Abstract: A method and system are described for improving the speed and efficiency of obtaining conversational search results. A user may speak a phrase to perform a conversational search or a series of phrases to perform a series of searches. These spoken phrases may be enriched by context and then converted into a query embedding. A similarity between the query embedding and document embeddings is used to determine the search results including a query cutoff number of documents and a cache cutoff number of documents. A second search phrase may use the cache of documents along with comparisons of the returned documents and the first query embedding to determine the quality of the cache for responding to the second search query. If the results are high-quality then the search may proceed much more rapidly by applying the second query only to the cached documents rather than to the server.

Abstract: A method and system are described for improving the speed and efficiency of obtaining conversational search results. A user may speak a phrase to perform a conversational search or a series of phrases to perform a series of searches. These spoken phrases may be enriched by context and then converted into a query embedding. A similarity between the query embedding and document embeddings is used to determine the search results including a query cutoff number of documents and a cache cutoff number of documents. A second search phrase may use the cache of documents along with comparisons of the returned documents and the first query embedding to determine the quality of the cache for responding to the second search query. If the results are high-quality then the search may proceed much more rapidly by applying the second query only to the cached documents rather than to the server.

Abstract: A method and system are described for improving the speed and efficiency of obtaining conversational search results. A user may speak a phrase to perform a conversational search or a series of phrases to perform a series of searches. These spoken phrases may be enriched by context and then converted into a query embedding. A similarity between the query embedding and document embeddings is used to determine the search results including a query cutoff number of documents and a cache cutoff number of documents. A second search phrase may use the cache of documents along with comparisons of the returned documents and the first query embedding to determine the quality of the cache for responding to the second search query. If the results are high-quality then the search may proceed much more rapidly by applying the second query only to the cached documents rather than to the server.

Abstract: Embodiments may provide a cache for query results that can adapt the cache-space utilization to the popularity of the various topics represented in the query stream. For example, a method for query processing may perform receiving a plurality of queries for data and requesting data responsive to at least one query from a data cache comprising a temporal cache, wherein the temporal cache is configured to store data based on a topic associated with the data and is configured to retrieve data based on a topic, and wherein the data cache is configured to retrieve data responsive to at least one query from the computer system.

Abstract: Embodiments may provide a cache for query results that can adapt the cache-space utilization to the popularity of the various topics represented in the query stream. For example, a method for query processing may perform receiving a plurality of queries for data and requesting data responsive to at least one query from a data cache comprising a temporal cache, wherein the temporal cache is configured to store data based on a topic associated with the data and is configured to retrieve data based on a topic, and wherein the data cache is configured to retrieve data responsive to at least one query from the computer system.

Abstract: Embodiments may provide a cache for query results that can adapt the cache-space utilization to the popularity of the various topics represented in the query stream. For example, a method for query processing may perform receiving a plurality of queries for data, determining at least one topic associated with each query, and requesting data responsive to each query from a data cache comprising a plurality of partitions, including at least a static cache partition, a dynamic cache partition, and a temporal cache partition, the temporal cache partition may store data based on a topic associated with the data, and may be further partitioned into a plurality of topic portions, each portion may store data relating to an associated topic, wherein the associated topic may be selected from among determined topics of queries received by the computer system, and the data cache may retrieve data for the queries from the computer system.

Abstract: The present invention concerns a novel method to efficiently score documents (texts, images, audios, videos, and any other information file) by using a machine-learned ranking function modeled by an additive ensemble of regression trees. A main contribution is a new representation of the tree ensemble based on bitvectors, where the tree traversal, aimed to detect the leaves that contribute to the final scoring of a document, is performed through efficient logical bitwise operations. In addition, the traversal is not performed one tree after another, as one would expect, but it is interleaved, feature by feature, over the whole tree ensemble. Tests conducted on publicly available LtR datasets confirm unprecedented speedups (up to 6.5×) over the best state-of-the-art methods.

Abstract: Embodiments may provide a cache for query results that can adapt the cache-space utilization to the popularity of the various topics represented in the query stream. For example, a method for query processing may perform receiving a plurality of queries for data, determining at least one topic associated with each query, and requesting data responsive to each query from a data cache comprising a plurality of partitions, including at least a static cache partition, a dynamic cache partition, and a temporal cache partition, the temporal cache partition may store data based on a topic associated with the data, and may be further partitioned into a plurality of topic portions, each portion may store data relating to an associated topic, wherein the associated topic may be selected from among determined topics of queries received by the computer system, and the data cache may retrieve data for the queries from the computer system.

Abstract: Embodiments may provide a cache for query results that can adapt the cache-space utilization to the popularity of the various topics represented in the query stream. For example, a method for query processing may perform receiving a plurality of queries for data, determining at least one topic associated with each query, and requesting data responsive to each query from a data cache comprising a plurality of partitions, including at least a static cache partition, a dynamic cache partition, and a temporal cache partition, the temporal cache partition may store data based on a topic associated with the data, and may be further partitioned into a plurality of topic portions, each portion may store data relating to an associated topic, wherein the associated topic may be selected from among determined topics of queries received by the computer system, and the data cache may retrieve data for the queries from the computer system.

Abstract: The present invention concerns a novel method to efficiently score documents (texts, images, audios, videos, and any other information file) by using a machine-learned ranking function modeled by an additive ensemble of regression trees. A main contribution is a new representation of the tree ensemble based on bitvectors, where the tree traversal, aimed to detect the leaves that contribute to the final scoring of a document, is performed through efficient logical bitwise operations. In addition, the traversal is not performed one tree after another, as one would expect, but it is interleaved, feature by feature, over the whole tree ensemble. Tests conducted on publicly available LtR datasets confirm unprecedented speedups (up to 6.5×) over the best state-of-the-art methods.