Moscow, Russia
Moscow, Russia

Yandex is a Russian Internet company which operates the largest search engine in Russia with about 60% market share in that country. It also develops a number of Internet-based services and products. Yandex ranked as the 4th largest search engine worldwide, based on information from Comscore.com, with more than 150 million searches per day as of April 2012, and more than 50.5 million visitors daily as of February 2013. The company's mission is to provide answers to any questions users have or think about . Yandex also has a very large presence in Ukraine and Kazakhstan, providing nearly a third of all search results in those markets and 43% of all search results in Belarus.The Yandex.ru home page has been rated as the most popular website in Russia. The web site also operates in Belarus, Kazakhstan, Ukraine and Turkey. Another company, Yandex Labs, is a wholly owned division of Yandex that is located in the San Francisco Bay Area. In 2014, Yandex announced plans to open a research and development office in Berlin, Germany. Wikipedia.


Time filter

Source Type

There is disclosed a method for generating a content recommendation for a given user of a recommendation system. The method is executable at a recommendation server. The method comprises: receiving, by the recommendation server, from an electronic device associated with the given user a request for the content recommendation; responsive to the request generating, by the recommendation server, a set of content recommendations for the given user, the generating being executed by a prediction module of the recommendation server, the prediction module having been trained using a training set of training events, such that for each given training event from the training set of training events: at least one user-nonspecific feature is used as a first input parameter for the prediction module training, the at least one user-nonspecific feature having been retrieved from a latest version of a snapshot archive available at a time of the given training event occurring, the latest version of the snapshot archive having been generated prior to the time of the given training event occurring; at least one user-specific feature is used as a second input parameter for the prediction module training, at least one user-specific feature available at the time of the given training event occurring; transmitting at least a sub-set of the set of content recommendations to the electronic device.


A method of generating a recommended subset of items for a user of an electronic device, the method being executed at a server, the method comprises: acquiring, by the server, user events associated with a plurality of users, the user events comprising indications of user queries associated with the plurality of users; for each of the user queries, generating, by the server, a ranked predicted items list that comprises at least some items from a set of potentially recommendable items, such that each particular item within the ranked predicted items list has an associated rank; for each item within a plurality of ranked predicted items lists, generating, by the server, an item score based on a totality of ranks associated therewith; acquiring, by the server, a request for the recommended subset of items; and generating, by the server, the recommended subset of items from the set of potentially recommendable items, the generating the recommended subset of items comprises selecting, by the server, at least one item within the plurality of ranked predicted items lists as the recommended subset of items based on the item scores of the items within the plurality of ranked predicted items lists.


A method of generating a recommended subset of items for a user of an electronic device, the method being executed at a server, the method comprises: identifying, by the server, a first subset of items within a set of potentially recommendable items based on item features of items within the set of potentially recommendable items; acquiring, by the server, a request for the recommended subset of items; identifying, by the server, a second subset of items within the set of potentially recommendable items based on user events associated with the user, each item within the second subset of items being different from any item within the first subset of items; and generating, by the server, the recommended subset of items, the recommended subset of items comprising at least some items from the first subset of items and at least some items from the second subset of items.


Patent
Yandex | Date: 2017-04-05

There are disclosed methods and systems for text-to-speech synthesis for outputting a synthetic speech having a selected speech attribute. First, an acoustic space model is trained based on a set of training data of speech attributes, using a deep neural network to determine interdependency factors between the speech attributes in the training data, the dnn generating a single, continuous acoustic space model based on the interdependency factors, the acoustic space model thereby taking into account a plurality of interdependent speech attributes and allowing for modelling of a continuous spectrum of the interdependent speech attributes. Next, a text is received; a selection of one or more speech attribute is received, each speech attribute having a selected attribute weight; the text is converted into synthetic speech using the acoustic space model, the synthetic speech having the selected speech attribute; and the synthetic speech is outputted as audio having the selected speech attribute.


A system for and a method of generating a first and a second simplified borders of a first and a second graphical objects having respectively a first and a second original borders comprising curved border portions being located in a close proximity, the method executable on a computing device, the method comprising: applying a divider having cells (906) to both the first and the second graphical objects such that both original borders are split into a plurality of original fragments, at least some of them being original curved fragments (902, 904); anchoring intersections (602, 604, 622, 624) of the first and the second original borders with the cells (906); responsive to at least one cell houses two curved fragments of different graphical objects, generating instructions to render both simplified objects by using one graphical element for rendering similar simplified fragments (1002, 1004) in both objects, if both simplified curved borders are similar.


A computer-implemented system for processing a user device request to process a user data portion, the system comprising a server having a processor, the processor having a user space and a kernel space, the processor configured to perform receiving the request to process the user data portion from a user device, reading the user data portion from a database at the server, allocating space at the processor to define a sandbox environment defining a kernel space commands set of the processor to perform processing of the user data portion, isolating the processor within the sandbox environment in order to perform isolated execution of the request by the kernel space commands set, processing the user data portion within the sandbox environment, de-isolating the sandbox environment from the user space by returning an indication of a processed user data portion and writing the indication to the user space of the processor.


There is provided a method and a system for conducting a search and presenting results. The method can be executed at a server. The method comprises receiving a search query from an electronic device associated with a user; responsive to the search query, generating a search query result set, the search query result set including a vertical search result; determining a confidence level that the vertical search result is the most relevant to the search query; responsive to the confidence level being above a pre-determined threshold, causing the electronic device to display exclusively the vertical search result.


A method and system is described including presenting a first SERP version to a first set of users and a second SERP version to a second set of users; assessing first and second measures of user interactions with the first and second version of the SERP respectively, the user interactions being of a pre-selected type; computing a first and a second distribution of the first and second measures of user interactions and analyzing the first and second distributions conjointly for determining a change in user interactivity. The analyzing can include determining a set of ratios; determining a lowest ratio indicative of a smallest relative change and a highest ratio being indicative of a largest relative change within the set of ratios; determining the magnitude of the change in user interactivity based on the lowest and highest ratios; and determining a significance of the magnitude of the change in user interactivity.


A method of ranking elements of a first network resource for a first user is disclosed. The method is executable on a server, the method comprising receiving an indication of the elements of the first network resource, receiving an indication of a client device of the first user, responsive to the indication of the elements at least partially matching an indication of elements of a second network resource and the indication of the client device of the first user at least partially matching an indication of a client device of a second user, the second user having engaged on the client device in a past interaction with the elements of the second network resource, acquiring information associated with the past interaction of the second user, determining a ranking of the elements of the first network resource by relevance to the first user and a most relevant one of the elements.


A method and system for processing a user request for a recommended area of interest includes the steps of receiving the request including an indication of an electronic device geo-location and a user defined search constraint; receiving data associated with photographs associated with geo-objects, the data comprising geo-location coordinates of the photographs, the geo-location coordinates of the photographs being in proximity with the device geo-location; computing a plurality of region representations based on the geo-location coordinates of the photographs, each region representation being associated with a unique photograph density calculation parameter, the computing comprises determining a potential area of interest in each region representation, each region representation being a candidate for an optimal region representation; determining the optimal region representation based on the user defined search constraint; and displaying to the user the recommended area of interest that corresponds to the potential area of interest of the optimal region representation.

Loading Yandex collaborators
Loading Yandex collaborators