Automated company matching
An automated predictive analytics system disclosed herein provides a novel technique for industry classification. In one implementation, the analytic system provides a method of matching company names, the method including generating a provider-side company dataset, for each record in a client's company records, identify a list of match candidates from the provider-side company dataset that has overlapping company signature fragments, for each client-side company, compute match scores against the list of provider-side match candidates on each signature fragment and form an overall match score, for each client-side company, use the match scores to rank candidates and keep a predetermined number of top predetermined matches, and display the predetermined number of top predetermined matches using a user interface.
1. A method of matching company names, the method comprising:
receiving a provider side company dataset having multiple provider companies, each of the multiple provider companies having a respective company record;
generating provider side company signature fragments for each of the multiple provider companies based on the respective provider company records;
receiving a client side company dataset having multiple client companies, each of the multiple client companies having a respective client company record;
identify a provider side inexact match candidate based on a comparison of the provider side company signature fragments to the provider company record for at least one of the multiple client companies of the client side company dataset, the provider side inexact match candidate including an overlap of the provider side company signature fragment with the respective client side company record;
for each of the client side companies:
compute a match score for the provider side inexact match candidate;
rank the provider side company the match score, and
generate multiple top matches of the client side companies based on the ranking of the provider side match candidates; and
output the multiple top matches.
2. The method of claim 1 , wherein generating the provider side company dataset further comprises:
generating provider company signatures;
generating provider company signature fragments;
generating provider company name tokens;
generating a provider company signature; and
generating the provider side company dataset based on the provider company signature.
3. The method of claim 1 , wherein the comparison of the provider side company signature fragments to the client company record for the at least one of the multiple client companies of the client side company dataset includes comparing multiple subsequences of the provider side company signature fragments to respective multiple subsequences of the client side company dataset.
4. The method of claim 3 , wherein the multiple subsequences includes comparing two fuzzy subsequences.
5. The method of claim 3 , wherein the multiple subsequences includes comparing a fuzzy subsequence and a non-fuzzy subsequence.
6. The method of claim 1 , wherein the comparison of the provider side company signature fragments to the company record for the at least one of the multiple client companies of the client side company dataset includes determining a Levenshtein distance.
7. The method of claim 1 , wherein the comparison of the provider side company signature fragments to the company record for the at least one of the multiple client companies of the client side company dataset includes determining a cosine similarity.
8. The method of claim 1 , wherein the comparison of the provider side company signature fragments to the company record for the at least one of the multiple client companies of the client side company dataset includes determining a Euclidean distance.
9. The method of claim 1 , further comprising:
identifying multiple provider side inexact match candidates based on the comparison of the provider side company signature fragments to each of the client company records; and
computing the match score for each of the provider side match candidates based on the multiple provider side inexact match candidates.
10. The method of claim 9 , further comprising:
weighting each of the multiple provider side inexact match candidates; and
computing the match score for each of the provider side inexact match candidates based on the weighted multiple inexact match candidates.
11. The method of claim 1 , further comprising:
generating client side company signature fragments for each of the multiple client companies based on the respective client company records; and
identifying the provider side inexact match candidate based on a comparison of the provider side company signature fragments to each of the client side company signature fragments.
12. The method of claim 11 , further comprising:
identifying multiple provider side inexact match candidates based on the comparison of the provider side company signature fragments to each of the client side company signature fragments; and
computing the match score for each of the provider side match candidates based on each of the provider side inexact match candidates.
13. The method of claim 1 , wherein the multiple top matches are output to a remote computing device, network, or server.
14. The method of claim 1 , wherein the multiple top matches are output to a display of a user interface.