Abstract: Techniques for comparing customer records to identify linked customer records are provided. The techniques may include a computer system identifying a pair of customer records and comparing the customer records to determine if the customer records are linked records. The computer system may analyze a first set of data values from first corresponding data fields from the pair of records to determine a first distance between the first set of data values, analyze a second set of data values from second corresponding data fields from the pair of records to determine a second distance between the second set of data values, and combine the first distance and the second distance into a combined distance value which is representative of a distance between the pair of customer records by using a combination function which provides a maximum value cap for the first distance and a weight for the first distance.

Abstract: The present disclosure extends to temporally consistent group keys for groups of linked records in a data store. In embodiments, records in the data store can be categorized as persistent or temporary. Each record carries a unique record identifier and a timestamp depicting the record's time of creation. The group key for a group is the record identifier for the oldest persistent record in the group. If a group has no persistent records, the group key is the record identifier for the oldest temporary record in the group. If two or more records have identical timestamps, the group key is the first record identifier in a lexicographical ordering of the record identifiers for records having the earliest timestamps.

Abstract: Techniques for comparing customer records to identify linked customer records are provided. The techniques may include a computer system identifying a target group of electronic customer records which have data pertaining to a customer, receiving a hierarchical blocker, and analyzing a customer record against the hierarchical blocker and extracting one or more data values from data fields identified by the hierarchical blocker. The computer system may create block keys from the extracted data values, associate the block keys with the customer record, create a block key group of customer records which are associated with a particular block key, and compare pairs of records from the block key group to identify records which are likely to be linked records because they pertain to a single customer entity.

Abstract: Techniques for comparing customer records to identify linked customer records pertaining to a single customer entity are provided. The techniques include identifying a target group of electronic customer records having data fields containing data pertaining to a customer, identifying one or more focused blockers identifying a data value for an electronic customer record data field, and analyzing the target group of electronic customer records to identify a focused group of electronic customer records containing the focused blocker data value. The techniques also include comparing pairs of electronic customer records from the focused group of electronic customer records to identify linked records which pertain to a single customer entity.

Abstract: The present disclosure extends to comparing two or more names in a database of contact records. In embodiments, systems of the present disclosure execute a comparison of submitted names to determine if the records should be linked and/or merged. Embodiments of the present disclosure can compensate for irregularities in the database of contact records, including inconsistent name prefixes and suffixes, misspellings, typographical errors, misordered first/last names, and nicknames.

Abstract: In some embodiments, a method can comprise accessing one or more neighborhood data structures, at least a portion of the one or more neighborhood data structures are active. In many embodiments, the one or more neighborhood data structures each comprise a label comprising a record identifier, the record identifier corresponding to a specified record of the one or more records, a neighborhood comprising a list of record identifiers for other records linked to the specified record of the one or more records by a pairwise link, and a message for each of the other records linked to the specified record of the one or more records, the message comprising the label and an activity state indicator, the activity state indicator indicating if the neighborhood data structure is active. Other embodiments of related methods and systems are also provided.

Abstract: A computer-implemented method is disclosed for efficiently processing a large number of records. In the method, a computer system may obtain a plurality of records and count the number of records thereof corresponding to each block of a plurality of blocks. The computer system may also identify a plurality of partitions corresponding to selected blocks of the plurality of blocks. Each partition of the plurality of partitions may be substantially uniform in processing time. The computer system may then distribute a workload associated with a block or partition to each node of a plurality of nodes contained within the computer system. Each node may then process the block or partition in parallel such that each node completes the processing within a selected period of time.

Abstract: A computer-implemented method is disclosed for efficiently processing records with unseen data. In the method, a computer system may obtain a plurality of records and a decision tree generated in a learning process. The decision tree may include a distinction node having multiple paths extending therefrom. After arriving at the distinction node with one or more records, the computer system may determine that the one or more records correspond to data of a type not seen by the distinction node in the learning process. Thereafter, the computer system may depart the distinction node via each of the multiple paths and eventually reach multiple leaf nodes of the decision tree. Each of the multiple leaf nodes may correspond to a probability distribution. Accordingly, the computer system may combine the probability distribution of each of the multiple leaf nodes to obtain a hybrid probability distribution corresponding to the one or more records.

Abstract: In some embodiments, a method can comprise accessing one or more neighborhood data structures, at least a portion of the one or more neighborhood data structures are active. In many embodiments, the one or more neighborhood data structures each comprise a label comprising a record identifier, the record identifier corresponding to a specified record of the one or more records, a neighborhood comprising a list of record identifiers for other records linked to the specified record of the one or more records by a pairwise link, and a message for each of the other records linked to the specified record of the one or more records, the message comprising the label and an activity state indicator, the activity state indicator indicating if the neighborhood data structure is active. Other embodiments of related methods and systems are also provided.

Abstract: A computer-implemented method is disclosed for efficiently processing a large number of records. In the method, a computer system may obtain a plurality of records and a decision tree. The decision tree may include a distinction node corresponding to a distinction requiring completion of a computation. Due to the fact that the computation may be, in the overall context of the process, computationally expensive, it may initially be left uncomputed. Accordingly, if the distinction node is never reached when records are being processed, no computation time gets wasted. However, if and when the distinction node is reached, the computer system may complete the computation and make the distinction based on results of the computation.

Abstract: The present invention extends to methods, systems, and computer program products for resolving pairwise links to groups. Embodiments of the invention use an iterative algorithm to transform a collection of pairwise links to groups of records that correspond to the same entity. The algorithm can be stopped after any number of iterations for an increasing accurate approximation result. The algorithm essentially guarantees a correct solution for groups of size up to the number of iterations. This algorithm scales linearly on the size of the record set, with little impact from the number of links.

Abstract: Systems and methods are disclosed for classifying records by sorting records based on both quantitative and qualitative information at a node in a decision tree. Technologies are also disclosed for determining whether records are linked together by applying quantitative and qualitative information at the same nodes in a decision tree. Furthermore, improvements to decision trees are disclosed in terms of the generation and/or training of decision trees that harnesses additional information in the quantitative and qualitative aspects that a unit of data relevant to a single node, and/or the relationships between these aspects, may provide a machine learning algorithm.

Abstract: Techniques for avoiding duplicate comparisons while comparing customer records to identify linked customer records pertaining to a single customer entity are provided. The techniques include the computer system comparing a first electronic customer record with a second electronic customer record to determine if the first electronic customer record and the second electronic customer record pertain to a single customer entity if the computer system identifies a common blocker key corresponding to a selected blocker from a data field in the first electronic customer record and from a data field in the second electronic customer record and if the computer system does not identify a common blocker key corresponding to an additional lower order blocker from another data field in the first electronic customer record and from a data field in the second electronic customer record.

Abstract: The present invention extends to methods, systems, and computer program products for resolving pairwise links to groups. Embodiments of the invention use an iterative algorithm to transform a collection of pairwise links to groups of records that correspond to the same entity. The algorithm can be stopped after any number of iterations for an increasing accurate approximation result. The algorithm essentially guarantees a correct solution for groups of size up to the number of iterations. This algorithm scales linearly on the size of the record set, with little impact from the number of links.

Abstract: The present disclosure extends to methods, systems, and computer program products for determining customer linkages between a plurality of customer profiles and providing missing cost values in the attribute fields.

Abstract: A computer-implemented method is disclosed for efficiently processing a large number of records. In the method, a computer system may obtain a plurality of records and a decision tree. The decision tree may include a distinction node corresponding to a distinction requiring completion of a computation. Due to the fact that the computation may be, in the overall context of the process, computationally expensive, it may initially be left uncomputed. Accordingly, if the distinction node is never reached when records are being processed, no computation time gets wasted. However, if and when the distinction node is reached, the computer system may complete the computation and make the distinction based on results of the computation.

Abstract: A computer-implemented method is disclosed for efficiently processing set-based attributes. In the method, a computer system may obtain a plurality of records and a decision tree. The decision tree may include a distinction node corresponding to a comparison of two attributes. The distinction node may have a match path and a no match path extending therefrom. After arriving at the distinction node, the computer system may initiate a process wherein each member of a first set corresponding to a first of the two attributes is to be compared to each member of a second set corresponding to a second of the two attributes. The computer system may depart the distinction node via the match path after the process reveals that at least one member of the first set matches at least one member of the second set.

Abstract: Techniques for avoiding duplicate comparisons while comparing customer records to identify linked customer records pertaining to a single customer entity are provided. The techniques include the computer system comparing a first electronic customer record with a second electronic customer record to determine if the first electronic customer record and the second electronic customer record pertain to a single customer entity if the computer system identifies a common blocker key corresponding to a selected blocker from a data field in the first electronic customer record and from a data field in the second electronic customer record and if the computer system does not identify a common blocker key corresponding to an additional lower order blocker from another data field in the first electronic customer record and from a data field in the second electronic customer record.

Abstract: Systems and methods are disclosed for classifying records by sorting records based on both quantitative and qualitative information at a node in a decision tree. Technologies are also disclosed for determining whether records are linked together by applying quantitative and qualitative information at the same nodes in a decision tree. Furthermore, improvements to decision trees are disclosed in terms of the generation and/or training of decision trees that harnesses additional information in the quantitative and qualitative aspects that a unit of data relevant to a single node, and/or the relationships between these aspects, may provide a machine learning algorithm.

Abstract: A computer-implemented method is disclosed for efficiently processing records with unseen data. In the method, a computer system may obtain a plurality of records and a decision tree generated in a learning process. The decision tree may include a distinction node having multiple paths extending therefrom. After arriving at the distinction node with one or more records, the computer system may determine that the one or more records correspond to data of a type not seen by the distinction node in the learning process. Thereafter, the computer system may depart the distinction node via each of the multiple paths and eventually reach multiple leaf nodes of the decision tree. Each of the multiple leaf nodes may correspond to a probability distribution. Accordingly, the computer system may combine the probability distribution of each of the multiple leaf nodes to obtain a hybrid probability distribution corresponding to the one or more records.