Patent Grant
10242105
This application claims foreign priority to Chinese Patent Application No. 201310242522.8 filed on 19 Jun. 2013, entitled “Method and Apparatus for Implementing Comment Ranking by Search Engine,” which is hereby incorporated by reference in its entirety.
The present disclosure is related to the field of the Internet, and more particularly to a method and an apparatus of implementing comment ranking by a search engine.
With the development of the Internet in the world, there are more and more people accepting the Internet and relying on the Internet for their work and daily lives. Activities previously requiring face to face interaction are gradually being replaced by the Internet.
The development of the Internet has generated a variety of applications, one of which is electronic commerce, i.e., e-commerce.
E-commerce utilizes computer technology, network technology, and remote communicating technology to achieve electronics, digitalization, and networking for the entire transaction process. The term e-commerce is usually referred to as a novel business model by which worldwide commercial and trade activities are transacted in an open network environment such as the Internet. In e-commerce, various commercial and trade activities, such as on-line shopping by a consumer, on-line transactions between merchants, on-line payment as well as all kinds of business activities, trade activities, financial activities and related activities of integrated services are conducted based on the application of browsers/servers, where a face to face meeting between a buyer and a seller is not required.
With the development of e-commerce, there are more and more consumers, i.e., users, choosing to purchase various products and services such as clothes, digital products, home appliances, books and lottery tickets online. However, it is not easy for a user to choose the most satisfying item from a wide range of products and services on the Internet, especially when the transaction is made when the user cannot see the physical object or is unfamiliar with the merchant.
There are a lot of websites that provide information related to products and services to help the user make a decision, among which the most convenient and effective one is offering the comments provided by other users with respect to these products or services. Accordingly, the user can make his/her decision based on a massive amount of comments about the product or the service from a massive amount of other users.
A comment search engine based on user comments is a key solution for resolving this issue. When the user clicks a product or a product attribute on a webpage, the comment search engine will perform a search in an index file according to a keyword, and return the most relevant comment information.
However, as a number of comments with respect to the desired product or the desired service may be large, the user may not have sufficient time or energy to read through all the comments. Thus, how the user can rapidly and efficiently retrieve a helpful comment that facilitates decision making has become a problem.
In addition, such commenting platforms are open to the users. In other words, the contents of the comments are not restricted. Consequently, a lot of spam contents such as advertisements, contents irrelevant to the product, or contents with very little useful information may be contained in the growing massive amount of comments with respect to the product. Thus, how to help the user to effectively eliminate or ignore the comments containing spam contents has also become an issue.
The conventional techniques solve these issues by ranking the comments in accordance with time. Specifically, the most recent comment is displayed at the top, and the earlier comment is displayed at the bottom. In this way, the most recent comment can be read by the user in his first priority, and the comments that are published earlier will be read in a low priority as the user scrolls down the page or flips the page.
However, users usually become impatient in the operation of flipping the list of pages during the browsing process. Prior experiences show that users often start to leave the page when the users begin to browse the third page, and a proportion of users leaving the page become higher as the number of pages browsed increases.
Therefore, an excellent content published earlier that is helpful to the user's decision making may be unread by the user. Accordingly, the conventional sorting method that only considers a sequence of publishing time cannot help the user to quickly reach high quality content, and is not helpful for reducing the time of decision making for the user.
Moreover, sometimes the user may not want to browse each comment one by one, but rather wants to focus on reading comments with respect to a specific attribute or a specific aspect of the product or the service. Such a demand may be satisfied by performing an emotional analysis of the comments.
The emotional analysis of the comment is referred to performing a structural analysis with respect to the content of the comment so as to obtain a description of the overall product and attributes in each dimension expressed by the users through their comments.
Therefore, the present techniques efficiently provide good quality and useful comment contents to the user and reduce a number of pages that the user flips.
This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify all key features or essential features of the claimed subject matter, nor is it intended to be used alone as an aid in determining the scope of the claimed subject matter. The term “techniques,” for instance, may refer to apparatus(s), system(s), method(s) and/or computer-readable instructions as permitted by the context above and throughout the present disclosure.
The present techniques improve diversity of results in the direction technology.
The present disclosure provides a method and an apparatus for implementing comment ranking by a search engine.
The present disclosure provides an example method for ranking comments by the search engine. One or more terms describing a target object are extracted from a comment in connection with the target object. One or more useful attributes included in the comment are obtained according to the terms describing the target object. One or more scoring indexes to score the comment are determined according to a number of the useful attributes. The comment is scored according to the scoring indexes. The comment is ranked according to the score.
The present disclosure also provides an example apparatus for ranking comments by the search engine. An extracting module extracts one or more terms describing a target object from a comment in connection with the target object. An obtaining module obtains one or more useful attributes included in the comment according to the terms describing the target object. A determining module determines one or more scoring indexes to score the comment according to a number of the useful attributes. A scoring module scores the comment according to the indexes. A ranking module ranks the comment according to the score.
Compared with the conventional techniques, the present techniques enable the users to read helpful comment information quickly and effectively. The users are able to understand the product, service, and the merchant better and thus make a better shopping decision. The present techniques not only reduce a number of pages that the users flip when the users browse the comments and improve the efficiency in on-line shopping but also help a website to effectively manage the comments of its products and services.
The FIGs are used to provide illustration of the present disclosure and thus are part of the present disclosure. The illustrative examples and their explanations are used to illustrate the present disclosure and should not be used to limit the present disclosure. In the FIGs, the same reference numerals refer to the same or similar parts.
Conventional techniques, when ranking a comment (or “points of view”), do not consider the following factors: an amount of useful product attributes in the comment, i.e., useful information in the comment, a readability of a sentence as the sentence that is too long or redundant is inconvenient for a user to obtain useful information, an amount of useless attributes in the comment such as an attribute of logistics or an attribute of advertisement, and a field of the product attribute as the useful attributes are different for different products such that standby time is a very important product attribute to mobile phones but is meaningless for products such as lotions.
The present techniques, when ranking a comment, take into account the amount of useful information in the comment based on an emotional analysis. The amount of useful information in the comment based on the emotional analysis is obtained by segmenting the comment to an attribute granularity level through the emotional analysis, obtaining evaluation of the useful attributes in the comment, and calculating the amount of information that evaluates the useful attributes, i.e., the amount of useful information in the comment.
For example, in a comment relating to a specific lotion, a sentence “buying it for a classmate” does not provide any evaluation of the product and thus is useless. In contrast, another sentence that “this lotion provides good moisturizing effects” provides an evaluation of a useful attribute about the moisturizing effect of the product, and thus is useful information to users.
Thus, the information that really affects a user's understanding of a target object, i.e., useful information in the comment, may be given more weight during the process of ranking comments. When the useful information is included in the consideration, the displaying order of the comments can be sorted in a way so that contents having more useful information can be shown at a higher priority position. Accordingly, the number of pages that a user needs to check for the comments may be reduced. The present techniques not only save the user's time and effort but also enable the website to manage and use the comments more effectively.
In an example implementation of the present techniques, a computing device includes one or more processors or central processing units (CPUs), input/output interfaces, network interfaces, and memories.
The memory may include non-permanent memory, random access memory (RAM) and/or nonvolatile memory, e.g., read-only memory (ROM) or flash memory (flash RAM) as used in a computer readable medium. The memory can be regarded as an example of a computer readable medium.
The computer readable medium includes permanent and non-permanent as well as removable and non-removable media capable of accomplishing a purpose of information storage by any method or technique. The term of information may be referred to as computer executable instructions, a data structure, a program module or any kind of data. Examples of the computer readable medium or the computer storage medium may include, but are not limited to, phase-change memory (PRAM), static random-access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically-erasable programmable read-only memory (EEPROM), flash memory or any other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disk (DVD) or any other optical storage media, cassette tape, diskette or any other magnetic storage device, or any other non-transmission medium which can be used to store information and accessed by the computing device. As defined herein, the computer readable medium does not include transitory medium such as a modulated data signal and a carrier wave.
The detailed descriptions of the present disclosure are provided below with the example embodiments and accompanying drawings in order to make the objectives, technical solutions and advantages of the present disclosure more clear.
In the following descriptions, a use of “an example embodiment,” “the example embodiment,” “an example,” “the example,” etc., may indicate that a specific feature, structure, characteristic, property, element, or limit is included in the specific example embodiment or example. Nevertheless, it is not necessary that each embodiment or example needs to include such specific feature, structure, characteristic, property, element or limit. In addition, although the phrase “in an example embodiment” is repeatedly used, which may refer to the same embodiment, it does not necessarily need to refer to the same embodiment.
In the following description, some technical features known to one of ordinary skill in the art are omitted for simplicity.
At 110, one or more terms describing a target object are extracted from a comment in connection with the target object. The term “object” refers to the object in connection with the comment.
For example, a user may provide comments with respect to a specific product or a service in an application of e-commerce. Hence, the target object in the example is the specific product or service. Sometimes the user may provide comments with respect to different merchants for other users' reference. In this case, the object is the merchant itself. For another example, the user may provide comments with respect to different restaurants at a specific evaluation website. In this case, the object of the comment is the different restaurants.
For another example, the user may provide comments with respect to a specific news article on a specific news website. In this case, the object is the news article.
For another example, the user may provide comments with respect to a status of his/her friend, another user, or a content posted by the friend on a social network. In this case, the object is the status of his/her friend, another user, or the content posted by the friend on the social network.
It should be noted that the examples of objects described above are merely illustrative rather than restrictive. A person of ordinary skill in the art may easily find other embodiments of the term “object” according to the present disclosure.
In the present disclosure, a term used by the users to describe the target object, e.g., a product, a service or a merchant in the comments, may be referred to as an emotional term.
For example, an operation of word segmentation may be applied to the users' comments first. The operation of word segmentation divides a sentence into different terms. There are various methods for word segmentation, such as a Forward Maximum Matching Method, a Backward Maximum Matching method, and a Bidirectional Maximum Matching method.
Then, a search can be performed according to the terms obtained from the operation of word segmentation in a list containing emotional terms, also known as “an emotional term dictionary”, which is maintained on a server side, to determine the emotional terms included in the comments.
At 120, one or more useful attributes included in the comment are obtained according to the terms describing the target object.
The term “useful attribute” as used herein, for example, describes an attribute which most users are concerned about with respect to the target object such as a product. The useful attributes vary with respect to different types of objects. For example, the useful attributes vary with respect to the products that belong to different fields or categories. In the case of a product such as a mobile phone, the useful attributes may include a standby time, a screen, functionality, etc. In the case of a different product such as a lotion, the useful attributes may include a moisturizing effect, a whitening effect, and a nourishing effect.
According to an example embodiment of the present disclosure, the useful attributes of the target object may be obtained by using a data mining technology to extract them from the existing massive amount of comments with respect to the target object.
According to another example embodiment of the present disclosure, the useful attributes may be preset.
A useful attribute is an aspect (or a characteristic) of the target object, e.g., a product, a service, or a merchant, etc.
An emotional term refers to as a thought or a point of view expressed by the user in the comment about the target object. For example, with respect to a useful attribute “nourishing effect,” the emotional terms in the user's comments may be words such as “nourishing,” “dry,” “great,” etc.
For example, under the category of the lotion, the useful attributes may include an overall product, a nourishing effect, an irritant, etc. Accordingly, for example, a corresponding relationships between normalized useful attributes, the useful attributes, normalize emotional terms, and the emotional terms may be shown in Table 1:
The relationship between the normalized useful attributes and the useful attributes as well as the relationship between the normalized emotional terms and emotional terms may refer to the following
For example, the user's comments in connection with a specific lotion is “generally speaking, I quite like this lotion. It's nourishing with a light smell, and it's mild.” In the comment, “quite like” is the emotional term, “generally speaking” is a useful attribute which is corresponded to a normalized useful attribute of “overall product,” “nourishing” is the emotional term whose corresponding term of useful attribute does not appear in the comment and whose corresponding normalized useful attribute is “nourishing effect.” Therefore, sometimes emotional terms appear in the comment without directly accompanying corresponding terms of useful attributes.
In a general situation, a corresponding useful attribute may be obtained directly from the emotional term.
In some circumstances, obtaining a corresponding useful attribute directly from the emotional term is not possible. For example, the user's comment includes “the money and the quantity are both reasonable,” wherein the emotional term is “reasonable,” and determining a corresponding useful attribute therefrom is not possible. At this point, a search with respect to the corresponding useful attribute of the emotional term may be performed in the comment. One useful attributed is obtained as “money” and its corresponding normalized useful attribute is obtained as “price.” Another useful attribute is obtained as “quantity” and a corresponding normalized useful attribute is obtained as “quantity.”
In another circumstance, the useful attribute may not be obtained from the emotional term and the corresponding useful attributes may not be found out in the comments. In other words, there is no objectively corresponding useful attributes of such emotional term. For example, an emotional term “sharp” is found in the user's comments with respect to the lotion. Such an emotional term does not have a corresponding useful attribute. In this case, the emotional term is ignored without performing further analysis.
According to an example embodiment of the present disclosure, a table recording the corresponding relationship between at least two of the normalized useful attribute, the useful attribute, the normalized emotional term, and the emotional term may be maintained at a database at a server side to be searched during the process of ranking the comments.
At 130, one or more scoring indexes that score the comment are determined according to a number of the useful attributes.
If the number of the useful attributes in the comment is larger than or equal to 1, operations at 220 are performed. At 220, the present techniques determine that the scoring indexes include the number of the useful attributes.
An example formula for calculating the number of useful attributes included in the comment is as follows:
SD=γ (1)
In the formula, γ represents the number of the useful attributes obtained from the operations at 110.
According to an example embodiment of the present disclosure, when the scoring indexes includes the number of the useful attributes, the scoring indexes may further include at least one of the items: a readability of the comment, a useless attribute in the comment, and a commenting time.
The readability of the comment refers to an index of whether the comment is easy to read for the users. Normally, a sentence that is too long, a sentence with an overly short clause, or a sentence with too many clauses is considered as difficult to read for the users. Consequently, the readability is low. According to an example embodiment of the present disclosure, an example formula for calculating the comment readability is as follows:
In the formula, the symbol NL represents a normalized length of all of the clauses in the user's comment, and a length represents a number of characters in a specific clause of the user's comment. A clause is a part of a sentence that is separated by a punctuation mark. In an example embodiment of the present disclosure, a number of one Chinese character is 1, and a number of one English alphabet is also 1, e.g., a length of a term “my” is 2. In the formula, α represents a reasonable average value with respect to a readable sentence for ordinary people, and β represents a reasonable variance with respect to the readable sentence for ordinary people. Both α and β are constants. The formula (2) uses the normalized length of the user's comment to describe the readability of the sentence.
The useless attribute of the comment refers to an attribute that lowers the score in relation to the ranking of the comment.
According to an example embodiment of the present disclosure, the useless attribute may be predetermined and applicable to all categories, or set based on experiences or practical applications.
For example, while the useful attributes may be used for analysis purpose and displayed to the user, the useless attributes are only used for ranking. The user may click the useful attributes on a webpage so that the comments are sorted in an order according to the useful attributes.
According to an example embodiment of the present disclosure, when the comment includes the useful attributes, an example formula for calculating the ranking factor for the useless attributes of the comment is as follows:
In the formula, the symbol FS represents a ranking factor for the useless attributes, ε represents a number of useless attributes included in the comment, and δ and ε are constants. The constants δ and ε have constant values, and each is determined through multiple experiments so that the larger the ε value is, the smaller the score of formula (5) below becomes. The useless attributes may also be defined manually according to different ranking requirements. For example, most of the attributes displayed on some instant messaging tools are advertising attributes, which can be considered as useless attributes. In the case where an application does not desire too much logistical information, the attributes such as the speed of the logistic service or the service attitude of the logistic service are considered as the useless attributes. The useless attributes may be universal to all products, which are used to calculate the score of ranking and are not used for displaying the attribute tags to the user.
When the number of the useful attributes in the comment is 0, operations at 230 are performed. At 230, the scoring indexes are determined to include an index other than the useful attributes. According to an example embodiment of the present disclosure, the scoring indexes other than the index of useful attributes may includes at least one of the following items: the readability of the comment, the ranking factors for useless attributes in the comment, and the commenting time.
The useless attribute of the comment refers to an attribute that lowers the score of ranking for the comment. According to an example embodiment of the present disclosure, when the comment does not include useful attributes, an example formula for calculating the ranking factor for the useless attribute of the comment is as follows:
In the formula, the symbol FSN represents the ranking factor for the useless attributes, ε represents a number of useless attributes included in the comment, and δ and ε are constants. The constants δ and ε are constant values, and each is determined through multiple experiments so that the larger the ε value is, the smaller the score of formula (6) below becomes.
At 140, the comment is scored according to the scoring indexes. In the scoring, scoring indexes such as the number of the useful attributes, the readability of the sentence, and the useless attributes of the sentence, may be considered in the step of scoring.
According to an example embodiment of the present disclosure, when the comment contains useful attributes and the scoring indexes include the number of useful attributes, the readability of the comment, and the ranking factors for useless attributes in the comment, an example formula for calculating the score of final ranking is as follows:
score=SD/NL*FS (5)
In the formula, the symbol SD represents the number of the useful attributes, the symbol NL represents the normalized length of all clauses in the user's comment, and the symbol FS represents the ranking factor for the useless attributes which is calculated according to the formula (3).
According to another example embodiment of the present disclosure, the comment does not include the useful attributes and the scoring indexes include one or more scoring indexes other than the useful attributes. The scoring indexes other than the useful attributes may include: the readability of the comment and the ranking factor for useless attributes in the comment. In this case, an example formula for calculating the score of final ranking is as follows:
In the formula, the symbol “num” represents a number of the clauses, and θ is a constant which is determined by multiple experiments so that a ranking result may mostly meet an expectation. The symbol score represents the ranking score. The symbol NL represents the normalized length of all clauses in the comment which is calculated according to the formula (2) and the symbol FSN represents the ranking factor for the useless attributes which is calculated according to the formula (4). As shown in the formula (4), there will be more negative impacts to the comment or the score of the corresponding comment becomes lower when the number of the useless attribute is higher.
According to an example embodiment of the present disclosure, the commenting time may be considered as a factor during the operation of calculating the final ranking score. For example, the commenting time may be normalized and then used in calculating the final ranking score.
When a specific comment has a higher score, it indicates that the comment has more useful information that helps the user to make a decision, such as a shopping decision.
At 150, the comment is ranked based on the score. According to one example embodiment of the present disclosure, the comments are displayed in a descending order of scores from high to low.
At a displaying webpage 300, multiple displaying tags 302 are shown at a top portion of the webpage 300, and a list of comments is displayed below the displaying tag 302. When a button “all” 304 at the left of the displaying tag 302 is clicked by the user, all of the comments are displayed in a sorted order according to the present techniques. When one of the displaying tags 302 is clicked by the user, the comments corresponding to the corresponding displaying tag are displayed in an order sorted according to the present techniques.
In
It is noted that specific terms of the useful attribute 310 included in the comments are not necessary the same. For example, with respect to a useful attribute of product quality, one user may refer to it as “a product,” another user may refer to it as “a thing,” and another user may not even mention the product directly, etc.
Moreover, the emotional terms or descriptions used by the users may not necessarily be the same even if the same or similar emotional attitudes are expressed. For example, one user may use the term “very good,” another user may use the term “excellent,” and another user may user the term “worked as expected,” etc.
Therefore, although the meaning of combination of the useful attribute 310 and the emotional term 320 found in the comments may be the same or similar, the expressions may not necessarily be the same. As shown in
Such purpose may be fulfilled by the displaying tags 302. The displaying tags 302 may include two parts. One part is the normalized useful attribute, and the other part is the normalized emotional term. For example, the displaying tag “good quality” 306 as shown in the
According to an example embodiment of the present disclosure, the user may click one of the displaying tags 302 on the webpage to view the comments corresponding to the particular displaying tag. The comments are displayed in an order sorted or ranked by the method of the present disclosure.
According to another example embodiment of the present disclosure, portions corresponding to a displaying tag may be highlighted in the list of comments for display.
Thus, the present techniques not only rank the comments in consideration of the useful attributes of the target object but also display the ranked results according to different displaying tags, which further help the user to make a fast shopping decision.
It should be noted that the displaying tags as shown in
The apparatus 400 may include one or more processors or data processing units 402 and memory 404. The memory 404 may store therein a plurality of modules including an extracting module 410, an obtaining module 420, a determining module 430, a scoring module 440, and a ranking module 450. The extracting module 410 extracts one or more terms describing a target object from a comment in connection with the target object. The obtaining module 420 obtains one or more useful attributes included in the comment according to the terms describing the target object. The determining module 430 determines one or more scoring indexes that score the comment according to a number of the useful attributes. The scoring module 440 scores the comment according to the scoring indexes. The ranking module 450 ranks the comment according to the score.
According to an example embodiment of the present disclosure, the determining module 430 may further include a first determining sub-module that determines that the scoring indexes that score the comments includes the number of the useful attributes in the comments, when the number of the useful attributes is larger than or equal to 1.
According to one example embodiment of the present disclosure, the scoring indexes that score the comments may further include at least one of the items: a readability of the comment, a ranking factor of the useless attributes in the comments, and a commenting time.
According to an example embodiment of the present disclosure, the determining module 430 may further include a second determining sub-module that determines that the scoring indexes that score the comments includes one or more scoring indexes other than the useful attributes in the comments, when the number of the useful attributes is 0.
According to an example embodiment of the present disclosure, the scoring indexes other than the useful attributes may include at least one of the items: the readability of the comment, the ranking factor of the useless attributes in the comments, and the commenting time.
According to an example embodiment of the present disclosure, the apparatus 400 may further include a displaying module that displays one or more displaying tags to the user. The display tag may include the normalized useful attribute and the normalized term that describes the target object. The displaying tag may be clicked by the user and a sorted list of the user comments corresponding to the displaying tag is thereby displayed.
According to one example embodiment of the present disclosure, the target object may be a product, a service, or a merchant.
The present techniques may be typically applicable to e-commerce activities. However, the present techniques are not limited to such use and may be implemented in other Internet uses, such as a news website, a social network, etc., as long as any application that contains the comments of a specific matter or phenomenon, i.e. the target object. A person with ordinary skill in the art may readily think of other Internet applications of the present disclosure.
A person of ordinary skill in the art should understand that the embodiment of the present disclosure may be provided as a method, a system, or a product of computer program. Therefore, the present techniques may be implemented in a form of hardware, software, or a combination of hardware and software. In addition, the present disclosure may be implemented as the product of computer program that stored on a computer readable storage medium, which includes but not limit to: a disk memory, a CD-ROM or an optical memory, including computer executable instruction.
The embodiments described above are only example embodiments of the present disclosure, and are not intended to limit the scope of the present disclosure. Various modifications and alternations may be made to the present disclosure by a person with ordinary skill in the art. Any modifications, replacements and improvements that fall within the spirit and the principle of the present disclosure shall be included in the claim scope of the present disclosure.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2013 1 0242522 | Jun 2013 | CN | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 7389289 | Solaro | Jun 2008 | B2 |
| 8417713 | Blair-Goldensohn | Apr 2013 | B1 |
| 8645395 | Mushtaq | Feb 2014 | B2 |
| 8655938 | Smith | Feb 2014 | B1 |
| 8671098 | Salvetti | Mar 2014 | B2 |
| 8700480 | Fox | Apr 2014 | B1 |
| 8725495 | Peng | May 2014 | B2 |
| 8744855 | Rausch | Jun 2014 | B1 |
| 8799773 | Reis | Aug 2014 | B2 |
| 8918312 | Rehling | Dec 2014 | B1 |
| 8949211 | Lu | Feb 2015 | B2 |
| 8983975 | Kenton | Mar 2015 | B2 |
| 9129008 | Kuznetsov | Sep 2015 | B1 |
| 9135665 | England | Sep 2015 | B2 |
| 9342846 | Maddali | May 2016 | B2 |
| 20050193335 | Dorai | Sep 2005 | A1 |
| 20070038646 | Thota | Feb 2007 | A1 |
| 20070067294 | Ward | Mar 2007 | A1 |
| 20070156672 | Wolff et al. | Jul 2007 | A1 |
| 20080040383 | Kuwata et al. | Feb 2008 | A1 |
| 20090311658 | Polivka | Dec 2009 | A1 |
| 20100262597 | Han | Oct 2010 | A1 |
| 20110082848 | Goldentouch | Apr 2011 | A1 |
| 20110258556 | Kiciman et al. | Oct 2011 | A1 |
| 20120143597 | Mushtaq et al. | Jun 2012 | A1 |
| 20120259866 | Austin et al. | Oct 2012 | A1 |
| 20120278767 | Stibel et al. | Nov 2012 | A1 |
| 20130073545 | Jain | Mar 2013 | A1 |
| 20130151236 | Iofinov | Jun 2013 | A1 |
| 20130159413 | Davis | Jun 2013 | A1 |
| 20150052098 | Kveton | Feb 2015 | A1 |
| Number | Date | Country |
|---|---|---|
| 2008097286 | Apr 2008 | JP |
| 2008234090 | Oct 2008 | JP |
| WO2011019295 | Feb 2011 | WO |
| Entry |
|---|
| PCT Search Report and Written Opinion dated Jan. 19, 2016 for PCT Application No. PCT/US14/43032, 9 pages. |
| The European Office Action dated Apr. 24, 2018 for European Patent Application No. 14742007.9, a counterpart foreign application of U.S. Appl. No. 14/308,580, 6 pages. |
| The Chinese Office Action dated Dec. 27, 2016 for Chinese patent application No. 201310242522.8. a counterpart foreign application of U.S. Appl. No. 14/308,580, 8 pages. |
| The Chinese Office Action dated Aug. 10, 2017 for Chinese patent application No. 201310242522.8. a counterpart foreign application of U.S. Appl. No. 14/308,580, 9 pages. |
| The Extended European Search Report dated Feb. 9, 2017 for European Patent Application 14742007.9, 7 pages. |
| The Japanese Office Action dated Jun. 19, 2018, for Japanese patent application No. 2016-521555, a counterpart foreign application of U.S. Appl. No. 14/308,580, 6 pages. |
| Tanimoto et. al., “Evaluation of Evaluative Expression Dictionaries for Visualizing Reputation,” Proceedings of the 3rd Forum on Data Engineering and Information Management [online], The Institute of Electronics, Information and Communication Engineers, Data Engineering Technical Committee, Aug. 4, 2011, Retrieved from: Internet <URL: http://dbevent.jpn.org/deim2011/proceedings/pdf/f2-6.pdf> 9 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20140379682 A1 | Dec 2014 | US |
