Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialog patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialog patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: Systems and methods are disclosed for partitioning data into n partitions in a database used for information retrieval. According to certain embodiments implemented with at least one processor, a perceived load is determined for each of a plurality of substrings, each of the substrings corresponding to one or more entries in the database. The plurality of substrings are ordered from highest to lowest perceived load. Data associated with each of the first n substrings is assigned to a separate partition, wherein n is greater than one. Moreover, data associated with each of the remaining substrings is assigned to the n partitions based on a score. The score is computed by calculating, for each partition, a load variance and an entropy weight for the candidate substring. The load variance and the entropy weight are combined, and the candidate substring is assigned to the partition that has the lowest combined score.

Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialogue patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialogue patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: Systems and methods are disclosed for identifying points-of-interest using customized query prediction. According to certain embodiments, a partial query is received from a client via a network. The partial query may comprise a plurality of characters associated with a full query and location data associated with the client. The received partial query may be analyzed using at least one dictionary. Based on the analysis of the received partial query, at least one query may be predicted. Moreover, at least one point-of-interest may be identified based on the at least one predicted query. Further, data associated with the at least one identified point-of-interest may be transmitted over the network to the client prior to receiving the full query from the client.

Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialog patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialog patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: Systems and methods are disclosed for identifying points-of-interest using customized query prediction. According to certain embodiments, a partial query is received from a client via a network. The partial query may comprise a plurality of characters associated with a full query and location data associated with the client. The received partial query may be analyzed using at least one dictionary. Based on the analysis of the received partial query, at least one query may be predicted. Moreover, at least one point-of-interest may be identified based on the at least one predicted query. Further, data associated with the at least one identified point-of-interest may be transmitted over the network to the client prior to receiving the full query from the client.

Abstract: Systems and methods are disclosed for partitioning data into n partitions in a database used for information retrieval. According to certain embodiments implemented with at least one processor, a perceived load is determined for each of a plurality of substrings, each of the substrings corresponding to one or more entries in the database. The plurality of substrings are ordered from highest to lowest perceived load. Data associated with each of the first n substrings is assigned to a separate partition, wherein n is greater than one. Moreover, data associated with each of the remaining substrings is assigned to the n partitions based on a score. The score is computed by calculating, for each partition, a load variance and an entropy weight for the candidate substring. The load variance and the entropy weight are combined, and the candidate substring is assigned to the partition that has the lowest combined score.

Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialogue patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialogue patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialogue patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialogue patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialogue patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialogue patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Abstract: A facility for the reconciliation of data records pertaining to business entities. One or more fitness functions are applied to fields contained in two conflicting data records to assess the similarity of each field. The results of the fitness functions are then weighted and combined to assess the likelihood that the two data records are associated with the same business entity. When the weighted fitness functions are applied to conflicting data records, the fitness functions generate a confidence level that the compared records are associated with the same business entity. If the confidence level exceeds a certain threshold, the facility accepts that the data records refer to the same business entity and synthesizes a business record from the data records.

Abstract: A system and method to automatically generate display text for advertisements. The generated creatives may be distributed in text, banner, email, or other advertising formats through online advertising distribution networks. To generate creatives, the system utilizes a record of the grammatical structure (or “grammar”) of creatives and combinations of two or more words (“N-grams”) to populate each grammar. To generate grammars of creatives, the system analyzes a large set of creatives that have previously been used in advertisements to populate grammar and N-gram datasets. Once the grammar and N-gram datasets are populated with a sufficient number of grammars and N-grams, the system is able to generate new creatives for advertisements. The system populates the selected grammars using N-grams that are obtained from the N-gram dataset in order to form new creatives.

Abstract: A facility for the reconciliation of data records pertaining to business entities. One or more fitness functions are applied to fields contained in two conflicting data records to assess the similarity of each field. The results of the fitness functions are then weighted and combined to assess the likelihood that the two data records are associated with the same business entity. When the weighted fitness functions are applied to conflicting data records, the fitness functions generate a confidence level that the compared records are associated with the same business entity. If the confidence level exceeds a certain threshold, the facility accepts that the data records refer to the same business entity and synthesizes a business record from the data records.