Abstract: Methods and systems disclosed herein describe customizing searching. Search queries may be customized according to a user's preferences. A user may emphasize or indicate that additional weight should be given to one or more terms in a search query. Terms that are weighted higher may have a larger impact on the results that are returned in response to the search query. In addition to changing the terms in a search query, a user may provide a weight for each term. Each term in a search query may be weighted to varying degrees, giving a user more control over the results that are returned. The weights may be used with machine learning techniques to generate a vector representation of a search query. The vector representation of the search query may be compared with vector representations of search objects to determine results that match the search query.

Abstract: Described are natural language web browsers configured to provide a natural language interface to an existing web browser or a newly created web browser. That is, users interact with the web browser employing speech (or text) as input and output instead of using a mouse and keyboard as input and an electronic display screen as output. Embodiments of a natural language web browser converts user questions, statements, and/or commands into actions, reads the resulting HTML and converts the HTML into natural language descriptions to provide to the user.

Abstract: Prioritizing crawled data in a document store for reprocess operations. Reprocessing occurs upon a triggering change to configurations. Prioritization is based on the status of the crawled data with respect to an ACL. During reprocessing, the crawled data is reprocessed in an order defined by assigned priority levels.

Abstract: A computer-implemented method for enabling static markup crawlability for a web application dynamically updating web pages client-side may be provided. The method comprises providing a web application identifiable by a web address comprising both, a static content block and dynamic content, and generating a static HTML page and pre-rendering the static content block of the web application such that at least a portion of static content of the web application is crawlable.

Abstract: Utilizing a computing device to automatically locate a link to a target web page whose link from a referring web page is broken. A computing device receives a request to access a link to a target web page from a referring web page. The computing device receives a notification indicating that the link to the target web page is broken. The computing device locates a new link to the target web page utilizing one or more keywords. The computing device provides access to the target web page.

Abstract: A system for returning synthetic database query results. The system may include a memory unit for storing instructions, and a processor configured to execute the instructions to perform operations comprising: receiving a query input by a user at a user interface; determining, based on natural language processing, a type of the query input; determining, based on the received query input and a database language interpreter, an output data format; returning, based on a generation model and the output data format, a result of the query input; providing, to a plurality of training models and based on the determined query type, the query input and the result; and training the training models, based on the query input and the result.

Abstract: A system and method for crawling a wide area computer network for retrieving contextual information. The system includes a data processing arrangement including a communication interface for accessing the wide area computer network and a web crawling module. The web crawling module is operable to receive an input file including plurality of Uniform Resource Identifiers (URIs) via a user interface, identify accessible URIs from the plurality of accessible URIs, obtain one or more features associated with accessible URIs, and execute an event for crawling the accessible URIs, the event includes creating resource cluster for executing the event, acquiring electronic record file to be executed by the resource cluster, and crawling the accessible URIs, furthermore, the web crawling module processes the extracted web content to identify contextual data. The system further includes a database arrangement communicably coupled to the data processing arrangement to aggregate contextual data therein.

Abstract: A system is provided that can execute one or more scripts to cause a discrepancy resolution processor to be executed on a first exported image file (having a raster-based image format such as PNG), and a second exported image file (having a vector-based image format such as SVG) to programmatically compare, fix, and flag discrepancies or differences between the first exported image file and the second exported image file.

Abstract: Embodiments seek to improve prefetch hinting using time-dependent, machine-generated hints. Some embodiments operate in context of client machines in communication hinting machines that can develop information about whether and how resources are used in network transactions over time by collecting “resource samples.” Each resource sample can identify rendering status information of a resource at a sample time. The time-based samples can be used to compute time-based probabilities for the resources, indicating, for example, the likelihood of a resource being used to render a web page at some subsequent time. Time-dependent hints can be generated as a function of the time-based probabilities, and the time-dependent hints can be used to improve prefetching by optimizing the hinting information with respect to a particular request time (e.g., the prefetching hints for rendering a web page can be generated in a manner that accounts for when the web page is being rendered).

Abstract: A computer system for identifying execution paths of analytic elements comprises computer-executable instructions that configure the computer system to identify an orphan analytic element and a second analytic element associated with a network-connected software application. The system can also be configured to identify, with a computer processor, one or more common attributes associated with the orphan analytic element and the second analytic element. Based upon the one or more common attributes, the computer system can identify an execution path for the orphan analytic element. Additionally, the system can be configured to execute, at the one or more computer processors, the network-connected software application, capture network communications generated by the network-connected software application, and generate observed execution paths based on the captured network communications.

Abstract: A retrieval method including outputting a first retrieval result to a result area, the first retrieval result being obtained by referring to a first database based on a first retrieval query entered into a retrieval area, outputting a second retrieval result and the first retrieval result in the result area, the second retrieval result being obtained by referring the first database based on a second retrieval query in response to acquiring a delimiter character and the second retrieval query entered next to the first retrieval query in the retrieval area.

Abstract: Techniques for determining user authorization to access data in dynamically generated database structures are presented. A method can commence upon receiving database statements from a user that is authorized to access a set of data in a database. A physical plan derived from the received database statements to operate on the database is inspected for references to dynamically generated data structures such as aggregates. A modified physical plan having no aggregates and/or other altered database structures is used to issue low latency database statements to operation on the database for verifying user access authorization to the underlying data. In some cases, the foregoing database statements are based on a virtual multidimensional data model. In other cases, a low latency directive is included in the modified physical plan to facilitate a low latency authorization response.

Abstract: Aspects of the disclosure relate to resource allocation and rebating during in-flight data masking and on-demand encryption of big data on a network. Computer machine(s), cluster managers, nodes, and/or multilevel platforms can request, receive, and/or authenticate requests for a big data dataset, containing sensitive and non-sensitive data. Profiles can be auto provisioned, and access rights can be assigned. Server configuration and data connection properties can be defined. Secure connection(s) to the data store can be established. Sensitive information can be redacted into a sanitized dataset based on one or more data obfuscation types. State point information for previously reached safe points can be stored and progressively released such that only the incomplete portion(s) of task(s) need to be resubmitted. The encrypted data can be transmitted, in response to the request, to a source, a target, and/or another computer machine and can be decrypted back into the sanitized dataset.

Abstract: Embodiments of this application disclose a service type determining method and a related device. A first device obtains a dataset, where the dataset includes at least two 5-tuples. The first device determines a target feature (the target feature is used to indicate a service type corresponding to a target triplet) corresponding to the target triplet based on the dataset. In this way, the first device can identify the service type corresponding to the target triplet based on the target feature, to facilitate operations such as packet transmission traffic control or application identification based on the importance indicated by the service type.

Abstract: Aspects of the disclosure relate to resource allocation and rebating during in-flight data masking and on-demand encryption of big data on a network. Computer machine(s), cluster managers, nodes, and/or multilevel platforms can request, receive, and/or authenticate requests for a big data dataset, containing sensitive and non-sensitive data. Profiles can be auto provisioned, and access rights can be assigned. Server configuration and data connection properties can be defined. Secure connection(s) to the data store can be established. The big data dataset can be uncompressed based on a codec and uncompressed data blocks can be distributed for processing. Sensitive information can be redacted into a sanitized dataset based on one or more data obfuscation types. The encrypted data can be transmitted, in response to the request, to a source, a target, and/or another computer machine and can be decrypted back into the sanitized dataset.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for identifying broken network connections. In one aspect, a system includes front-end server(s) that receive data specifying, for multiple different user interactions with one or more application links that link to a given application, presentation durations specifying how long application content linked to by the application link was presented following the multiple different user interactions with the application link(s). Back-end server(s) that communicate with the front end server(s) can classify each application link as broken or working based on application of a machine learning model to the presentation durations for the application link. The machine learning model can be generated using labeled training data. The back-end server(s) can generate and output an alert identifying an application link as a broken link based on the application link being classified as broken by the machine learning model.

Abstract: For a multi-tenant database accessible by a plurality of separate organizations, a system is provided for capturing organization specificities in a model for the multi-tenant database. The system includes a neural network. The system is configured to: receive an organization encoding for one or more separate organizations making previous search queries into the multi-tenant database; generate a vector matrix from the organization encoding to embed organization specificities for training a model of the neural network; and using the vector matrix, train the model of the neural network for processing a present search query into the multi-tenant database. In some embodiments, the model of the neural network is global across the separate organizations accessing the database.

Abstract: A method and an apparatus for provisioning optimized content to customers are disclosed. The method includes determining at least one attribute associated with a customer active on a web interface associated with an enterprise. A plurality of baseline contents and a plurality of content elements are accessed from a database and at least one baseline content and at least one content element are selected based on the at least one attribute associated with the customer. A customized advertisement is generated using the at least one baseline content and the at least one content element.

Abstract: A method, product, and system for inserting intermediate content page. An exemplary method comprises displaying to a user a first webpage followed by a second webpage; and in response to an instruction to return to a previous location from the second webpage, displaying an intermediate content webpage that comprises a feed element displaying sponsored or organic content. Another method comprises monitoring browsing activity of a user, in which the user reaches from a first webpage to a second webpage, and causing an intermediate content webpage to be displayed the user instead of the first webpage in response to a user instruction to implement the previous location functionality at the second webpage.

Abstract: A server receives a query identifying one or more attributes of an entity. The server accesses a set of candidate templates for answering the query based on the one or more attributes of the entity, each candidate template having one or more fields, wherein each field is associated with at least one constraint. The server obtains a set of information that answers the query and selects a template from the set of candidate templates, where the selected template has a maximum number of fields with constraints that are satisfied by the set of information. The server then generates a phrase by adding the set of information to the one or more fields of the selected template, such that the phrase comprises an answer to the query. Finally, the server communicates the phrase to a client device.

Abstract: Aspects of the disclosure relate to resource allocation and rebating during in-flight data masking and on-demand encryption of big data on a network. Computer machine(s), cluster managers, nodes, and/or multilevel platforms can request, receive, and/or authenticate requests for a big data dataset, containing sensitive and non-sensitive data. Profiles can be auto provisioned, and access rights can be assigned. Server configuration and data connection properties can be defined. Secure connection(s) to the data store can be established. Sensitive information can be redacted into a sanitized dataset based on one or more data obfuscation types. RAM requirements and current RAM allocation can be diagnosed. Portion(s) of the current RAM allocation exceeding the RAM requirements can be rebated. The encrypted data can be transmitted, in response to the request, to a source, a target, and/or another computer machine and can be decrypted back into the sanitized dataset.

Abstract: Implementations of the present disclosure include actions of receiving a search query, identifying potential search results responsive to the search query, the potential search results corresponding to digital content stored in one or more computer-readable storage media, determining that the potential search results include user-generated content that is generated using one or more computer-implemented social services, receiving data associated with the user-generated content, the data including one or more scores, determining, based on the one or more scores, that the user-generated content is to be provided as a search result, generating search results, the search results including web-based search results and at least a portion of the user-generated content, and transmitting the search results to a client computing device for display to the searching user.

Abstract: Character string analysis and classification can be useful in a variety of contexts, including examining web URLs to determine whether a URL indicates that a user is attempting to take a particular action on an electronic service platform. In some cases, however, URLs or other string data may have “noise” in them, such as random sub-strings, that prevents a string from being properly classified. Sometimes it may be useful to classify a string into a category, however, and it may be important to do this quickly (e.g. during an active user interaction with a website). Learning tables allowing for O(1) lookup can be established by tokenizing strings and then using probability analysis to eliminate tokens that appear an insufficient number of times. This allows for quick and accurate string classification, which may be useful in numerous circumstances.

Abstract: An extraction unit-extracts a specific request from among requests that do not match with a profile on the basis of a similarity to a request to a server, where the profile determines whether the request is an attack. Further, a determination unit determines whether the specific request extracted by the extraction unit meets a predetermined condition indicating that the specific request is continuously transmitted from a certain number or more of transmission sources. Furthermore, a control unit relearns the profile if the determination unit determines that the specific request meets the predetermined condition.

Abstract: In an approach to personalizing a news feed, a computing device monitors a user accessing content. The computing device determines a personal knowledge graph for the user based on the accessed content. Responsive to receiving a new piece of content that the user has not accessed, the computing device determines a novelty score for the new piece of content based on the personal knowledge graph. The computing device filters the new piece of content based on the novelty score.

Abstract: The present invention relates to a deep learning system suitable for searching data sources for specific content. In particular, the present invention relates to an unconventional machine-implemented process, leveraging a machine learning algorithm, to provide a technology that searches data sources and recognizes feature data associated with one or more predetermined classifiers.

Abstract: Disclosed are methods and systems for ranking content. In one aspect, a method of ranking content for display includes identifying, via hardware processing circuitry, interactions by a single account with content pairs, each of the content in the content pairs included in a plurality of content, aggregating, via the hardware processing circuitry, the identified interactions across a plurality of accounts, associating, via the hardware processing circuitry, probabilities with each content in the plurality of content based on the aggregated interactions, ranking, via the hardware processing circuitry, the plurality of content based on the associated probabilities; and selecting, via the hardware processing circuitry, content ranked above a threshold for display.

Abstract: Systems, methods, and computer-readable storage media for categorizing items based on attributes of the item and a shape of the item, where the shape of the item is determined from an image of the item. An exemplary system configured as disclosed herein can receive a request to categorize an item, the item having a plurality of attributes, and receive an image of the item. The system can identify, via a processor configured to perform image processing, a shape of the item based on the image, and transform the plurality of attributes and the shape of the item, into a plurality of quantifiable values. The system can then categorize the item based on the quantifiable values.

Abstract: Systems and methods of task implementation are extended as provided herein and target the web crawling process through a step of submitting a request by a customer to a web crawler. The systems and methods allow a more complex request for a web crawler to be defined in order to receive more specific data. In one aspect, a method for data extraction and gathering from a Network by a Service provider infrastructure include the following steps: checking the parameters of a request received from a User's Device, adjusting the request parameters according to pre-established Scraping logic, selecting a Proxy according to the criteria of the pre-established Scraping logic, sending the adjusted request to the Target through the selected Proxy, checking metadata received from the Target, and forwarding the data to the User's device.

Abstract: An example system can include a server that includes or is associated with a keyword recommendation module. The keyword recommendation module can be configured to select keywords for a search engine for use in a search engine marketing campaign, wherein the search engine provides more traffic to Internet content of a publisher than other search engines, for the keywords. The module can also be configured to determine, per keyword, user engagement with the Internet content resulting from the traffic provided by the search engine for the keywords, according to one or more of time spent viewing the Internet content, page views of the Internet content, and dwell times. The module can also be configured to score, per keyword, the keywords according to the determined user engagement with the Internet content, and generate keyword recommendations according to the scoring of the keywords.

Abstract: A method of identifying and presenting accounts and postings of a social media system may include: retrieving, via at least one communication interface of the social media system, account information for a plurality of social media accounts; classifying, based on the retrieved account information, the plurality of social media accounts according to account type; ranking, based on the retrieved account information, the plurality of social media accounts according to perceived importance or influence; storing, in an account database, a representation of a subset of the plurality of social media accounts having a ranking greater than a predetermined threshold; retrieving, via the at least one API, postings of the stored social media accounts; and generating, for display on a user system, a user interface to present the retrieved social media postings for the stored accounts organized according to at least one of: account type, or ranking.

Abstract: The present teaching, which includes methods, systems and computer-readable media, relates to ranking content from multiple disparate sources including a person's personal data sources and non-personal data sources. The disclosed techniques may include obtaining a plurality sets of content associated with a request from a person, each of which being from a separate data source, and applying a model for each set of content to obtain a set of features for each piece of content in the set of content, wherein the model is specific to a data source from where the set of content comes from. Each set of features for each piece of content of the set of content may be normalized with respect to a common space to generate a normalized feature set. Further, a score for each piece of content from a set of content may be estimated based on the normalized feature set for the piece of content, and based on the score of the piece of content, each piece of content of the plurality sets of content may be ranked.

Abstract: Disclosed is a system for context-based abbreviation disambiguation, the system comprising: an ontological databank represented into a multi-dimensional space, a synonym databank, a glossary databank, and a server arrangement. The server arrangement is configured to obtain a text comprising abbreviations and concept phrases, extract a target abbreviation from the abbreviations, obtain potential expansions for the target abbreviation, calculate a synonym match score for potential expansions, using synonyms of the potential expansions and the concept phrases, calculate a concept match score, using concepts relating to the potential expansions and the concept phrases, calculate a context match score for the potential expansions using a comparison module, and determine one of the potential expansions as a valid expansion of the target abbreviation based on at least one of the: synonym match score, concept match score and context match score.

Abstract: Systems and methods are described for generating criteria for a plurality of web applications in an online application store, translating the criteria into at least one rule, the at least one rule based on predefined categories defined by the online application store, obtaining, metadata associated with a plurality of websites, determining, using the metadata and the at least one rule, whether any of the websites in the plurality of websites, includes code that executes a feature associated with the at least one rule, and displaying the icon as a selectable listing in the online application store.

Abstract: A first compute server of a distributed cloud computing network receives a request from a first client device for an object to be handled by an object worker that includes a single instantiation of a piece of code that solely controls reading and writing access to the first object. A determination is made that the object worker is instantiated for the object and is currently running in the first compute server, and the piece of code processes the first request. The first compute server receives a message to be processed by the first object worker from a second compute server. The message includes a second request for the object from a second client device connected to the second compute server. The piece of code processes the message and transmits a reply to the second compute server.

Abstract: A hybrid docking station determines whether native video data exists and can be passed through to a video port or whether a virtual video processor should be activated to provide virtual video data to a video port. For example, a laptop is connected to a hybrid docking station using a USB™ 3.0 connection. The hybrid docking station recognizes that the USB™ 3.0 connection includes a native video data and passes the native video data to a DisplayPort™. By avoiding activating a virtualized video processor and using native video data, the laptop avoids installing software to communicate with the virtualized video processor and communicates with one or more displays using a native video channel. By avoiding installing software, it simplifies IT's and user's usage and experience with universal docking station.

Abstract: Techniques for providing explanations of candidate search queries are described. The queries can be created using query intent clustering in an automated sourcing tool. In an example embodiment, disclosed is a system that obtains one or more current candidate member profiles used as a basis for a search on member profiles in an online system. Additionally, the system extracts one or more attributes from the one or more current candidate member profiles. Moreover, the system identifies query intent clusters based on the extracted one or more attributes. Furthermore, the system generates a search query based on the identified query intent clusters. Next, an explanation of the search query can be displayed on a display device of the system. In some embodiments, the online system hosts a social networking service that includes the member profiles, and the identified query intent clusters include skills clusters.

Abstract: Systems and methods for generating search results based on non-linguistic tokens are disclosed. In embodiments, a computer-implemented method comprises: mapping, by a computing device, a non-linguistic token to documents during indexing based on associations in a token database between the non-linguistic token and variations of a multi-word term; receiving, by the computing device, the multi-word term in a search query; determining, by the computing device, the non-linguistic token associated with the multi-word term using the token database; and generating, by the computing device, search results based on the multi-word term and the non-linguistic token.

Abstract: Systems and methods of task implementation are extended as provided herein and target the web crawling process through a step of submitting a request by a customer to a web crawler. The systems and methods allow a more complex request for a web crawler to be defined in order to receive more specific data. In one aspect, a method for data extraction and gathering from a Network by a Service provider infrastructure include the following steps: checking the parameters of a request received from a User's Device, adjusting the request parameters according to pre-established Scraping logic, selecting a Proxy according to the criteria of the pre-established Scraping logic, sending the adjusted request to the Target through the selected Proxy, checking metadata received from the Target, and forwarding the data to the User's device.

Abstract: A method and a system for performing topic-based aggregation of web content. The method includes receiving a plurality of Uniform Resource Identifiers, wherein each of the plurality of Uniform Resource Identifiers points to a web location related to a subject-matter; crawling the web location, to fetch web content related to the subject-matter, wherein the web content includes one or more elements; generating an electronic record file for storing the web content, and meta-information associated with the crawling of the web location; identifying a first type and at least a second type of the one or more elements; processing the first type of element to obtain a topic-based web content corresponding to a category associated with the subject-matter, and the at least one second type of element to obtain operational data; and aggregating the topic-based web content into a database.

Abstract: Examples relate to systems and methods for dynamically generating web pages for displaying database items based on metadata included in network URL requests. Some examples disclosed herein may enable receiving a URL comprising a plurality of segments, comparing each segment of the plurality of segments to known segment values, identifying, based on the comparison, searchable segments and one or more searchable categories for the searchable segments, independent of the order of the searchable segments within the URL, determining a search pattern based on the searchable segments and searchable segment categories, performing an inventory search based on the search pattern to return a list one or more inventory items, building web page metadata items based on the search pattern, and generating a web page comprising the web page metadata items and the list of inventory items.

Abstract: Altitude based device management is provided herein. A method can comprise transmitting, by a mobile device comprising a processor, a signaling message to a network device of a wireless network. The signaling message can comprise first data indicating a device type of the mobile device and second data indicating a distance measurement of the mobile device with respect to a reference point. The method can also comprise implementing, by the mobile device, a first instruction related to a power setting and a second instruction related to an operating parameter. The first instruction and the second instruction can be received from the network device and can be based on the device type of the mobile device and the distance measurement of the mobile device.

Abstract: Various methods and systems for processing web crawling queries using a web crawling prioritization model based on classification operation performance. A classification operation for organizing products in a product listing platform is accessed. A web crawling engine is accessed for the classification operation. The web crawling engine operates based on a web crawling query prioritization model that supports determining web crawling priority scores that indicate a predicted performance improvement for classification operations executed with known data and web crawled data to be retrieved from executing a web crawling query operation. Using the web crawling prioritization model, a web crawling priority score is determined for a web crawling query for the corresponding classification operation. The classification operation is associated with a product in a product listing platform and known data for the product.

Abstract: An analytics reporting computing device for generating a region report of a geographic region is described. The analytics reporting computing device is configured to receive a request for a region report, the request identifying a geographic region including a plurality of sub-regions and at least one selected region characteristic, and identify each sub-region. The analytics reporting computing device is also configured to retrieve transaction data and influence data associated with the geographic region, the influence data including electronic data representing user online social behavior. The analytics reporting computing device is further configured to, for each sub-region, determine a respective index score of the selected region characteristic based upon portions of the transaction data and influence data associated with each sub-region.

Abstract: A method of providing changes in healthcare policy information is provided, including: maintaining a first database, the first database containing records including copies of websites, the websites pertaining to policy information; a first server scanning current versions of the websites on the Internet and identifying which of the websites have been changed to a second server; and updating the records in the first database; a second server generating a report of the changes made to each of the identified websites; and the second server providing access to a user to generate reports relating to the websites.

Abstract: Techniques and mechanisms for detecting and deducing of urgent messages in low-power and lossy networks (LLNs) using a correlation analysis of the nodes within a network and machine learning (ML) models. Utilizing these techniques, a field network director (FND) of the network can determine neighboring devices within the network. ML models may be utilized to determine that based upon receipt of a power outage notification (PON) message and/or a power restoration notification (PRN) message from nodes, neighboring nodes of the nodes may also have suffered a power outage and/or a subsequent power restoration, even if the FND did not receive a corresponding PON message and/or a corresponding PRN message from the neighboring nodes of the network. Thus, loss of power and subsequent power restoration may be handled for large numbers of neighboring nodes within the network, even when only a few PON messages and/or subsequent PRN messages are received.

Abstract: A current sharing, in a system by a first user, of first content with at least one other user can be detected. A navigation path data structure indicating at least one navigation path from a second content to the first content within a network environment can be automatically determined. Responsive to determining the navigation path data structure indicating the at least one navigation path from the second content to the first content within the network environment, performance of at least one action can be automatically initiated based on the determined navigation path data structure.

Abstract: A communication method and system according to the present invention generates a unique cryptographically secure URI in response to receiving a user post. The user post and URI are stored temporarily. The URI is sent to an intended recipient. In response to a first instance of accessing the URI, the content is retrieved and sent to the intended recipient. The original uploaded content and URI are then deleted. In response to subsequent instances of accessing the URI, random content determined in part by a current environmental state of the communication, is returned. Recent subsequent instances can return thematically similar content.

Abstract: A current sharing, in a system by a first user, of first content with at least one other user can be detected. A navigation path data structure indicating at least one navigation path from a second content to the first content within a network environment can be automatically determined. Responsive to determining the navigation path data structure indicating the at least one navigation path from the second content to the first content within the network environment, performance of at least one action can be automatically initiated based on the determined navigation path data structure.

Abstract: In some embodiments, a computer-implemented method for retrieving a similar document from a corpus of documents is provided. The method may include receiving a search document comprising a set of words; applying a first encoder to generate a first vector; applying a second encoder to generate a second vector; determining a first similarity between the first vector of the search document and the first vector of each document of the corpus of documents; determining a second similarity between the second vector of the search document and the second vector of each document of the corpus of documents; generating a first ranked list of documents based on the first similarity; generating a second ranked list of documents based on the second similarity; applying a voting algorithm to determine a score associated with each document; and outputting a third ranked list of documents based on the determined score.