Multivariate Transaction Classification

Abstract

Embodiments relate to transaction classification based on analyzing multiple variables. For purchase transactions, such variables may include, but are not limited to: purchase location, source system, line of business, cost center, functional area, supplier capability, project description, account description, organization, department, custom parameters, and so on. Embodiments may rely on one or more classification schemes employed alone or in combination, such as statistical classification, semantic classification, and / or knowledge base classification. In purchase transactions, classification based on multivariate analysis facilitates the identification of purchased items or services, thereby bringing precision in the classification and assignment of centralized classification codes. Particular embodiments may include features that allow users to review / correct category assignments via a feedback loop linked to past categorizations. This revision feature can increase the transparency of current transactions, allow future classifications to be modified for continuous improvement, and provide a user-driven measure of system performance.

Description

technical field [0001] Embodiments of the invention relate to classification of transactions, and in particular, to systems and methods employing multivariate classification of transactions. Background technique [0002] Unless indicated to the contrary herein, the approaches described in this section are not prior art to the claims in this application and are not admitted to be prior art by inclusion in this section. [0003] Classification is the process of assigning classification codes from one or more taxonomies to transactions. An example of a taxonomy is the United Nations Standard Products and Services Code (UNSPSC). Such a taxonomy may also provide a framework for categorizing and performing analysis on spend within an organization. [0004] The classification process can involve organizing transactions (such as invoices and purchase orders) by business dimensions and their hierarchies to ensure a consistent and complete understanding of purchasing activity. This...

AI Technical Summary

Problems solved by technology

[0006] Such a univariate approach is simple but can lead to imprecision

For example, the reliance of expense classifications on item descriptions tends to cover direct costs, but may fall short because item descriptions of indirect costs are ignored

[0007] Similarly, basing classification solely on general ledger (G / L) accounts can lead to a bias towards 'why' something is bought rather than 'what' is actually bought

Finally, categorization by supplier alone can lead to oversimplification, since suppliers tend to sell more than one thing

As a result, alternative advantages such as combined shipping and services may remain unseen and unutilized

Images