Computer-based systems configured to pre-train language models for entity resolution and methods of use thereof
In some embodiments, the present disclosure provides an exemplary method that may include steps of receiving a dataset of entity records, identifying, a candidate entity record of the plurality of entity records, utilizing a set of predefined rules to generate a first augmented record, and a second augmented record, utilizing, at least one contrastive loss optimization functions to train parameters of an unsupervised self-contrastive machine learning language model to distinguish between similar entity records representing a same entity and dissimilar entity records.
1 . A computer-implemented method comprising:
receiving, by at least one processor, a dataset of entity records, the dataset comprising a plurality of entity records associated with one or more entities, wherein each entity record comprises at least one element;
identifying, by at least one processor, a candidate entity record of the plurality of entity records;
utilizing, by the at least one processor, first predefined rules of a set of predefined rules to generate a first augmented record by augmenting the at least one element of the candidate entity record with a first augmentation type;
wherein the first augmentation type represents a positive contrast between the first augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilizing, by the at least one processor, second predefined rules of a set of predefined rules to generate a second augmented record by augmenting the at least one element of the candidate entity record with a second augmentation type;
wherein the second augmentation type represents a negative contrast between the second augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilizing, by the at least one processor, at least one contrastive loss optimization to train parameters of an unsupervised self-contrastive machine learning language model to distinguish between similar entity records representing a same entity and dissimilar entity records represented different entities based at least in part on the at least one element of each entity record;
wherein the at least one contrastive loss optimization trains the parameters based at least in part on:
the positive contrast between the first augmented record and the candidate entity record, and
the negative contrast between the second augmented record and the candidate entity record; and
utilizing, by the at least one processor, an index engine to index the entity records determined to have a positive contrast with the candidate entity record;
wherein the index engine indexes the identified entity records into at least one database table so as to merge the entity records having the positive contrast with the candidate entity record.
2 . The computer-implemented method of claim 1 wherein, the set of predefined rules to generate an augmented record call for a random insertion of the at least one element of the candidate entity record.
3 . The computer-implemented method of claim 1 wherein, the set of predefined rules to generate an augmented record call for a random deletion of the at least one element of the candidate entity record.
4 . The computer-implemented method of claim 1 wherein, the set of predefined rules to generate an augmented record call for a random swap of the at least one element of the candidate entity record.
5 . The computer-implemented method of claim 1 wherein, the set of predefined rules to generate an augmented record call for a span deletion of the at least one element of the candidate entity record.
6 . The computer-implemented method of claim 1 wherein, the set of predefined rules to generate an augmented record call for randomization of the at least one element of the candidate entity record.
7 . The computer-implemented method of claim 1 wherein, the index engine utilizes a hamming distance to determine a positive contrast with the candidate entity record.
8 . At least one non-transitory computer-readable storage medium having encoded thereon software instructions that, when executed by at least one processor, cause the at least one processor to perform steps to:
receive, by at least one processor, a dataset of entity records, the dataset comprising a plurality of entity records associated with one or more entities, wherein each entity record comprises at least one element;
identify, by at least one processor, a candidate entity record of the plurality of entity records;
utilize, by the at least one processor, first predefined rules of a set of predefined rules to generate a first augmented record by augmenting the at least one element of the candidate entity record with a first augmentation type;
wherein the first augmentation type represents a positive contrast between the first augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilize, by the at least one processor, second predefined rules of a set of predefined rules to generate a second augmented record by augmenting the at least one element of the candidate entity record with a second augmentation type;
wherein the second augmentation type represents a negative contrast between the second augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilize, by the at least one processor, at least one contrastive loss optimization to train parameters of an unsupervised self-contrastive machine learning language model to distinguish between similar entity records representing a same entity and dissimilar entity records represented different entities based at least in part on the at least one element of each entity record;
wherein the at least one contrastive loss optimization trains the parameters based at least in part on:
the positive contrast between the first augmented record and the candidate entity record, and
the negative contrast between the second augmented record and the candidate entity record;
utilize, by the at least one processor, an index engine to identify the entity records determined to have a positive contrast with the candidate entity record;
wherein the index engine indexes the identified entity records into at least one database table so as to merge the entity records having the positive contrast with the candidate entity record.
9 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the set of predefined rules to generate an augmented record call for a random insertion of the at least one element of the candidate entity record.
10 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the set of predefined rules to generate an augmented record call for a random deletion of the at least one element of the candidate entity record.
11 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the set of predefined rules to generate an augmented record call for a random swap of the at least one element of the candidate entity record.
12 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the set of predefined rules to generate an augmented record call for a span deletion of the at least one element of the candidate entity record.
13 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the set of predefined rules to generate an augmented record call for randomization of the at least one element of the candidate entity record.
14 . The at least one non-transitory computer-readable storage medium of claim 8 , wherein the index engine utilizes a hamming distance to determine a positive contrast with the candidate entity record.
15 . A system comprising:
a non-transient computer memory, storing software instructions; and
at least one processor of a first computing device associated with a user;
wherein, when the at least one processor executes the software instructions, the first computing device is programmed to:
receive, by at least one processor, a dataset of entity records, the dataset comprising a plurality of entity records associated with one or more entities, wherein each entity record comprises at least one element;
identify, by at least one processor, a candidate entity record of the plurality of entity records;
utilize, by the at least one processor, first predefined rules of a set of predefined rules to generate a first augmented record by augmenting the at least one element of the candidate entity record with a first augmentation type;
wherein the first augmentation represents a positive contrast between the first augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilize, by the at least one processor, second predefined rules of a set of predefined rules to generate a second augmented record by augmenting the at least one element of the candidate entity record with a second augmentation type;
wherein the second augmentation type represents a negative contrast between the second augmented record and the candidate entity record based at least in part on the set of predefined rules;
utilize, by the at least one processor, at least one contrastive loss optimization to train parameters of an unsupervised self-contrastive machine learning language model to distinguish between similar entity records representing a same entity and dissimilar entity records represented different entities based at least in part on the at least one element of each entity record;
wherein the at least one contrastive loss optimization trains the parameters based at least in part on:
the positive contrast between the first augmented record and the candidate entity record, and
the negative contrast between the second augmented record and the candidate entity record;
utilize, by the at least one processor, an index engine to identify the entity records determined to have a positive contrast with the candidate entity record;
wherein the index engine indexes the identified entity records into at least one database table so as to merge the entity records having the positive contrast with the candidate entity record.
16 . The system of claim 15 , wherein the set of predefined rules to generate an augmented record call for a random insertion of the at least one element of the candidate entity record.
17 . The system of claim 15 , wherein the set of predefined rules to generate an augmented record call for a random deletion of the at least one element of the candidate entity record.
18 . The system of claim 15 , wherein the set of predefined rules to generate an augmented record call for a random swap of the at least one element of the candidate entity record.
19 . The system of claim 15 , wherein the set of predefined rules to generate an augmented record call for a span deletion of the at least one element of the candidate entity record.
20 . The system of claim 15 , wherein the set of predefined rules to generate an augmented record call for randomization of the at least one element of the candidate entity record.
21 . The system of claim 15 , wherein the index engine utilizes a hamming distance to determine a positive contrast with the candidate entity record.