Abstract: Technologies are described herein for extracting quotes from customer reviews regarding collections of items. An identifier of an individual item is received and customer reviews regarding a collection of items containing the individual item are retrieved. A collection of sentences is parsed from the retrieved customer reviews, and sentences that discuss the individual item are identified. A quote is extracted from the identified sentences and displayed to a customer in conjunction with information regarding the individual item. Similarly, individual items belonging to a collection of items are identified, and customer reviews regarding the individual items are retrieved. A collection of sentences are parsed from the retrieved customer reviews, and those sentences that discuss topics relevant to the collection of items are identified. A number of quotes are extracted from the identified sentences and displayed to a customer in conjunction with information regarding the collection of items.

Abstract: The top rated products in a product category are identified by computing a score for each product in the category. The score is computed by calculating a time decay factor for each product category. The time decay factor is utilized to decay the ratings of each product as a function of time. A relevance value is also computed for each product. The relevance value is utilized to modify the ratings for each product in view of the recency of the ratings. The score may be computed for each product using the ratings for the product, the computed time decay factor, and/or the relevance value. The computed scores are utilized to rank the products in the product category.

Abstract: Techniques for calculating edge strengths for edges of a social graph are described herein. These techniques may determine an affinity between a first user of a social graph and a second user of the social graph. Based at least in part on this affinity, a strength of an edge connecting the first user and the second user on the social graph may be calculated and assigned to the corresponding edge.

Abstract: Technologies are described herein for classifying sentences or phrases as expressing positive or negative sentiment based on machine learning from training data comprising sentences manually labeled as to sentiment. A list of terms is generated from the manually labeled sentences and sentiment scores are determined for the terms in the list of terms based on the manually labeled sentences. A collection of sentences or phrases may then be classified as to sentiment utilizing one or more logistic regression classifiers trained on the sentiment scores determined for the terms in the list of terms. The classified collection of sentences may be further analyzed to determine an overall majority sentiment regarding a topic discussed in the sentences and/or to extract specific sentences or phrases expressing a particular sentiment for display to a customer.

Abstract: An interactive system enables users to generate and submit lists of items (e.g., products) for viewing by other users. The user-generated lists are selectively suggested to users of the system based on their respective behaviors, such as their item viewing activities, searches, and/or item purchases. The system may monitor the behaviors of users who access particular user-generated lists, and based on the monitored behaviors, generate measures of the effectiveness of particular lists (e.g., effectiveness at driving item purchases). The measures of effectiveness may be considered in selecting particular lists to suggest.

Abstract: Disclosed are various embodiments generating network pages for search engines from data protected from search engines. A portion of data from a corpus of data that is protected from indexing by a search engine is extracted. A first network page is generated based at least in part on the portion of data. The first network page is configured for the search engine to index the portion of data. The first network page omits a context for the portion of data from the corpus of data. The first network page includes one or more links to a second network page that is protected from indexing by the search engine. The second network page provides access to the corpus of data.

Abstract: Systems and methods are provided for providing an answer to a question left unanswered in an electronic discussion forum. In some embodiments, a question that has not been answered may be identified in a first discussion forum. A second discussion forum in which to present the question may be determined, where the second discussion forum is more likely than the first discussion forum to result in an answer to the question. The question may then be presented in the second discussion forum, and one or more responses to the question may be received in the second discussion forum. It may then be determined whether one of the one or more responses includes a good answer to the question, and when the determination is that one of the one or more responses includes a good answer, the good answer may be presented in the first discussion forum.

Abstract: Improved user interfaces and methods are provided for presenting item recommendations to a user when the user selects an item to add to a shopping cart. In response to the user's selection, a page generation process generates and returns a page that includes a plurality of recommendations sections, each of which displays item recommendations generated using a different respective recommendation or selection algorithm (e.g., recommendations based on shopping cart contents, recommendations based on purchase history, etc.). These recommendation algorithms may be selected based on collected information regarding the user. The page may also include a shopping cart section which displays contents of the shopping cart.

Abstract: Techniques for embedding metadata into a piece of content are described herein. With use of the embedded metadata, an application takes one or more actions specified by the embedded metadata upon selection of the content. In some instances, the content comprises an image, video, or any other form of content that a user may consume. Using the example of an image, the techniques may embed metadata within the image to create an image file that includes both the image and the embedded metadata. Then, when an application of a computing device selects (e.g., receives, opens, etc.) the image file, the application or another application may perform one or more actions specified by the metadata.

Abstract: Disclosed are various embodiments of generating item recommendations. A user submitted query associated with a category of items in an electronic repository is received, each of the items being associated with reviews. Reviews relevant to the user submitted query are identified. Reviews relevant to the user submitted query are displayed with a user voting element. The user voting element allows a user to vote whether the review is relevant to the user query. Further relevance to the user submitted query is determined based at least upon the vote.

Abstract: Systems and methods are provided for providing an answer to a question left unanswered in an electronic discussion forum. In some embodiments, a question that has not been answered may be identified in a first discussion forum. A second discussion forum in which to present the question may be determined, where the second discussion forum is more likely than the first discussion forum to result in an answer to the question. The question may then be presented in the second discussion forum, and one or more responses to the question may be received in the second discussion forum. It may then be determined whether one of the one or more responses includes a good answer to the question, and when the determination is that one of the one or more responses includes a good answer, the good answer may be presented in the first discussion forum.

Abstract: Improved user interfaces and methods are provided for presenting item recommendations to a user when the user selects an item to add to an electronic shopping cart. In response to the user's selection, a page generation process generates and returns a page that includes a condensed shopping cart portion and a recommendations portion. The condensed shopping cart portion displays the contents of the shopping cart, but omits some of the item information and/or controls provided on a shopping cart page, thereby preserving more screen real estate for the display of item recommendations. The condensed shopping cart portion also includes a link to the shopping cart page.

Abstract: Tag suggestions enable a hosting entity such as a website to determine one or more tags to suggest to a user for association with a particular item within an electronic catalog. After this determination, the hosting entity may suggest the determined tags to the user. To determine these tags, the hosting entity may employ techniques to determine items related to the particular item. The hosting entity then suggests some or all of the tags associated with the related items. Additionally or alternatively, the hosting entity may determine certain metadata associated with the particular item. The entity then may suggest this metadata, or some related phrase or tag, to the user for association with the particular item. However the tag suggestions are determined, the hosting entity may rank the tag suggestions to determine which tags to present to the user or to determine an order in which to present the tags.

Abstract: Customer-annotated catalog pages enable a user of a merchant website to annotate an image that illustrates one or more items found within an electronic catalog. This annotation may include associating an illustrated item with a hyperlink to a page that is associated with the illustrated item. This annotation may also include commenting on the illustrated item or describing the image as a whole. In some instances, the user may herself upload the image to the merchant web site to enable this annotation. Additionally, other users of the merchant website may comment on the image, the illustrated items, or may associate hyperlinks to associated pages for one or more of the illustrated items. By providing such customer-annotated catalog pages, users of the merchant website are able to view a group of potentially-related items and are able to easily navigate to these items' associated pages within the electronic catalog or otherwise.

Abstract: Tag suggestions enable a hosting entity such as a website to determine one or more tags to suggest to a user for association with a particular item within an electronic catalog. After this determination, the hosting entity may suggest the determined tags to the user. To determine these tags, the hosting entity may employ techniques to determine items related to the particular item. The hosting entity then suggests some or all of the tags associated with the related items. Additionally or alternatively, the hosting entity may determine certain metadata associated with the particular item. The entity then may suggest this metadata, or some related phrase or tag, to the user for association with the particular item. However the tag suggestions are determined, the hosting entity may rank the tag suggestions to determine which tags to present to the user or to determine an order in which to present the tags.

Abstract: An interactive system enables users to generate and submit lists of items (e.g., products) for viewing by other users. The user-generated lists are selectively suggested to users of the system based on their respective behaviors, such as their item viewing activities, searches, and/or item purchases. The system may monitor the behaviors of users who access particular user-generated lists, and based on the monitored behaviors, generate measures of the effectiveness of particular lists (e.g., effectiveness at driving item purchases). The measures of effectiveness may be considered in selecting particular lists to suggest.

Abstract: Information retrieval systems and methods are disclosed for presenting items of interest to a user of a computer-based data repository, where the items are presented in the form of lists generated by other users. Users of a data repository may generate lists of items from the repository that are accessible by other users. In various embodiments, users generating lists may add supplemental commentary about the items and other personalizing information. The information retrieval systems and methods store the user-generated lists for future presentation to other users. When a user of the data repository submits a search query, items returned in the query's results are interpreted to be items of interest to the user, and in response, one or more lists determined to be similar to the items of interest are presented to the user.

Abstract: Creation and placement of tag forums for tags enables a user to discuss aspects of both tags and elements associated with the tags, and further enables a user to view discussions from other users. The tag forum may be generated and/or placed based in part on user interaction or other factors related to the elements and the tags. Tag forums may be placed on, for example, selective tag web pages or associated element web pages within a website, among other possibilities.

Abstract: Creation, association, and/or placement of search forums for searches strings enable users to discuss aspects of both searches and elements associated with the searches, and further enable the users to view discussions from other users. The search forum may be generated from variations of the search string. In addition or alternatively, search strings may be associated with existing item forums, category forums, and tag forums. Search forums may be placed with, for example, search results or with items, categories, and/or tags, among other possibilities, to provide user access to the search forums.

Abstract: Information retrieval systems and methods are disclosed for presenting items of interest to a user of a computer-based data repository, where the items are presented in the form of lists generated by other users. Users of a data repository may generate lists of items from the repository that are accessible by other users. In various embodiments, users generating lists may add supplemental commentary about the items and other personalizing information. The information retrieval systems and methods store the user-generated lists for future presentation to other users. The online activities of a user of the data repository are monitored, and, in response to the user indicating an interest in an item, one or more lists determined to be similar to the indicated item are presented to the user.