Abstract: A system answers natural language questions using retrieval-augmented generation. The system stores a set of domain specific documents in a vector database. The system receives a natural language question. The system retrieves a subset of documents relevant to the natural language question from the vector database. The system determines prior knowledge information required in addition to the subset of documents retrieved from the vector database for answering the natural language question. The system generates a prompt for a machine learning based language model including instructions to the machine learning based language model to refrain from using prior knowledge obtained by the machine learning based language model during training of the machine learning based language model. The receives a response generated by executing the machine learning based language model based on the prompt. The system performs an action based on the response.

Abstract: A method and apparatus capable of identifying a unique user across multiple devices in a computer network are provided. Device-specific and behavioral features associated with an event and a device are extracted. The device-specific features form a device signature associated the device. Hardware mobile device identifiers (IDs) are also associated with mobile application devices. Over a period of time, the behavioral features of such devices are monitored. Similarity scores between various devices are calculated based on the behavioral features and device types. The devices in the computer network are clustered and a device graph is generated representing the connections between the devices based on the similarity scores. A unique user ID associated with the multiple devices is generated from the device graph.

Abstract: A method and system target an audience segment in real-time based on social activities of users. A set of keywords associated with a brand and a set of keywords associated with a first plurality of users is compared. When there is match between the two sets of keywords, a seed audience segment that includes a subset of the first plurality of users is generated. User profiles of users from the seed audience segment based on features are generated. Subsequently, model files of features, feature threshold scores, and an overall feature threshold score corresponding to the user profiles are calculated and provided to a real-time bidding (RTB) server. When the RTB server receives a bid request for a cookie, the RTB server computes a model score corresponding to the cookie based on the model files and accepts the bid request if the model score is above the overall feature threshold score.

Abstract: Embodiments of a method for determining a mapping are illustrated. In an embodiment, the method includes receiving a log record from a tracking component that is located on a plurality of web pages. The method further includes determining a first mapping between a plurality of anchors associated with the plurality of users. The method also includes determining a second mapping between the plurality of users based on the first mapping.

Abstract: Embodiments of a method for generating reports are illustrated. In an embodiment, the method includes receiving a log record from a tracking component that is located on a plurality of web pages. The method includes extracting a plurality of user features for a plurality of users based on the at least one log record. The method further includes determining a first mapping between the plurality of users and a plurality of user features, and a second mapping between the plurality of users and a plurality of advertisement campaign descriptors. The method also includes merging the first mapping and the second mapping to create a merged data model, and analyzing the merged data model to generate reports.

Abstract: An advertisement campaign optimization system, method, and a computer product for determining real time bidding data corresponding to one or more advertisement campaigns targeting one or more users are disclosed. In an embodiment, the method comprises receiving a bid request for the one or more advertisement campaigns. User interest data corresponding to the one or more users is determined. One or more bidding rules based on a first data set and a second data set are also determined. The first data set comprises data corresponding to the one or more users and the second data set comprises data corresponding to the one or more advertisement campaigns. The method determines real time bidding data based on the user interest data and the one or more bidding rules.

Abstract: Embodiments for a method for ranking social quality of content published on a plurality of web pages are provided. In an embodiment, the method includes receiving at least one log record from a tracking component on at least one web page. The one log record is indicative of at least one user activity on the at least one web page. The method further includes aggregating the at least one log record corresponding to preferably each of the plurality of web pages based on one or more parameters. The method also includes assigning a first score for preferably each of the plurality of web pages based on the aggregating. The first score is indicative of a social quality of content published in the at least one web page. The method includes ranking the plurality of web pages based on the first score.

Abstract: Embodiments of a method for determining a mapping are illustrated. In an embodiment, the method includes receiving a log record from a tracking component that is located on a plurality of web pages. The method further includes determining a first mapping between a plurality of anchors associated with the plurality of users. The method also includes determining a second mapping between the plurality of users based on the first mapping.

Abstract: The present invention provides a method and a system for providing targeted advertisements to a user. The method includes determining a set of signals corresponding to at least one online activity associated with the user and then analyzing the set of signals to identify one or more user interests. The set of signals includes at least one of a share signal, a view signal, a search signal, and a click signal. Further, the method includes tagging the user with at least one ad exchange cookie based on the identified user interests and an available advertisement pool. Lastly, the method includes displaying, by an advertisement network, an advertisement to the user based on the ad exchange cookie.

Abstract: The invention provides a business method and a system to perform focused online shopping and sharing the online shopping experience with other users. The shopping-related information from various sources, and the search query, is converted into directed acyclic graph forests. These graphs are then compared to identify the search results that correspond to the shopping criteria. The sharing of online shopping experience includes sharing of search results between multiple users, discussing the search results through instant messaging (IM), revision of relevant items by any or all users and the flexibility of online-purchase by any user.

Abstract: A system and method are provided for delivering financial services in a substantially automated and efficient manner. In one embodiment, an integrated Internet-based common digital platform automates and simplifies a mortgage closing process by receiving transaction information and merging the same with a set of electronic documents. The common digital platform then selectively permits users to access and edit certain ones of the documents, permits electronic signature of the documents, records transfers of negotiable instruments associated with the transaction, and stores the electronic documents in an electronic vault. The common digital platform also includes a dynamic pricing engine containing updated cost estimates for various closing cost elements in multiple geographical regions and generates closing cost estimates based on the geographical location of the real estate.

Abstract: A general purpose expert system architecture for diagnosing faults in any one of a plurality of machines includes a machine information database containing information on characteristics of various components of the machines to be diagnosed and a sensory input database which contains vibration data taken at predetermined locations on each of the machines. The system knowledge base contains a plurality of general rules that are applicable to each of the plurality of machines. The generality of diagnosis is accomplished by focusing on components that make up the machine rather than individual machines as a whole. The system architecture also permits diagnosis of machines based on other parameters such as amperage, torques, displacement and its derivatives, forces, pressures and temperatures. The system includes an inference engine which links the rules in a backward chaining structure.