Automatic CRM Record Matching via Multi-Tier Weighted Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CRM systems require extensive manual data entry for matching communication records with CRM records, leading to partial, missing, and inaccurate data due to the complexity of matching the same salesperson, customer, or product across multiple records.
Innovation Solution
A multi-tier learning approach is applied to automatically match communication records with CRM records by computing global, company, and user weights using reference vectors, enabling accurate assignment and insertion of communication records into CRM systems without user intervention, utilizing a support vector machine for classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual data entry is used to match communication records with CRM records, then data accuracy can be maintained through human judgment, but extensive manual effort and time are required leading to partial and missing data
Solution Approach 1:
The system performs automatic matching of communication records with CRM records using machine learning algorithms, enabling the system to serve itself without requiring manual human intervention for each matching decision, thus dramatically improving productivity while maintaining consistency
Solution Approach 2:
The system incorporates feedback mechanisms where user corrections and manual adjustments to automatic matching results are captured and used to retrain and improve the machine learning models, continuously enhancing data accuracy over time while maintaining high automation levels
2Productivity
If automatic matching algorithms are implemented to improve data entry efficiency, then productivity increases, but matching accuracy deteriorates due to the complexity of identifying the same salesperson, customer, or product across multiple records
Solution Approach 1:
The system transforms the matching problem into a parameter-based comparison by extracting and comparing key attributes (salesperson names, customer names, product names, contact information) from records, using multi-tiered feature extraction to handle variations and improve matching precision automatically
Solution Approach 2:
The system uses composite matching criteria combining multiple data sources and verification methods, integrating information from communication records, CRM records, and external data sources to create a robust matching decision that improves accuracy beyond what single-method algorithms could achieve
3Quantity of substance
If extensive manual data entry is performed to ensure complete data capture, then data completeness improves, but the complexity of matching records increases and leads to partial and missing data
Solution Approach 1:
The system segments the complex matching task into distinct processing stages: extraction of key entities, normalization of data formats, comparison against multiple criteria, and hierarchical decision-making, making the overall complex process manageable and accurate through structured breakdown
Solution Approach 2:
The system creates a universal matching framework that handles multiple types of records (salespeople, customers, products, contacts) using the same core algorithmic approach, reducing complexity by applying a single versatile solution rather than separate specialized processes for each record type
Data Source
AI summary
A method for data processing includes obtaining from multiple different organizations (24, 52, 54, 56) customer relationship management (CRM) records (110) and communication records (90). Reference vectors (130) of feature values are computed for the communication records in a training set. Global weights are computed for the feature values by evaluating the reference vectors for all of the different organizations. For each organization, respective company weights are computed by evaluating specifically the reference vectors computed over the CRM records and communication records belonging to the organization. For each person belonging to a given organization, respective user weights are computed for the feature values by evaluating specifically the reference vectors computed over the communication records that identify the person as the user. The weights are applied in order to assign the communication records that are not in the training set to respective ones of the CRM records.