news content.

[0008] In addition, the present invention relates to a method and system for providing news content that calculates the importance of each cluster according to the number of clustered sentences from news content and provides each cluster according to the calculated importance.

[0009] In addition, the present invention relates to a method and system for providing news content that lists clustered sentences from news content in chronological order. means of solving the problem

[0010] To solve the problem described above, a method for providing news content according to the present invention may include: a step of collecting news content; a step of separating a plurality of sentences included in the news content according to sentence units; a step of clustering the separated plurality of sentences to generate a plurality of news clusters; and a step of calculating a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, and providing the plurality of news clusters based on the calculated weight.

[0011] In addition, the news content providing system according to the present invention includes: a communication unit connected to a server in which news content is stored; and a control unit that collects the news content from the server, separates a plurality of sentences included in the news content according to sentence units, and clusters the separated plurality of sentences to generate a plurality of news clusters, wherein the control unit calculates a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, and can provide the plurality of news clusters based on the calculated weight.

[0012] In addition, a program stored on a computer-readable recording medium according to the present invention is executed by one or more processes in an electronic device and is a program stored on a computer-readable recording medium, wherein the program may include instructions for performing the steps of: collecting news content; separating a plurality of sentences included in the news content according to sentence units; clustering the separated plurality of sentences to generate a plurality of news clusters; and calculating a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, and providing the plurality of news clusters based on the calculated weight. Effects of the invention

[0013] According to various embodiments of the present invention, a news content provision method and system can effectively provide the content of an event commonly covered in different news content by collecting news content, dividing the collected news content into sentence units, and clustering a plurality of divided sentences.

[0014] In addition, according to various embodiments of the present invention, a news content provision method and system can effectively provide the event of the highest interest among the various events covered in the news content by calculating the importance of each cluster based on the number of clustered sentences from the news content and providing each cluster according to the calculated importance.

[0015] In addition, according to various embodiments of the present invention, a method and system for providing news content can effectively provide a process in which a specific event covered in news content changes over time by arranging and providing clustered sentences from news content in chronological order. Brief explanation of the drawing

[0016] FIG. 1 illustrates a news content provision system according to the present invention. FIGS. 2 to 4 illustrate an embodiment in which news content is provided to a user. FIG. 5 is a flowchart illustrating a method for providing news content according to the present invention. FIG. 6 illustrates an example of clustering multiple sentences included in news content. FIGS. 7 to 10 illustrate an embodiment providing a news cluster. Specific details for implementing the invention

[0017] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are assigned the same reference number regardless of the drawing symbols, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not have distinct meanings or roles in themselves. Furthermore, in describing the embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the present invention.

[0018] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0019] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0020] A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0021] In this application, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0022] FIG. 1 illustrates a news content provision system according to the present invention. FIGS. 2 to 4 illustrate an embodiment in which news content is provided to a user.

[0023] Referring to FIG. 1, a news content providing system (100) according to the present invention collects news content (1), separates a plurality of sentences included in the news content (1) into sentence units, clusters the separated plurality of sentences to generate a plurality of news clusters (3), and provides the generated plurality of news clusters (3) so that a user can check them.

[0024] Here, news content (1) is content produced and distributed by a company or institution such as a media company, and may include information related to various categories such as society, economy, politics, culture, sports and entertainment.

[0025] These news contents (1) may include text data and image data, and in particular, the text data may include information related to multiple sentences entered by a media company.

[0026] Therefore, it can be understood that the news content (1) includes multiple sentences, and processing these multiple sentences can be understood as processing the text data included in the news content (1).

[0027] Additionally, the news content (1) may further include information related to the time when the news content (1) is produced and posted on a news content server or web server provided by a media company or portal, etc. (or the time of production), and information related to the category to which the news content (1) belongs.

[0028] Meanwhile, the news cluster (3) is formed by clustering multiple sentences separated from multiple news contents (1) according to the similarity between the sentences, and each news cluster (3) may include one or more sentences separated from the news contents (1).

[0029] At this time, the news cluster (3) includes information related to a core keyword representing the topic of the sentences belonging to the news cluster (3), and each news content (1) may include an index, which is information related to the location of the sentences belonging to the news cluster (3).

[0030] Additionally, the news cluster (3) may include information related to the time of publication (or the time when the news cluster was created) of the news content (1) containing the sentence belonging to the news cluster (3), and may also include an access path (e.g., a link) to the news content (1) containing the sentence belonging to the news cluster (3).

[0031] Accordingly, the news content provision system (100) may prioritize providing major news clusters (3) based on the number of sentences in each news cluster (3), or classify and provide news clusters (3) related to specific topics based on key keywords for each news cluster (3).

[0032] Additionally, the news content provision system (100) may provide multiple news clusters (3) arranged in the order in which the sentences were written based on an index for each news cluster (3), or may provide multiple news clusters (3) arranged in the order in which each news cluster (3) was created (or the order in which the news content (1) containing the sentences belonging to each news cluster (3) was posted).

[0033] Referring to FIG. 2, in one embodiment, a news content providing system (100) can effectively provide issues that occurred on a given day to a user by using news content collected on a specific day (e.g., daily) to create a plurality of news clusters and providing the created clusters so that the user can view them (e.g., sub-topic view).

[0034] At this time, the news content provision system (100) may select and provide multiple news clusters corresponding to a specific issue based on core keywords belonging to each news cluster.

[0035] In this regard, referring to FIG. 3, each news cluster includes information related to a core keyword (e.g., Subtopic title) representing the topic of a sentence belonging to the news cluster, and sentences belonging to the news cluster (e.g., Sentence). The subtopic view provided to the user can be implemented to provide sentences belonging to a specific news cluster selected according to user input, along with a core keyword corresponding to each news cluster.

[0036] Referring again to FIG. 2, in another embodiment, a news content providing system (100) can generate a plurality of news clusters using news content collected over an entire period (or a predetermined period (e.g., the last 10 years)) (e.g., Timeline), and among the generated plurality of news clusters, select a plurality of news clusters corresponding to a specific issue based on a core keyword belonging to each news cluster and provide them so that a user can verify them (e.g., news article timeline view), thereby providing the user with major content regarding a specific issue that occurred during the entire period.

[0037] In this regard, referring to FIG. 4, each news cluster includes information related to a core keyword representing the topic of a sentence belonging to the news cluster (e.g., Subtopic title), sentences belonging to the news cluster (e.g., Sentence), and information related to the date on which news content corresponding to each sentence was produced (or published) (e.g., date), and the news article timeline view provided to the user can be implemented so that core keywords corresponding to each date's news cluster in relation to a specific issue are provided.

[0038] Referring again to FIG. 1, the news content providing system (100) according to the present invention may include a communication unit (110), a storage unit (120), a control unit (130), and an output unit (140).

[0039] The communication unit (110) can be connected via a wired or wireless network to a news content server or web server equipped with a media company or a portal, etc. Accordingly, the communication unit (110) can receive news content (1) from the news content server or web server.

[0040] Alternatively, the communication unit (110) may receive information related to a website provided on a news content server or a web server. Through this, the control unit (130) may collect news content (1) by crawling the news content (1) from a website provided on a news content server or a web server through the communication unit (110).

[0041] The storage unit (120) may store commands and data for performing the operation of the news content provision system (100) according to the present invention. For example, the storage unit (120) may store news content (1) collected through the communication unit (110). Additionally, the storage unit (120) may store information related to a plurality of sentences separated from the news content (1) and a plurality of news clusters (3) in which the plurality of sentences are clustered.

[0042] The control unit (130) can control the overall operation of the news content provision system (100) according to the present invention. For example, the control unit (130) can collect news content (1), separate the news content (1) into a plurality of sentences, and cluster the separated plurality of sentences to generate a plurality of news clusters (3).

[0043] Accordingly, the control unit (130) can classify and process a plurality of news clusters (3) according to predetermined conditions and provide them for the user to view. At this time, providing a plurality of news clusters (3) can be understood as providing the results of classifying and processing the plurality of news clusters (3) (e.g., graphs, shapes, text, etc.) so that the user can visually view them.

[0044] The output unit (140) can output a plurality of news clusters (3) generated by the control unit (130). At this time, the output unit (140) can output the plurality of news clusters (3) in various forms such as graphs, text, and shapes.

[0045] Based on the configuration of the news content provision system (100) examined above, the method of providing news content will be explained in more detail below.

[0046] FIG. 5 is a flowchart illustrating a method for providing news content according to the present invention. FIG. 6 illustrates an embodiment of clustering a plurality of sentences included in news content. FIG. 7 to 10 illustrate an embodiment of providing news clusters.

[0047] Referring to FIG. 5, the news content providing system (100) according to the present invention collects news content (S100) and can separate a plurality of sentences included in the news content according to sentence units (S200).

[0048] Specifically, the news content provision system (100) can collect news content from a pre-equipped news content server or a web server in which multiple news contents are stored.

[0049] For example, the news content provision system (100) can collect news content by crawling a website provided by a web server equipped to provide multiple news contents. In this case, the news content provision system (100) may collect news content by category related to the news content provided on the website.

[0050] As another example, the news content provision system (100) may receive news content from a media server equipped to produce and publish multiple news contents, or collect news content by crawling a website provided by the media server. In this case, the news content provision system (100) may collect news content by category related to news content, and may also collect news content for each of multiple different media servers.

[0051] Furthermore, the news content provision system (100) can obtain multiple sentences included in each of the multiple news contents by separating each of the multiple news contents, which are composed of multiple sentences, into sentence units.

[0052] For example, a news content provision system (100) can extract text data containing multiple sentences from each of multiple news contents and perform a series of sentence separation algorithms on the extracted text data to obtain multiple sentences contained in each of the multiple news contents.

[0053] At this time, the sentence separation algorithm can be implemented to separate multiple sentences based on punctuation marks (e.g., periods, quotation marks, parentheses, etc.) or to separate multiple sentences based on terminal endings and terminal marks recognized through morphological analysis.

[0054] In one embodiment, the sentence separation algorithm may utilize algorithms such as KIWI, OKT (Open Korean Text), and KoalaNLP (Korean Scala NLP (Natural Language Processing)). Additionally, the sentence separation algorithm may utilize different algorithms depending on the language provided in the news content, and a translation tool for translating the language provided in the news content into another language may also be used.

[0055] Furthermore, the news content provision system (100) can assign an index to each of the multiple sentences separated from the specific news content based on the position of each sentence in the specific news content.

[0056] For example, the news content providing system (100) may assign an index to each of the multiple sentences separated from each news content according to the order of the sentences in the news content. That is, in one embodiment, the news content providing system (100) may assign '0' as an index to the first sentence recognized first in the news content, and assign '1' as an index to the second sentence recognized following the sentence assigned index '0'.

[0057] As another example, the news content provision system (100) may perform a filtering operation to be described later on each of the multiple sentences separated from each news content, and may assign an index to each of the multiple sentences based on the order of the filtered multiple sentences.

[0058] A news content providing system (100) according to the present invention may cluster a plurality of separated sentences to generate a plurality of news clusters (S300), calculate a weight for each of the plurality of news clusters based on a plurality of sentences belonging to each of the plurality of news clusters, and provide the plurality of news clusters based on the calculated weight (S400).

[0059] Specifically, the news content provision system (100) can perform filtering by removing sentences that have a predetermined keyword corresponding to a unique sentence of the news content from among a plurality of sentences separated from each of the plurality of news contents.

[0060] For example, the news content provision system (100) can analyze each of the multiple sentences separated from the news content and extract one or more keywords corresponding to each sentence. Accordingly, the news content provision system (100) can compare the one or more keywords extracted above with a predetermined keyword (e.g., 'date, news' or 'name, reporter', etc.) that corresponds to the unique sentence of the news content, and filter by removing the sentence corresponding to the unique sentence of the news content from among the multiple sentences based on the comparison result.

[0061] According to an embodiment, the news content providing system (100) can filter by removing sentences corresponding to unique sentences of the news content, such as 'This is the news dated 0000 / 0 / 0' and 'This has been reporter 000', among a plurality of sentences separated from the news content.

[0062] As another example, the news content provision system (100) may filter by removing the sentence corresponding to the unique sentence of the news content through semantic analysis of each of the multiple sentences separated from the news content.

[0063] In this case, the news content provision system (100) may use sentence similarity measurement algorithms such as NLU (Natural Language Understanding), SBERT (Sentence-Bidirectional Encoder Representations from Transformers), and STS (Semantic Textual Similarity), depending on the embodiment.

[0064] Accordingly, the news content provision system (100) can compare a unique sentence of the news content with each of the multiple sentences separated from the news content, and if the similarity based on the comparison result exceeds a predetermined threshold, it can remove and filter the sentence.

[0065] Furthermore, the news content provision system (100) can cluster multiple sentences separated from news content into multiple news clusters based on the similarity between different sentences, and provide the multiple news clusters so that the user can check them.

[0066] For example, a news content provision system (100) can compare the similarity between multiple sentences separated from news content and, based on the comparison result, cluster sentences whose similarity exceeds a predetermined threshold into one news cluster.

[0067] At this time, the news content provision system (100) can measure the similarity between multiple sentences based on the sentence similarity measurement algorithm described above, and cluster sentences whose similarity exceeds a threshold into one news cluster according to the measurement result.

[0068] Accordingly, the news content provision system (100) may include one or more sentences separated from different news content in each of the multiple news clusters.

[0069] Referring to FIG. 6, as another example, the news content providing system (100) may separate sentences 1 to 10 from the first news content (11) and sentences a to j from the second news content (12).

[0070] Accordingly, the news content provision system (100) can obtain sentences 1 to 10 and sentences a to j, which are multiple sentences from multiple news contents (11, 12), and sequentially measure the similarity for each of the multiple sentences to generate multiple news clusters (21, 22, 23).

[0071] According to an embodiment, the news content providing system (100) can sequentially compare a plurality of sentences based on an index assigned to each sentence. That is, the news content providing system (100) can compare sentence 1 with sentence a, then compare sentence 1 with sentence 2 and sentence b, and then compare sentence 1 with sentence 3 and sentence c.

[0072] Through this, the news content provision system (100) can generate a first cluster (21) to which other sentences similar to sentence 1 (e.g., sentence 2) belong based on the similarity calculated according to the comparison result, and can perform clustering again on the remainder of the multiple sentences that do not belong to the first cluster (21).

[0073] Accordingly, the news content provision system (100) can compare sentence a and sentence b, then compare sentence a with sentence 3 and sentence c, then compare sentence a with sentence 4 and sentence d, and then compare sentence a with sentence 5 and sentence e.

[0074] Through this, the news content provision system (100) can generate a second cluster (22) to which other sentences similar to sentence a (e.g., sentence 4) belong based on the similarity calculated according to the comparison result, and can perform clustering again on the remainder of the multiple sentences that do not belong to the second cluster (22).

[0075] The news content provision system (100) can repeat the above process to generate multiple sentence clusters (21, 22, 23) to which multiple sentences separated from multiple news contents (11, 12) belong.

[0076] As another example, the news content provision system (100) can compare the similarity between multiple sentences separated from news content and generate multiple news clusters based on the comparison result.

[0077] Accordingly, the news content provision system (100) can filter by removing news clusters among the previously generated multiple news clusters in which the number of sentences belonging to each news cluster is less than a predetermined limit value (e.g., 1 or 2).

[0078] As another example, the news content provision system (100) can generate multiple tokens for each of the multiple sentences separated from the news content.

[0079] To this end, the news content provision system (100) can divide each of the multiple sentences into morphemes or words and convert each divided morpheme or word into a predetermined form to generate multiple tokens.

[0080] Accordingly, the news content provision system (100) can compare the similarity between sentences based on multiple tokens divided from each of the multiple sentences, and generate multiple news clusters based on the comparison result.

[0081] Furthermore, the news content provision system (100) classifies the previously collected multiple news contents according to predetermined conditions, generates multiple news clusters based on the multiple news contents classified according to the conditions, and can provide the multiple news clusters generated according to the conditions so that a user can check them.

[0082] For example, a news content provision system (100) can collect and classify multiple news contents by media company (or media company server) and generate multiple news clusters by media company for each of the classified multiple news contents.

[0083] Through this, the news content provision system (100) can provide news clusters so that the differences in topics that are important to each media outlet can be confirmed by comparing the differences in news clusters generated for each media outlet, or it can provide news clusters so that the reliability of information provided by a media outlet can be confirmed by listing the news clusters generated for a specific media outlet in chronological order.

[0084] As another example, the news content provision system (100) can classify multiple news contents by category provided by a media company or website (e.g., economy, politics, culture, sports, entertainment, etc.) and generate multiple news clusters for each of the classified multiple news contents.

[0085] Through this, the news content provision system (100) can provide news clusters by sorting the news clusters generated by category according to the order of the magnitude of the weights calculated for each news cluster, so that topics of high interest by category can be identified.

[0086] Furthermore, the news content provision system (100) can calculate weights based on the number of sentences belonging to each of the multiple news clusters, and sort the multiple news clusters according to the order of the calculated weights so that the user can check them.

[0087] For example, the news content provision system (100) can calculate the number of sentences belonging to each of the multiple news clusters and determine a weight based on the number of calculated sentences.

[0088] That is, the news content provision system (100) can provide the news cluster to which the sentences are most numerous so that the user can check it first. Additionally, the news content provision system (100) may provide only news clusters whose weight size exceeds a predetermined threshold, or provide a predetermined number of news clusters in order of increasing weight.

[0089] Furthermore, the news content provision system (100) can specify a core keyword for each of the multiple sentences, group multiple news clusters by topic of the sentences according to the specified core keyword based on the specified core keyword, and provide multiple news clusters by grouped topics so that the user can check them.

[0090] For example, the news content provision system (100) can specify one of the multiple tokens for sentences belonging to each news cluster as a core keyword.

[0091] In one embodiment, the news content providing system (100) can identify one of the tokens corresponding to the object (or, subject) of each sentence among a plurality of tokens for sentences belonging to each news cluster as a core keyword.

[0092] In another embodiment, the news content providing system (100) may specify as a core keyword any one token corresponding to a predetermined core keyword list among a plurality of tokens for sentences belonging to each news cluster. At this time, the core keyword list may be arbitrarily determined.

[0093] Accordingly, the news content provision system (100) can create multiple topic groups by grouping multiple news clusters that have the same specific core keyword for each news cluster into one topic group.

[0094] Through this, the news content provision system (100) can provide multiple news clusters grouped according to topic groups so that the user can view them.

[0095] Furthermore, the news content provision system (100) can calculate the average value of the index for multiple sentences belonging to each of the multiple news clusters, and sort the multiple news clusters according to the size of the average value of the calculated index so that the user can check them.

[0096] For example, a news content provision system (100) can calculate the average value of the index for a news cluster by summing the indices of multiple sentences belonging to each news cluster and dividing by the number of multiple sentences belonging to the news cluster.

[0097] Accordingly, the news content provision system (100) can sort multiple news clusters in ascending order according to the magnitude of the average value of the previously calculated index, so that the user can check the multiple news clusters according to the time order in which the sentences were written within the news content.

[0098] Furthermore, the news content provision system (100) may have a time period associated with the collection of news content designated, and at each time period, collect multiple newly posted news content, obtain multiple sentences from the multiple news content collected according to the time period and cluster them into multiple news clusters, and arrange the multiple news clusters in order according to the time period so that the user can check them.

[0099] For example, a news content provision system (100) can collect multiple news contents posted on a given day and create multiple news clusters for the collected multiple news contents.

[0100] Through this, the news content provision system (100) can arrange multiple news clusters according to the time order in which the news clusters were created (or news content was collected) and provide multiple news clusters to the user so that they can check the process of a specific event unfolding over time or an event that is attracting people's interest.

[0101] Based on the above configurations, the news content provision system (100) can provide the previously generated multiple news clusters in various forms so that the user can visually check them.

[0102] Referring to FIG. 7 (a), for example, a news content provision system (100) may provide a plurality of news clusters (e.g., topics of news) in the form of a graph showing weights (e.g., importance) for the plurality of news clusters in each time period for a time period (e.g., date) related to the collection of news content.

[0103] Through this, the news content provision system (100) can provide multiple news clusters to check how people's interest or the importance of news covered by media outlets changes over time.

[0104] Referring to Fig. 7(b) as another example, the news content providing system (100) may provide a plurality of news clusters (e.g., news topics) in the form of a bar graph of weights (e.g., importance) according to the number of sentences belonging to each news cluster.

[0105] At this time, multiple news clusters may be the result of clustering multiple sentences accumulated from continuously collected news content, and the news content providing system (100) may highlight the newly created news cluster (e.g., NEW mark) when a news cluster different from the existing news cluster is created due to the occurrence of a new sentence (or event).

[0106] Through this, the news content provision system (100) can provide multiple news clusters so that it can quickly identify what major events are being covered by the media outlet.

[0107] Referring to Fig. 7 (c) as another example, the news content providing system (100) may provide a plurality of news clusters (e.g., Cluster 1 to 4) in the form of a shape based on the size of a weight value according to the number of sentences belonging to each news cluster, and may also provide a range of times when each of the plurality of news contents including sentences belonging to each news cluster is posted.

[0108] Through this, the news content provision system (100) can provide multiple news clusters to check the time range and importance of the relevant events covered by media outlets for each news cluster, and to quickly identify newly occurring events.

[0109] Referring to (d) of FIG. 7, as another example, the news content providing system (100) may provide a plurality of news clusters (e.g., news topics) in the form of a ratio of weights (e.g., importance) according to the number of sentences belonging to each news cluster.

[0110] In this case, the ratio for weighting may be the ratio of the number of sentences belonging to each news cluster to the total number of sentences belonging to multiple news clusters based on a specific core keyword.

[0111] Through this, the news content provision system (100) can provide multiple news clusters to check the status of people's interest or the importance of news covered by media outlets regarding specific key keywords.

[0112] Referring to FIG. 8, the news content provision system (100) can collect news content at a predetermined time interval (e.g., daily) to create multiple news clusters, and process each of the multiple news clusters created at each hour interval into a predetermined table form to provide them for the user to view.

[0113] At this time, the news content provision system (100) classifies and provides multiple news clusters generated for each category (1-3-1) based on news content collected by category, and in each category (1-3-1), among the multiple news clusters, can select and provide a predetermined number of news clusters according to the size of the weight.

[0114] At this time, the core keyword of the news cluster can be designated as the title (1-4-1), and any one of the sentences belonging to the news cluster (1-4-2) (or a sentence reconstructed by the token for the news cluster) can be provided together.

[0115] In this regard, with reference to FIG. 9, the news content providing system (100) may provide a specific news cluster provided in the form of a table, a graph (1-4-3) showing the relationship between other news clusters corresponding to the core keywords of the said news cluster, and a graph (1-4-4) showing the relationship between the said specific news cluster and other news clusters corresponding to each of the multiple news contents containing multiple sentences belonging to the said news cluster.

[0116] Additionally, referring further to FIG. 10, the news content providing system (100) can provide information related to multiple sentences belonging to a specific news cluster based on user input for any one of the multiple news clusters provided in a table form.

[0117] That is, when a specific news cluster is selected according to user input, the news content provision system (100) can provide a multiple sentences belonging to the news cluster, along with an access path (e.g., a link) to the news content containing each sentence.

[0118] Through the above configurations, the news content providing system (100) according to the present invention collects news content, divides the collected news content into sentence units, and clusters the divided multiple sentences, thereby effectively providing the content of an event that is commonly covered in different news content.

[0119] In addition, the news content providing system (100) according to the present invention calculates the importance of each cluster according to the number of clustered sentences from the news content and provides each cluster according to the calculated importance, thereby effectively providing the event with the highest level of interest among the various events covered in the news content.

[0120] In addition, the news content providing system (100) according to the present invention can effectively provide a process of a specific event covered in the news content changing over time by arranging and providing clustered sentences from the news content in chronological order.

[0121] Furthermore, the present invention described above can be implemented as a program that is executed by one or more processes in an electronic device and stored on a computer-readable recording medium.

[0122] Accordingly, the present invention can be implemented as computer-readable code or instructions on a medium on which a program is recorded. That is, various control methods according to the present invention can be provided in the form of integrated or individual programs.

[0123] Meanwhile, computer-readable media include all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SSD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0124] Furthermore, the computer-readable medium may be a server or cloud storage that includes a storage and is accessible to an electronic device via communication. In this case, the computer may download the program according to the present invention from the server or cloud storage via wired or wireless communication.

[0125] Furthermore, in the present invention, the computer described above is an electronic device equipped with a processor, namely a CPU (Central Processing Unit), and no special limitations are placed on its type.

[0126] Meanwhile, the above detailed description should not be interpreted restrictively in all respects but should be considered exemplary. The scope of the invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are included within the scope of the invention.

Claims

Claim 1 A method for providing news content, comprising: a step of collecting news content; a step of separating a plurality of sentences included in the news content according to sentence units; a step of clustering the separated plurality of sentences to generate a plurality of news clusters; and a step of calculating a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, and providing the plurality of news clusters based on the calculated weight, wherein the step of separating the plurality of sentences includes a step of assigning an index to each of the plurality of sentences separated from the news content based on the position of each sentence in the news content. Claim 2 A method for providing news content according to claim 1, wherein the step of generating the plurality of news clusters includes the step of performing filtering by removing sentences having a predetermined keyword corresponding to a unique sentence of the news content from among the plurality of sentences separated from each of the plurality of news contents. Claim 3 A method for providing news content according to claim 2, wherein the step of generating the plurality of news clusters further comprises the step of clustering the filtered plurality of sentences into the plurality of news clusters based on the similarity between different sentences. Claim 4 A method for providing news content according to claim 1, wherein the step of providing the plurality of news clusters comprises the step of calculating the weight based on the number of sentences belonging to each of the plurality of news clusters and sorting the plurality of news clusters in order of highest calculated weight. Claim 5 A method for providing news content according to claim 1, wherein the step of providing the plurality of news clusters comprises, for each of the plurality of sentences, specifying a key keyword, grouping the plurality of news clusters by topic of the sentences according to the specified key keyword based on the specified key keyword, and providing the plurality of news clusters by grouped topics. Claim 6 delete Claim 7 A method for providing news content according to claim 1, wherein the step of providing the plurality of news clusters comprises the step of calculating an average value of an index for the plurality of sentences belonging to each of the plurality of news clusters, and sorting the plurality of news clusters according to the magnitude of the average value of the calculated index. Claim 8 A method for providing news content according to claim 1, wherein the step of generating a plurality of news clusters comprises collecting and classifying a plurality of news contents by media outlet, and generating a plurality of news clusters by media outlet for each of the classified plurality of news contents. Claim 9 A news content providing system comprising: a communication unit connected to a server storing news content; a control unit that collects news content from the server, separates a plurality of sentences included in the news content according to sentence units, and clusters the separated plurality of sentences to generate a plurality of news clusters, wherein the control unit calculates a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, provides the plurality of news clusters based on the calculated weight, and assigns an index to each of the plurality of sentences separated from the news content based on the position of each sentence in the news content in order to separate the plurality of sentences. Claim 10 A program stored on a computer-readable recording medium that is executed by one or more processes in an electronic device, the program comprising instructions for performing the steps of: collecting news content; separating a plurality of sentences included in the news content according to sentence units; clustering the separated plurality of sentences to generate a plurality of news clusters; and calculating a weight for each of the plurality of news clusters based on the plurality of sentences belonging to each of the plurality of news clusters, and providing the plurality of news clusters based on the calculated weight, wherein the step of separating the plurality of sentences is characterized by assigning an index to each of the plurality of sentences separated from the news content based on the position of each sentence in the news content.

Images