Abstract: A system and a method are disclosed for recommending a change to improve performance of a target workflow. A workflow management system receives the target workflow intended to be used in a particular context to achieve a target result. The target workflow has a structure with a plurality of steps performed in a predefined order, but there may be options for modifying the workflow to lead to better performance (e.g., change type of action performed in a step, change order of steps, add a new step). The workflow management system identifies candidate workflows that are similar to the target workflow and identifies historical changes that have been made to these candidate workflows. Using a machine learning model, the workflow management system determines a change from one of the historical changes made to the candidate workflows associated with the highest expected impact when applied to the target workflow.

Abstract: A system and a method are disclosed for recommending a change to improve performance of a target workflow. A workflow management system receives the target workflow intended to be used in a particular context to achieve a target result. The target workflow has a structure with a plurality of steps performed in a predefined order, but there may be options for modifying the workflow to lead to better performance (e.g., change type of action performed in a step, change order of steps, add a new step). The workflow management system identifies candidate workflows that are similar to the target workflow and identifies historical changes that have been made to these candidate workflows. Using a machine learning model, the workflow management system determines a change from one of the historical changes made to the candidate workflows associated with the highest expected impact when applied to the target workflow.

Abstract: An entity summarization system is described herein that mines the Internet and other data source to provide answers to questions such as the relative sentiment of users towards various brands. The system uses a controlled vocabulary list describing a specific aspect of entities of interest. Given an entity name, the system scans the whole content corpus to collect statistics on the words that occur most frequently in the context of the entity name, taking into account proximity information, to produce a weighted list of vocabulary terms describing the entity. Two entities can be compared by normalizing and comparing their weighted term lists. In some embodiments, the system performs these procedures efficiently by leveraging an N-gram web model. Thus, the system provides an automated way to compare two entities to derive information about how users feel about the entities at any given time.

Abstract: Embodiment described herein are generally directed to a toolbar extension of a web browser that grabs a user's search engine query and suggests a refined search query known to yield better search results. The toolbar recognizes the web page the user is on as being associated with a search engine and retrieves the user's search query. The toolbar interacts with a refinement component on a server, and the refinement component determines a refined search query based on confidence scores assigned to data mined from a data center affiliated with different search engine (one related to the toolbar). The refined search query is returned and displayed in a search field of the toolbar, allowing the user to easily run the refined search on the different search engine.

Abstract: One or more search seeds for Web crawling operations are selected. In a directed graph with Web pages represented by vertices and links represented by edges, characteristics of vertices connected to potential seed vertices are considered in making a seed selection.

Abstract: An entity summarization system is described herein that mines the Internet and other data source to provide answers to questions such as the relative sentiment of users towards various brands. The system uses a controlled vocabulary list describing a specific aspect of entities of interest. Given an entity name, the system scans the whole content corpus to collect statistics on the words that occur most frequently in the context of the entity name, taking into account proximity information, to produce a weighted list of vocabulary terms describing the entity. Two entities can be compared by normalizing and comparing their weighted term lists. In some embodiments, the system performs these procedures efficiently by leveraging an N-gram web model. Thus, the system provides an automated way to compare two entities to derive information about how users feel about the entities at any given time.

Abstract: Embodiment described herein are generally directed to a toolbar extension of a web browser that grabs a user's search engine query and suggests a refined search query known to yield better search results. The toolbar recognizes the web page the user is on as being associated with a search engine and retrieves the user's search query. The toolbar interacts with a refinement component on a server, and the refinement component determines a refined search query based on confidence scores assigned to data mined from a data center affiliated with different search engine (one related to the toolbar). The refined search query is returned and displayed in a search field of the toolbar, allowing the user to easily run the refined search on the different search engine.

Abstract: A toolbar extension of a web browser suggests a user try searching a different search engine than the one on which the user has just conducted a search. The toolbar detects the user is frustrated with the search results returned by the search engine by monitoring a number of session heuristics. Analyzing historical user session, log, and click data of other users who have submitted the search query to different search engines, a different search engine is selected that has historically generated better search results. A suggestion is then presented to the user to run the search query in the more effective search engine. Such a suggestion may be presented in a number of ways, but one embodiment displays a window next to the text field of the toolbar and provides a link to conduct a search for the search query in the more effective search engine.

Abstract: A method for selecting one or more search seeds for Web crawling operations is provided. In a directed graph with Web pages represented by vertices and links represented by edges, characteristics of vertices connected to potential seed vertices are considered in making a seed selection.

Abstract: A host-based seed selection process considers factors such as quality, importance and potential yield of hosts in a decision to use a document of a host as a seed. A subset of a plurality of hosts is determined, including some but not all of the plurality of the hosts, according to an indication of importance of the hosts, according to an expected yield of new documents for the hosts, and according to preferences for the markets the hosts belong to. At least one seed is generated for each host of the determined subset of hosts, wherein each generated at least one seed includes an indication of a document in the linked database of documents. The generated seeds are provided to be accessible by a database crawler.