A method executes at a computer system to retrieve data from a database. Upon receiving a database query, the computer system translates the query into an intermediate representation, and estimates a compilation time to compile the intermediate representation into machine executable code. The query execution time to retrieve a result set is also estimated. In accordance with a determination that the query execution time and compilation time satisfy an interpretation criterion, the computer system invokes a byte code interpreter to interpret the intermediate representation and retrieve the result set from the database. In accordance with a determination that the query execution and compilation times satisfy one of a plurality of compilation criteria, the computer system compiles the intermediate representation to form machine code and executes the machine code to retrieve the result set from the database. In some cases, the query intermediate representation is optimized prior to compilation.

Methods and system for item matching are described. In one embodiment, compatibility-based text for an item may be accessed. A compatibility identifier may be identified based on the compatibility-based text. The compatibility identifier may be associated with an item cluster. The compatibility identifier may be used to identify a plurality of matching items. A result may be provided based on identification of the plurality of matching items. Additional methods and systems are disclosed.

In some implementations, systems and methods that are capable of customizing negative media searches using domain-specific search indexes are described. Data indicating a search query associated with a negative media search for an entity and a corpus of documents to be searched are obtained. Content from a particular collection of documents from among the corpus of documents is obtained and processed. Multiple scores for the entity are computed based on processing the content obtained from the collection of documents. The multiple scores are aggregated to compute a priority indicator that represents a likelihood that the collection of documents includes content that is descriptive of derogatory information.

In implementations of the subject matter described herein, a new approach for presenting a response to a message in a conversation is proposed. Generally speaking, in response to receiving a message in a conversation, the received message will be matched with one or more documents on the sentence basis. That is, the received message is compared with the sentences from a document(s), rather than predefined query-response pairs. In this way, a whole sentence may be selected from the document as a candidate response. Then the suitability of this sentence with respect to the ongoing conversation will be determined, and the response will be generated and rendered in an adaptive way based on the suitability. As a result, the user experiences may be significantly enhanced in the chatbot scenario.

Systems and methods for generating concept-based search summaries from a plurality of documents are provided. In one embodiment, a system may include interfaces to receive information identifying a meaning taxonomy including a normalized term and a search query including search terms. The system may be configured to identify documents relating to the search terms and normalized terms and display a concept-based summary of the documents, the summary including a syntactic structure associated with the normalized terms and search terms. In another embodiment, a method includes receiving a meaning taxonomy including normalized terms and search terms, identifying at least one document including the search terms and syntactic structures associated with the normalized terms, and display a search summary including the search terms and syntactic structures.

A document retrieval system that retrieves documents, with concepts of the documents taken into account, is provided. The document retrieval system (100) includes an input unit (101), a first processing unit (102), a storage unit (105), a second processing unit (103), and an output unit (104). The input unit (101) has a function of inputting a first document (20), the first processing unit (102) has a function of creating a first graph structure (21) from the first document (20), the storage unit (105) has a function of storing a second graph structure (11), the second processing unit (103) has a function of calculating a similarity between the first graph structure (21) and the second graph structure (11), the output unit (104) has a function of supplying information, the first processing unit (102) has a function of dividing the first document (20) into a plurality of tokens, a node and an edge of the first graph structure (21) have a label, and the label includes the plurality of tokens.

A method includes extracting variable length strings from text data, associating the extracted strings with indexes in an indexing structure that maintain identifiers for instances of the strings in the text data, selecting a set of the strings having corresponding indexes whose size exceeds a threshold size value, and determining whether to merge a first index corresponding to a first string with a second index corresponding to a second string, the second string being a sub string of the first string, wherein the determination is based at least in part on a comparison of a first size of the first index and a second size of the second index. The method further includes merging the first index with the second index to create a merged index in the indexing structure responsive to the determination, and processing queries to the text data utilizing the indexing structure with the merged index.

Generating textual entailment pair by a natural language processing (NLP) system. The NLP system receives two input texts, such as a question and a candidate answer. The NLP system queries a database and retrieves passages likely to include text that support the candidate answer. The NLP system generates parse trees and performs term matching on the passages and scores them according to the matching. The NLP system detects anchor pairs in the question and in the passage and aligns subgraphs (within the parse trees) of one to the other based on matching. The NLP system identifies aligned terms in the question and the passage that are not in the aligned subgraphs. The NLP system identifies text fragments, for the question and the passage, within the non-aligned segments of their respective parse trees, that connect the aligned term to the aligned portion of the subgraph.

Metadata of a storage object is managed. An example method for managing metadata of a storage object comprises: in response to receiving a request to access metadata of a storage object, obtaining a first identifier of the storage object from the request; determining, based on the first identifier, a second identifier of a page where the metadata is located, an address of the page and the second identifier being stored in association in an entry of a page table distributed among a plurality of storage devices with different access speeds; determining, from the plurality of storage devices, a storage device storing the entry; and obtaining the address of the page by accessing the entry from the storage device, so as to access the metadata from the page.

A master database module is on a master computer node. Slave database modules are on slave computer nodes connected to the master computer node via a network. A distributed database includes executable code executed by processors on the master computer node and the slave computer nodes to receive a distributed database query at the master computer node. A query execution plan is prepared at the master computer node. The query execution plan is deployed on the slave computer nodes. The query execution plan is executed on the slave computer nodes. The slave computer nodes each perform a single node query execution analysis to selectively produce a query cancellation command. The query cancellation command is propagated to the master computer node and the slave computer nodes. The query execution plan is cancelled on the master computer node and the slave computer nodes.