JP2019212083A - Information providing apparatus, information providing method, and program - Google Patents

Abstract

To provide a review document expected to be more useful for a user.SOLUTION: An information providing apparatus according to the present invention generates a summary of a review document based at least on a reference document other than the review document among the review document and the reference document. The reference document is an electronic document including words and phrases indicating contents of a question asked by a user to a third party (e.g., a search engine or a large number of anonymous users) by using a service providing an answer to a user's question. The information providing apparatus generates a summary of the review document and provides contents including the summary to a terminal apparatus (external apparatus) a user uses.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an information providing apparatus, an information providing method, and a program.

  Traditionally, information including user reviews on products is analyzed to extract the usage and characteristics of each product, associate the extracted usage with characteristics, and when the usage is selected by the user, it is associated with the selected usage A technology is known that ranks products according to the depth of association with a selected application based on the characteristics and outputs a list of products ranked higher (for example, Patent Document 1). reference).

JP 2012-168925 A

  However, in the conventional technique, there is a case where a recommendation is made so as to browse a review document of another user that is not necessarily useful to the user.

  The present invention has been made in view of the above problems, and an object thereof is to provide an information providing apparatus, an information providing method, and a program capable of providing a review document that is expected to be more useful to a user. Yes.

  One aspect of the present invention is a generation unit that generates a summary of the review document based on at least the reference document out of a review document and a reference document different from the review document, and the generation unit generates the review document A content providing unit that provides a content including the summary to a terminal device.

  According to one aspect of the present invention, a review document that is expected to be more useful to a user can be provided.

It is a figure which shows an example of the information provision system 1 containing the information provision apparatus 100 in 1st Embodiment. It is a figure which shows an example of a structure of the information provision apparatus 100 in 1st Embodiment. It is a figure which shows an example of the web page provided by the content provision part. It is a flowchart which shows the flow of a series of processes by the dictionary production | generation part 114. FIG. It is a figure which shows typically the process which classifies a review document. It is a figure which shows an example of a 1st frequent word and a 2nd frequent word. It is a figure which shows an example of a co-occurrence word. 4 is a flowchart showing a flow of a series of processes by a summary generation unit 116. It is a figure which shows an example of the result of the appearance frequency of the frequent word in a review document. It is a figure which shows an example of the result of the appearance frequency of the co-occurrence word in a review document. It is a figure which shows an example of the derivation | leading-out result of a score. It is a figure which shows an example of the web page of a hyperlink destination. It is a figure which shows an example of the summary based on a 2nd score. It is a figure which shows an example of the summary based on a 1st score. It is a figure which shows an example of the hardware constitutions of the information provision apparatus 100 of embodiment.

  Hereinafter, an information providing apparatus, an information providing method, and a program to which the present invention is applied will be described with reference to the drawings.

[Overview]
The information providing apparatus is realized by one or more processors. The information providing apparatus generates a summary of the review document based on at least the reference document among the review document and the reference document different from the review document.

  Review documents are specific items such as products or services sold at shopping sites, accommodation facilities such as hotels, sightseeing spots, companies that can buy and sell stocks, content such as animation and movies, and events such as festivals and fireworks displays. This is an electronic document that expresses the user's impressions, reviews, opinions, reports, etc.

  A reference document is a service that provides an answer to a user's question. This is an electronic document containing the indicated word or phrase. For example, when the user uses a search service, the reference document includes a query input by the user. For example, when the user uses the knowledge sharing service, the reference document includes a sentence posted by the user to the knowledge sharing service. The knowledge sharing service is a service in which information in which a question and an answer are combined can be shared (viewed) by an unspecified number of users.

  When the information providing device generates a summary of the review document, the information providing device provides content including the summary to a terminal device (external device) used by the user. The content including the summary is, for example, a web page on which a product review document is posted on a shopping site or the like. This makes it possible to provide a review document that is expected to be more useful to the user among a plurality of review documents.

[First Embodiment]
[overall structure]
The first embodiment will be described below. FIG. 1 is a diagram illustrating an example of an information providing system 1 including an information providing apparatus 100 according to the first embodiment. The information provision system 1 in 1st Embodiment is provided with the one or more terminal devices 10 and the information provision apparatus 100, for example. These devices are connected via a network NW.

  Each apparatus shown in FIG. 1 transmits and receives various information via the network NW. The network NW includes, for example, the Internet, a WAN (Wide Area Network), a LAN (Local Area Network), a provider terminal, a wireless communication network, a wireless base station, a dedicated line, and the like. Note that it is not necessary for all combinations of the devices shown in FIG. 1 to be able to communicate with each other, and the network NW may partially include a local network.

  The terminal device 10 is a terminal device including an input device, a display device, a communication device, a storage device, and an arithmetic device such as a mobile phone such as a smartphone, a tablet terminal, and various personal computers. The communication device includes a network card such as a NIC (Network Interface Card), a wireless communication module, and the like. In the terminal device 10, a user agent (UA) such as a web browser or an application program is activated, and transmits a request according to the user's input operation to the information providing device 100. Further, the terminal device 10 in which the UA is activated displays various images on the display device based on the information acquired from the information providing device 100.

[Configuration of Information Providing Device]
FIG. 2 is a diagram illustrating an example of the configuration of the information providing apparatus 100 according to the first embodiment. As illustrated, the information providing apparatus 100 includes, for example, a communication unit 102, a control unit 110, and a storage unit 130.

  The communication unit 102 includes a communication interface such as a NIC, for example. The communication unit 102 communicates with the terminal device 10 and the like via the network NW. For example, the communication unit 102 communicates with the terminal device 10 activated by a UA such as a web browser and receives a request from the terminal device 10.

  The control unit 110 includes, for example, a content providing unit 112, a dictionary generation unit 114, and a summary generation unit 116.

  For example, a hardware processor (or processor circuit) such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) executes a program (software) stored in the storage unit 130. It is realized by. In addition, some or all of these components are made by hardware (circuit unit: circuit) such as LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), and FPGA (Field-Programmable Gate Array). It may be realized or may be realized by cooperation of software and hardware. The above program may be stored in advance in the storage unit 130 or stored in a removable storage medium such as a DVD or a CD-ROM, and the storage medium is attached to the drive device of the information providing apparatus 100. By doing so, it may be installed in the storage unit 130 from the storage medium.

  The storage unit 130 is realized by a storage device such as an HDD (Hard Disc Drive), a flash memory, an EEPROM (Electrically Erasable Programmable Read Only Memory), a ROM (Read Only Memory), or a RAM (Random Access Memory). The storage unit 130 stores review document information 132, reference document information 134, and the like in addition to various programs such as firmware and application programs. The review document information 132 is information including the plurality of review documents described above, and the reference document information 134 is information including the reference documents described above.

  When the communication unit 102 receives a request from the terminal device 10, the content providing unit 112 controls the communication unit 102 to provide information corresponding to the request to the terminal device 10. For example, when the communication unit 102 receives an HTTP request from the terminal device 10 in which the web browser is activated as UA, the content providing unit 112 may provide the terminal device 10 with a web page. For example, when the communication unit 102 receives an API request from the terminal device 10 in which the application is activated as UA, the content providing unit 112 may provide content to the terminal device 10.

  FIG. 3 is a diagram illustrating an example of a web page provided by the content providing unit 112. In the illustrated example, web pages of websites (hereinafter collectively referred to as sales sites) such as shopping sites, auction sites, and flea market sites that sell products on the Internet are schematically shown. For example, a product image, a title, a price, a cart insertion button, and the like are displayed in the area R1 of the web page of the sales site, a review document summary is displayed in the area R2, and a review document is displayed in the area R3. Displayed in list format. The summary shown in the region R2 is generated by the summary generation unit 116 described later. This will be described later. In the example of FIG. 3, it has been described that the summary of the review document is displayed on the sales site. However, the present invention is not limited to this. For example, a reservation site for accommodation facilities, an information site for tourist sites, stocks, virtual currency, etc. Any website can be used as long as the review document created by the user is displayed as a list, such as an information exchange site related to finance. Further, these websites may be realized by application programs installed on a mobile phone or the like.

[Processing flow of dictionary generator]
The dictionary generation unit 114 generates a dictionary used when the summary generation unit 116 generates a summary. Hereinafter, a flow of a series of processing by the dictionary generation unit 114 will be described with reference to a flowchart. FIG. 4 is a flowchart showing a flow of a series of processes by the dictionary generation unit 114. The process of this flowchart may be repeatedly performed with a predetermined period, for example.

  First, for example, the dictionary generation unit 114 selects one target category (hereinafter referred to as a target category) from among a plurality of categories that are candidates for classification destinations of products posted on the sales site (S100). ).

  Next, the dictionary generation unit 114 acquires review documents for all products classified into the selected target category from the review document information 132 stored in the storage unit 130 (S102). In the review document information 132, for example, when a review target is a product sold on a sales site, the review document information 132 is classified into a plurality of categories (product categories) such as fashion, home appliances, food, audio, furniture, games, and sports. One or more review documents for each product being included are included. For example, when the category of “home appliance” is selected as the target category, the dictionary generation unit 114 is classified into “home appliance” from a plurality of review documents included in the review document information 132. A review document for each product such as a cleaner, a vacuum cleaner, and a refrigerator is acquired.

  Next, the dictionary generation unit 114 classifies the acquired review documents into a high evaluation review group in which high evaluation review documents are collected and a low evaluation review group in which low evaluation review documents are collected (S104).

  FIG. 5 is a diagram schematically illustrating a process of classifying review documents. As shown in the figure, the review document is associated with a higher score as the evaluation with respect to the review object is higher, and with a lower rating as the evaluation with respect to the review object is lower. This score is arbitrarily set by the user, for example, within a range from the lowest score (for example, 0.0) to the highest score (for example, 5.0). Therefore, for example, as shown in the figure, the dictionary generation unit 114 classifies review documents associated with a score of a certain first threshold value (for example, 4.0) or more into a high evaluation review group, and is smaller than the first threshold value. A review document associated with a score equal to or lower than the second threshold (for example, 2.0) may be classified into a low evaluation review group. The first threshold value and the second threshold value may be the same value. Moreover, the dictionary production | generation part 114 may classify the review document with which the score was matched into three or more review groups. The review document classified into the high evaluation review group is an example of “first review document”, and the review document classified into the low evaluation review group is an example of “second review document”.

  Next, the dictionary generation unit 114 extracts frequent words from the classified high evaluation review group and low evaluation review group (S106). A frequent word is a word (word) or a phrase (phrase) that appears frequently in a review document. For example, in each review group, a word or phrase whose number of appearances is greater than or equal to a threshold value or a predetermined number of words or phrases having a high number of appearances may be used as frequent words. Hereinafter, the frequent words extracted from the high evaluation review group will be referred to as “first frequent words”, and the frequent words extracted from the low evaluation review group will be referred to as “second frequent words”.

  FIG. 6 is a diagram illustrating an example of the first frequent word and the second frequent word. In the illustrated example, frequent occurrences obtained from review documents given to products classified as “air purifiers”, which are lower categories of the category “home appliances” (that is, the target category is “air purifiers”). Represents a word. For example, words such as “rapid”, “cheap”, and “early” are extracted as the first frequent words from the high evaluation review group, and “price” as the second frequent words from the low evaluation review group. And words such as “sound” and “defective product” are extracted. As shown in the example, when the same word (or phrase) appears frequently in both the high evaluation review group and the low evaluation review group such as “price” and “correspondence”, the first frequent appearance The same word may be included in both the word and the second frequent word.

  Next, the dictionary generation unit 114 extracts a co-occurrence word from the reference document indicated by the reference document information 134 stored in the storage unit 130 (S108). The co-occurrence word is a frequent word that co-occurs with a keyword indicating the target category in the reference document among a plurality of frequent words extracted from the high evaluation review group and the low evaluation review group. For example, when the target category is “air cleaner”, “air cleaner” representing the target category is a keyword indicating the target category.

  FIG. 7 is a diagram illustrating an example of a co-occurrence word. In the illustrated example, the reference document is a log of a query (set of a plurality of queries) input to a search site or the like. As described above, when the reference document is a query log and the keyword representing the target category is “air cleaner”, the dictionary generation unit 114 inserts a space, a punctuation mark or the like together with “air cleaner”, for example. Simultaneously input words and phrases are extracted as co-occurrence words. In the illustrated example, words such as “cleaning”, “cigarette”, “odor”, and “filter” are extracted as co-occurrence words.

  In the example described above, the reference document is described as a query log, but the present invention is not limited to this. For example, when there is a knowledge sharing service in which a third person posts an answer when a user asks a question, the reference document may be a sentence representing the content of a question posted to the knowledge sharing service.

  Next, based on the extracted frequent words and co-occurrence words, the dictionary generation unit 114 creates a high evaluation review dictionary (hereinafter referred to as a high evaluation dictionary Da) and a low evaluation review dictionary (hereinafter referred to as a low evaluation review dictionary). (Referred to as an evaluation dictionary Db) (S110).

  For example, the dictionary generation unit 114 includes a dictionary that collects frequent words extracted from the high evaluation review group (hereinafter, high evaluation frequent word dictionary) and a dictionary that collects frequent words extracted from the low evaluation review group (hereinafter, low evaluation). A frequent word dictionary) and a dictionary that collects co-occurrence words extracted from the reference document (hereinafter referred to as a co-occurrence word dictionary). Note that the dictionary generation unit 114, when extracting frequent words or co-occurrence words as phrases such as “sound is quiet” or “sound is quiet”, except for these particles except for particles and prefixes. Since the word and the predicate are essentially synonyms, the difference (particle shaking) such as the particle and prefix may be ignored and registered in the dictionary as the same phrase.

  Next, whether or not the dictionary generation unit 114 has generated a high-evaluation frequent word dictionary, a low-evaluation frequent word dictionary, and a co-occurrence word dictionary for all categories that are candidates for classification of products to be posted on the sales site. If it is determined that the dictionaries have not been generated for all categories, the target category is changed to another category, and the process returns to the process of S100 described above to newly generate a dictionary.

  On the other hand, if the dictionary generation unit 114 determines that the dictionary has been generated for all categories, that is, if the dictionary has been generated for each of the categories that are candidates for classification of the product, the processing of this flowchart ends. To do.

[Processing flow of summary generator]
When the dictionary generation unit 114 generates a dictionary, the summary generation unit 116 of the control unit 110 uses the dictionary to exemplify a plurality of review documents included in the review document information 132 as illustrated in FIG. A summary of a simple web page. Hereinafter, a flow of a series of processing by the summary generation unit 116 will be described with reference to a flowchart. FIG. 8 is a flowchart showing a flow of a series of processes by the summary generation unit 116. The process of this flowchart may be repeatedly performed according to the update period of the review document information 132, for example.

  First, the summary generation unit 116 uses the high evaluation frequent word dictionary, the low evaluation frequent word dictionary, and the co-occurrence word dictionary generated by the dictionary generation unit 114 for each review document included in the review document information 132. The appearance frequency (number of appearances) of the frequent word and the co-occurrence word is derived (S200).

  For example, the summary generation unit 116 derives the appearance frequency of the frequent words and the co-occurrence words using the same dictionary as the category of the product reviewed in the target review document. More specifically, when the review document from which the appearance frequency is derived is a review document for a product classified into the category “food”, the summary generation unit 116 includes a plurality of items generated by the dictionary generation unit 114. From the dictionary, select a dictionary generated by selecting the category “food” as the target category, and use the selected dictionary to generate frequent and co-occurrence words from the review document for which the appearance frequency is to be derived. The frequency of occurrence of is derived.

  FIG. 9 is a diagram illustrating an example of a result of the appearance frequency of a frequent word in a review document. In the illustrated example, among the plurality of first frequent words registered in the highly evaluated frequent word dictionary, the first frequent word “sound” has an appearance count of 74 times in the highly evaluated review group, and is low rated. Of the plurality of second frequent words registered in the frequent word dictionary, the second frequent word “scent” indicates that the number of appearances is five in the low evaluation review group. That is, it represents that the number of appearances of “sound” is 74 times out of the total number of appearances 466 times of the first frequent words in the high evaluation review document, and the second frequent words in the low evaluation review document. Of the 49 appearances, the “scent” appears 5 times. In this case, for example, the summary generation unit 116 derives the appearance frequency of the first frequent word “sound” as 74/466 and derives the appearance frequency of the second frequent word “scent” as 5/49. .

  FIG. 10 is a diagram illustrating an example of a result of the appearance frequency of co-occurrence words in a review document. In the illustrated example, the co-occurrence word “cleaning” out of a plurality of co-occurrence words registered in the co-occurrence word dictionary in the high evaluation review group has 21451 occurrences, and “co-occurrence” “cigarette”. For the word, the number of appearances is 13850, and for the co-occurrence word “odor”, the number of appearances is 13186. In this case, the summary generation unit 116 derives the appearance frequency of the co-occurrence word “cleaning” in the high evaluation review group as 21451/122879, and determines the appearance frequency of the co-occurrence word “cigarette” in the high evaluation review group, Derived as 13850/122879, the appearance frequency of the co-occurrence word “odor” in the highly evaluated review group is derived as 13186/122879.

  In the illustrated example, the co-occurrence word “odor” among the plurality of co-occurrence words registered in the co-occurrence word dictionary in the low evaluation review group has an appearance count of 13186 and is referred to as “filter”. For the co-occurrence word, the number of appearances is 12269, and for the co-occurrence word “electricity cost”, the number of appearances is 5364. In this case, the summary generation unit 116 derives the appearance frequency in the low evaluation review group of the co-occurrence word “odor” as 13186/47113, and the appearance frequency in the low evaluation review group of the co-occurrence word “filter” It is derived as 12269/47113, and the appearance frequency in the low evaluation review group of the co-occurrence word “electricity cost” is derived as 5364/47113.

  Next, the summary generation unit 116 derives a score for each word included in the review document for each product based on the appearance frequency of the derived frequent word and the appearance frequency of the co-occurrence word (S202). For example, the summary generation unit 116 derives, as a score (Score (i, c, kw)), a weighted average of the number of appearances of a frequent word and the number of appearances of a co-occurrence word based on the formula (1). In this case, the score tends to increase as the word tends to appear as a frequent word and also as a co-occurrence word.

  For example, in a review document, when a product or service sold on a shopping site or the like is a subject of review, i in the formula represents each product (or each service). Further, c represents a category, and kw represents a frequent word or a co-occurrence word included in the review document. FrqR (i) represents the number of reviews for a certain product i, and FrqR (i, kw) represents the number of reviews for the product i in which a frequent word or co-occurrence word appears. Further, for example, when the reference document is a query log, FrqQ (c) represents the total number of queries including a keyword representing a category (for example, a word such as “air purifier”) among queries appearing in the reference document. , FrqQ (c, kw) represents the number of queries including a keyword representing a category and a frequent word or co-occurrence word among queries appearing in the reference document. α is a coefficient representing a weight, and may be a coefficient such as FrqR (i) / FrqQ (c), for example. In addition, although Formula (1) demonstrated as what is a derivation type | formula which derives | leads-out a weighted average as a score, it is not restricted to this, For example, a geometric average, a harmonic average, or another aggregated index value is derived as a score. It may be a thing.

  Further, the summary generation unit 116 may derive a score for each word included in the review document for each product based only on the appearance frequency of the co-occurrence words. In this case, the summary generation unit 116 derives, for example, FrqQ (c, kw) / FrqQ (c) as a score.

  FIG. 11 is a diagram illustrating an example of a score derivation result. In the illustrated example, in the high evaluation review group, the word “compact” has the highest score and the word “sound” has the second highest score among the plurality of words included in the set of first frequent words and co-occurrence words. The score is large, and the word “price” represents the third highest score. In the illustrated example, in the low evaluation review group, the word “odor” has the highest score among the plurality of words included in the set of second frequent words and co-occurrence words, and the word “electricity cost” The second largest score, and the word “odor” represents the third largest score.

  Next, the summary generation unit 116 selects the upper predetermined number of words (frequent word or co-occurrence word) having a large score (S204). For example, in the case of the score result illustrated in FIG. 11 when the predetermined number is 3, the summary generation unit 116 selects the words “compact”, “sound”, and “price” for the high evaluation review group, For the low evaluation review group, the words “smell”, “electricity cost”, and “smell” are selected.

  Next, the summary generation unit 116 generates a summary of the review document in which the selected word appears as a frequent word or a co-occurrence word (S206). For example, when three words “compact”, “sound”, and “price” are selected for the high evaluation review group, the summary generation unit 116 selects “compact” from a plurality of review documents included in the high evaluation review group. A summary of high-review reviews (summary for high evaluation) including a set of review documents in which “Sound” appears, a set of review documents in which “Sound” appears, and a set of review documents in which “Price” appears. To be generated). Similarly, when the summary generation unit 116 selects three words “smell”, “electricity cost”, and “smell” for the low evaluation review group, among the plurality of review documents included in the low evaluation review group, A summary that includes a set of review documents in which "odor" appears, a set of review documents in which "electricity bill" appears, and a set of review documents in which "odor" appears is a summary that summarizes low-review reviews (low (Summary for evaluation). Thereby, the processing of this flowchart is completed. The high evaluation summary is an example of a “first summary”, and the low evaluation summary is an example of a “second summary”.

  When the high evaluation summary and the low evaluation summary are generated by the summary generation unit 116, the content providing unit 112 described above displays a web page including the high evaluation summary and the low evaluation summary (for example, a web page as illustrated in FIG. 3). Generate. Then, when the request for the web page on which the review document is posted is received by the communication unit 102, the content providing unit 112 controls the communication unit 102 to generate the high evaluation summary and the low evaluation summary generated in the request source terminal device 10. Send a web page containing the evaluation summary as a response. For example, the content providing unit 112 transmits a web page in which the high evaluation summary is displayed as “good” and the low evaluation summary is displayed as “bad” in the review summary in the region R2 of FIG. You can do it.

  Further, when displaying the high evaluation summary and the low evaluation summary, the content providing unit 112 displays the number of review documents associated with each word and a representative review among the review documents. Words and phrases that appear in the document may be displayed. A typical review document may be, for example, a review document having the largest number of browsing, or a review document having the largest number of times that a review has been reported as a reference from another user.

  For example, when the word “sound” is included in the top predetermined number of words having a large score extracted from the high evaluation review group, the content providing unit 112 includes a plurality of review documents included in the high evaluation review group. Of these, the number of review documents in which the word “sound” appears as a frequent word or co-occurrence word is displayed as the number of cases, and the review document is a representative review document among a plurality of review documents in which the word “sound” appears. You may display the title and summary text. In the example of FIG. 3, the number of review documents in which the word “sound” appears is “74”, and a representative review document among a plurality of review documents in which the word “sound” appears. The title and summary indicate that the sound is quiet.

  Further, the content providing unit 112 may associate a URL (Uniform Resource Locator) of a web page displaying a review document included in each summary as a list with a title of a typical review document as a hyperlink. That is, the content providing unit 112 includes a character or an image associated with a URL or the like that can be viewed as a summary for each review document, and provides the content to the terminal device 10.

  FIG. 12 is a diagram illustrating an example of a hyperlink destination web page. In the example shown in the drawing, the web page of the transition destination when the character string “easy to clean” is clicked or tapped in “good place (high evaluation summary)” is shown. In such a case, the content providing unit 112 posts a list of review documents in which the word “cleaning” appears as a frequent word or a co-occurrence word in the highly evaluated review group on the hyperlink destination web page. Good. This allows the user to click on or tap each word associated with a hyperlink in the summary without selecting the review document of interest from multiple review documents. Documents are aggregated individually and you can access pages displayed in a list format.

  According to the first embodiment described above, a review document and a reference document that is different from the review document are generated based on at least the reference document, and a web including the generated review document summary is generated. By providing a page or the like as content to the user terminal device 10, a review document that is expected to be more useful to the user can be provided.

  In general, in addition to catalog specifications for products to be purchased, etc., users tend to use a search site, a knowledge sharing site, or the like to find a feeling of use that can be first understood when actually used. In such a case, it can be considered that the words or phrases input as a query when searching for the feeling of use of the product directly represent what the user actually cares about. For this reason, as in this embodiment, the user really knows by analyzing a review document based on a reference document including a query log input to a search site or a question sentence posted on a knowledge sharing site. Review documents with the information you want can be provided as a summary. As a result, the number of users who refer to the review document is increased, and mismatches such that the actually purchased product is different from what is assumed can be eliminated, and the return rate of the product can be reduced. As a result, the store side (retailer side) that sells the product can reduce the delivery fee and the cost of return exchange, etc., so the business operator who opens a new store on the sales site or the business operator who continues to open the store The number of can be increased.

  Further, according to the first embodiment described above, since the review document is provided as a summary, the readability of the review document can be improved. For example, when a user browses a sales site or the like using the terminal device 10 having a small screen size such as a smartphone, it is restricted to display all review documents depending on the screen size of the terminal device 10. Only some review documents may be displayed. In such a case, the readability of the user's review document is likely to deteriorate. On the other hand, in the present embodiment, a query or the like that is easily input when the user examines the feeling of use is used as a co-occurrence word, and a review document selected based on the appearance frequency of the co-occurrence word is provided as a summary. The readability of the review document can be improved. As a result, the user can easily find a review document that serves as a reference, and the convenience of a website where the review document is displayed in a list format can be improved.

Second Embodiment
Hereinafter, a second embodiment will be described. The second embodiment is different from the first embodiment described above in that the review document included in the summary is highlighted. Hereinafter, differences from the first embodiment will be mainly described, and description of points common to the first embodiment will be omitted. In the description of the second embodiment, the same parts as those in the first embodiment will be described with the same reference numerals.

  The summary generation unit 116 in the second embodiment derives a score (hereinafter referred to as a first score) for each word included in the review document for each product based on the appearance frequency of the frequent words and the appearance frequency of the co-occurrence words. At the same time, a score (hereinafter referred to as a second score) is derived for each word included in the review document for each product based on the appearance frequency of the frequent words. For example, the summary generation unit 116 may derive FrqR (i, kw) / FrqR (i) as the second score. That is, the summary generation unit 116 increases the second score for words having a higher appearance frequency in each review group.

  Then, the summary generation unit 116 generates a summary of the review document in which the upper predetermined number of words having a large first score appears, and generates a summary of the review document in which the upper predetermined number of words having a large second score appears. .

  In the second embodiment, the content providing unit 112 compares the summary of the review document selected based on the first score with the summary of the review selected based on the second score, and part of the two summaries to be compared. Alternatively, when all the review documents are different, the review document different from the review document of the summary based on the second score is highlighted in the summary based on the first score.

  FIG. 13 is a diagram illustrating an example of a summary based on the second score. FIG. 14 is a diagram illustrating an example of a summary based on the first score. In the “good place (summary for high evaluation)” of the summary illustrated in FIG. 13, the appearance frequency in the high evaluation review group of “sound”, “compact”, “price”, “cheap”, and “small” is displayed. A review document containing the top five large words is included. On the other hand, the “good place (high evaluation summary)” of the summary illustrated in FIG. 14 has the top three appearance frequencies of “sound”, “compact”, and “price” having the highest appearance frequency. In addition to a review document that includes two words, review documents that include the words “cigarette” and “clean” that appear less frequently than words such as “cheap” and “small” are included.

  Further, the “bad place (summary for low evaluation)” of the summary illustrated in FIG. 13 includes “smell”, “break”, “fan”, “abnormal noise”, and “not sure” in the low evaluation review group. A review document including the top five words having the highest appearance frequency is included. On the other hand, in the “bad place (summary for low evaluation)” of the summary illustrated in FIG. 14, words having a high appearance frequency in the low evaluation review group such as “smell”, “break”, and “fan” are included. In addition to the review document including the review document, a review document including a word such as “smell” or “electricity cost” that does not necessarily occur frequently is included.

  In such a case, the content providing unit 112 highlights a review document different from the summary illustrated in FIG. 13 in the summary illustrated in FIG. In the example of FIGS. 13 and 14, the content providing unit 112 emphasizes the title “smoking cigarette” of a typical review document in which the word “cigarette” appears in “good place (high evaluation summary)”. Display and highlight the title “easy to clean” of a typical review document in which the word “cleaning” appears. More specifically, the content providing unit 112 highlights a character string indicating the title of a typical review document by adding an underline, increasing a character of the character string, or changing a color. You can do it. Further, the content providing unit 112 highlights the title “odor” of a typical review document in which the word “smell” appears in the “bad place (summary for low evaluation)” and the word “electricity cost”. Highlight the title “Electricity costs” of a typical review document that appears.

  According to the second embodiment described above, compared to the case where a summary of review documents is generated based on the appearance frequency of frequent words that frequently appear between review documents, the frequent words that appear frequently between review documents and the reference document When review document summaries are generated by comprehensively considering category keywords and co-occurrence words, and if review documents differ between summaries, highlight different review documents, Even a useful (useful) review document that has a small number of cases and is easily buried in another review document can prompt the user to browse.

<Hardware configuration>
The information providing apparatus 100 according to the above-described embodiment is realized by a hardware configuration as illustrated in FIG. 15, for example. FIG. 15 is a diagram illustrating an example of a hardware configuration of the information providing apparatus 100 according to the embodiment.

  The information providing apparatus 100 includes an NIC 100-1, a CPU 100-2, a RAM 100-3, a ROM 100-4, a secondary storage device 100-5 such as a flash memory and an HDD, and a drive device 100-6. Are connected to each other. The drive device 100-6 is loaded with a portable storage medium such as an optical disk. A program stored in a portable storage medium attached to the secondary storage device 100-5 or the drive device 100-6 is expanded in the RAM 100-3 by a DMA controller (not shown) or the like and executed by the CPU 100-2. Thus, the control unit 110 is realized. The program referred to by the control unit 110 may be downloaded from another device via the network NW.

  As mentioned above, although the form for implementing this invention was demonstrated using embodiment, this invention is not limited to such embodiment at all, In the range which does not deviate from the summary of this invention, various deformation | transformation and substitution Can be added.

  DESCRIPTION OF SYMBOLS 1 ... Information provision system, 10 ... Terminal device, 100 ... Information provision apparatus, 102 ... Communication part, 110 ... Control part, 112 ... Content provision part, 114 ... Dictionary production | generation part, 116 ... Review production | generation part, 130 ... Memory | storage part

Claims (10)

A generation unit that generates a summary of the review document based on at least the reference document among the review document and a reference document different from the review document;
A content providing unit that provides the terminal device with content including the summary generated by the generating unit;
An information providing apparatus comprising: The reference document includes a question content of a user using a service that provides an answer to the question.
The information providing apparatus according to claim 1. The services include a search service, or a knowledge sharing service that allows a combination of questions and answers to be shared by an unspecified number of users.
The question content includes
If the service includes the search service, a query entered by a user for the search service is included,
When the knowledge sharing service is included in the service, a sentence posted by a user as a question to the knowledge sharing service is included.
The information providing apparatus according to claim 2. The generator is
For each of the plurality of review documents having the same review target, the frequency of appearance of a co-occurrence word that co-occurs with the keyword representing the review target in the reference document is derived,
Generating the summary based at least on the frequency of occurrence of the derived co-occurrence word;
The information providing device according to any one of claims 1 to 3. The generation unit further includes:
Deriving frequent words that frequently appear between the plurality of review documents,
Deriving the appearance frequency of the frequent word for each of the plurality of review documents having the same review target,
Generating the summary based on the frequency of appearance of the frequent words and the frequency of occurrence of the co-occurrence words;
The information providing apparatus according to claim 4. Each of the plurality of review documents is associated with a higher score for a review object and a smaller score for a lower evaluation for the review object,
The frequent word includes a first frequent word that frequently appears in a first review document associated with the score that is equal to or higher than a first threshold among the plurality of review documents, and a second threshold that is smaller than the first threshold. A second frequent word that frequently appears in the second review document associated with the score of
The generator is
The appearance frequency of the first frequent word and the appearance frequency of the co-occurrence word are derived from the first review document, and the appearance frequency of the second frequent word and the occurrence frequency of the co-occurrence word are derived from the second review document. Deriving the appearance frequency,
Generating a summary of the first review document based on the derived appearance frequency of the first frequent word and the appearance frequency of the co-occurrence word;
Generating a summary of the second review document based on the derived appearance frequency of the second frequent word and the appearance frequency of the co-occurrence word;
The information providing apparatus according to claim 5. The generation unit uses, as a summary of the first review document, a summary including the upper predetermined number of the first review documents having a large weighted average of the appearance frequency of the first frequent words and the appearance frequency of the co-occurrence words. Generate
Generating a summary including the second predetermined number of second review documents having a large weighted average of the appearance frequency of the second frequent word and the appearance frequency of the co-occurrence word as a summary of the second review document;
The information providing apparatus according to claim 6. The content providing unit
Providing the content including the summary of the first review document and the summary of the second review document to the terminal device;
The information providing apparatus according to claim 6 or 7. Computer
Generating a summary of the review document based on at least the reference document among the review document and a reference document different from the review document;
Providing a content including the generated summary to a terminal device;
Information provision method. On the computer,
A process of generating a summary of the review document based on at least the reference document among the review document and a reference document different from the review document;
A process of providing content including the generated summary to a terminal device;
A program for running

Images