Accelerated Intelligent Enterprise with Timely Vendor Spend Analysis

Abstract

Disclosed are intelligent enterprise systems and methods that include an input / output (IO) interface configured to receive information related to accounting information, a processor interconnected with a memory configured to store the information received from the IO interface and analyze the information to provide timely vendor spend analysis, and a display device configured to display the provided timely vendor spend analysis.

Description

Cross - reference to related applications

[0001] This application claims the benefit of U.S. Patent Application No. 62 / 987,623, entitled "ACCELERATED INTELLIGENT ENTERPRISE INCLUDING TIMELY VENDOR SPEND ANALYTICS", filed on March 10, 2020, and is incorporated by reference in its entirety as if fully set forth herein.

Technical Field

[0002] The present invention is directed to enterprise systems, and more specifically, to intelligent enterprise systems including timely vendor spend analytics.

Summary of the Invention

[0003] Disclosed are intelligent enterprise systems and methods. The systems and methods include an input / output (IO) interface configured to receive information related to accounting information, a processor interconnected with a memory and configured to store the received information from the IO interface and analyze the information to provide timely vendor spend analytics, and a display device configured to display the provided timely vendor spend analytics.

[0004] The systems and methods may include received information including category - based spending. The systems and methods may include received information including vendor - based spending. The systems and methods may include received information including spending over a period such as a selected fiscal year.

[0005] The system and method may include received information processed by a processor to find a perfect match using nearest neighbor tendency scoring. The system and method may include received information processed by a processor to identify a match using a fuzzy matching algorithm. The system and method may include received information processed by a processor to identify a match using at least two fuzzy matching algorithms. The system and method may include matching based on cutoff scores. The system and method may include parent company information identified for at least one vendor.

[0006] The system and method may include performing text tokenization. The system and method may include learning an OEM based on the received information. The system and method may include determining a taxonomy category based on the received information.

[0007] The system and method may include the use of at least one hybrid model combining a support vector machine (SVM), a random forest (RF), a neural network (NN), and an expert system (ES).

[0008] The system and method may include a configuration in which the display device is configured to select vendor actual spending, peer comparison, market considerations, and normative considerations.

[0009] A more detailed understanding can be obtained from the following explanation, which is given as an example along with the attached drawings, and similar reference numerals in the drawings indicate similar elements. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 shows an example of a calculation system for evaluating vendor spending analysis.

[0011] [Figure 2]Figure 2 shows an illustrative screenshot of the system from Figure 1, illustrating vendor spending analysis by providing spending by category.

[0012] [Figure 3] Figure 3 shows an illustrative screenshot of the system from Figure 1, illustrating vendor spending analysis by providing vendor-specific spending.

[0013] [Figure 4] Figure 4 shows the vendor rollup flow corresponding to this enterprise in Figure 1.

[0014] [Figure 5] Figure 5 shows the flow for searching for OEMs from the products in the database in Figure 1.

[0015] [Figure 6] Figure 6 shows a screenshot of the enterprise system from Figure 1, illustrating the navigation windows for selecting vendor actual spending, peer comparison, market insights, and normative considerations.

[0016] [Figure 7] Figure 7 shows a screenshot of the enterprise system in Figure 1, illustrating the landing page for category spending as described herein.

[0017] [Figure 8] [Figure 9] Figures 8 and 9 show screenshots of the enterprise system in Figure 1, illustrating vendor spending after selecting vendor spending in the window of Figure 7.

[0018] [Figure 10] Figure 10 shows a screenshot of the enterprise system in Figure 1, illustrating vendor spending for a specific category selected in Figure 7 or Figures 8 and 9.

[0019] [Figure 11] FIG. 11 shows a screen depiction of the enterprise system of FIG. 1, showing additional details provided by selecting a particular vendor in the depiction of FIG. 10.

[0020] [Figure 12] FIG. 12 shows a screen depiction of the enterprise system of FIG. 1, showing additional details provided by selecting the vendor specified in the category of FIG. 10.

[0021] [Figure 13] [Figure 14] FIGS. 13 and 14 show a screen depiction of the enterprise system of FIG. 1, showing user-selectable toggles that enable showing spending trends, strategic value, vendor risk, preferred vendors, and exchange rates by depiction.

[0022] [Figure 15] FIG. 15 shows a screen depiction of the enterprise system of FIG. 1, showing a depiction of the actual spending of a particular vendor regardless of category.

[0023] [Figure 16] FIG. 16 shows a screen depiction of the enterprise system of FIG. 1, showing a depiction when proceeding to the actual spending of a particular vendor.

MODE FOR CARRYING OUT THE INVENTION

[0024] This invention relates to an intelligent enterprise that includes vendor spending analysis. The invention can accelerate enterprise intelligence using analytical artificial intelligence. This enterprise provides timely, transparent, and normative insights. Users can make smarter and faster decisions with reduced risk through the collaborative and normative intelligence of the enterprise. This enterprise may be used in conjunction with existing financial and vendor management solutions. This enterprise utilizes an independent overlay, a SaaS application that categorizes and visualizes vendor spending in conjunction with all financial systems.

[0025] In the following, the term “processor” includes, but is not limited to, single or multi-core general-purpose processors, dedicated processors, conventional processors, digital signal processors (DSPs), multiple microprocessors, one or more microprocessors associated with a DSP core, controllers, microcontrollers, one or more application-specific integrated circuits (ASICs), one or more field-programmable gate array (FPGA) circuits, other types of integrated circuits (ICs), systems-on-chip (SOCs), and / or state machines. In the following, the term “computer-readable storage medium” includes, but is not limited to, semiconductor memory devices such as registers, cache memory, read-only memory (ROM), dynamic random-access memory (D-RAM), static RAM (S-RAM), or other RAMs; magnetic media such as flash memory; hard disks, magneto-optical media; optical media such as CD-ROMs, digital versatile discs (DVDs), or Blu-ray discs (BDs); other volatile or non-volatile memory or other types of devices for electronic data storage. In the following, the term “memory device” refers to a device that can be configured to read and / or write data to one or more computer-readable storage media.

[0026] In the following, the term “display device” includes, but is not limited to, monitors or television displays, plasma displays, liquid crystal displays (LCDs), displays based on technologies such as front or back projection, light-emitting diodes (LEDs), organic light-emitting diodes (OLEDs), or digital light processing (DLP). In the following, the term “input device” includes, but is not limited to, keyboards, mice, trackballs, scanners, touchscreens, touchpads, stylus pads, and / or other devices that generate electronic signals based on interaction with a user. Input devices may operate using technologies such as Bluetooth, Universal Serial Bus (USB), PS / 2, or other technologies for data transmission.

[0027] Figure 1 shows an example calculation system 100 for determining vendor spending analysis. Vendor spending analysis for a service provider shows service provider performance in terms of aspects that are easily comparable to both ideal performance criteria and the performance of other service providers in the same industry or organization. As will be described in more detail below, vendor spending analysis is a weighted sum based on performance criteria and multiple weighting coefficients. Vendor spending analysis may be calculated by combining information about spending data from financial systems of metadata related to inter-firm transactions. For example, vendor metadata such as risk scores, diversity assessments, and other data fields may be used to contextualize spending data. By collecting these data points across multiple firms, peer comparisons of the collected information may be provided.

[0028] An example computing system 100 includes a processor 165, a memory device 170, and a communication bus 180 for data communication between components. The computing system 100 optionally includes several components, including a graphics subsystem (not shown), a network interface (not shown), an input interface 160, and a printer driver (not shown). The graphics subsystem may provide data to and / or drive a display device 140. The network interface provides a wired or wireless network connection that allows the computing system 100 to connect to a network-connected printer or the Internet. The input interface 160 may be, for example, a wireless connection, or a data port capable of receiving data from one or more input devices. The printer driver provides data to and / or drives a local printer.

[0029] The computing system 100 includes subsystems for evaluating service provider performance data, determining vendor spending analysis based on performance data, generating graphical output data for displaying results, and analyzing vendor spending analysis and related data. The subsystems include a performance data module, an aggregation module, a quantification module, a vendor spending analysis module, a display module, and a results analysis module. Modules may be implemented as software modules, dedicated processor elements, or a combination thereof. Suitable software modules include, for example, executable programs, functions, method calls, procedures, routines or subroutines, sets of processor-executable instructions, scripts or macros, objects, or data structures. Some or all of the modules may, as further examples, be implemented as modules or subroutines in spreadsheet programs such as Microsoft Excel®, database programs, or other types of programs. Data used and / or generated by the modules may be stored in the memory device 170. The processor 165 can be configured to operate and control the operation of the module, and can also be configured to control communication between the module and other components of the computing system 100, such as the printer driver, graphics subsystem, memory device 170, network interface, and input interface 160.

[0030] Generally, the accounting system 120 is used to summarize, analyze, and report on a company's financial operations. The accounting system 120 may include a ledger / expense / credit / revenue system 128, a liability system 126, assets 124, and a budget 122, as described below. The accounting system 100 uses information from the ledger / expense / credit / revenue system 128, the liability system 126, assets 124, and budget 122 to determine vendor expenditure analysis. Information within the accounting system 100 includes data supplied from the billing system, purchase orders, accounts payable, invoice details, general ledger charts, and business hierarchy. Data from these sources is cross-referenced by the system 100 to organize, classify, and enrich the data.

[0031] Generally, the ledger in the ledger-expense-credit-revenue system 128 may have a general ledger. A general ledger represents a record-keeping system for a company's financial data, including records of debit and credit accounts verified by a trial balance. A general ledger provides a record of each financial transaction conducted in the operation of a business. A general ledger holds the account information necessary to prepare a company's financial statements, and transaction data is separated by type into asset, liability, equity, revenue, and expense accounts.

[0032] In the ledger, expense, credit, and revenue system 128, an expense is a ledger record representing a payment or the occurrence of a liability in exchange for goods or services. The documentary evidence arising from the expense is a sales receipt or invoice.

[0033] Debits and credits are recorded in the ledger to document changes in value resulting from commercial transactions. A debit entry in an account represents a movement of value into that account, while a credit entry represents a movement of value out of an account. For example, a tenant paying rent to a landlord may debit the rent expense account corresponding to the landlord, and the landlord may credit the accounts receivable account corresponding to the tenant. In every transaction, both debits and credits occur for each party involved, and the sum of the debits and credits of each party for the same transaction is equal.

[0034] In the ledger, expense, credit, and revenue system 128, revenue is the income that a business earns from its ordinary business activities, usually from the sale of goods or services to customers. Revenue is also called sales or sales revenue.

[0035] Accounts Payable (AP) 126 is a general ledger account that represents a company's obligation to repay short-term debts to creditors or suppliers. AP 126 includes accounting for payments that a company owes to suppliers and other creditors. AP 126 represents the amount still unpaid to vendors or suppliers of goods or services received. The total outstanding balances to vendors are shown on the company's balance sheet as accounts payable.

[0036] Assets (124) are all resources owned by a business. Assets are tangible or intangible things that can generate value through ownership or control, and that, when held by a business, produce beneficial economic value. Simply put, assets represent the value of ownership that can be converted into cash.

[0037] A budget 122 represents a financial plan that includes both financial and non-financial information. Its most obvious features are revenue estimates, selling expenses to a given organization, and other expenses. Budgets can also include non-financial information, such as how many employees are considered necessary. While a budget 122 is a forecasting document, companies also use it as a financial control tool. Financial controls are tools for monitoring the activities of one's own business.

[0038] The vendor expenditure analysis module may be contained within the data system 150 and / or the accounting system 120. The vendor expenditure analysis may determine a weighting coefficient for each quantitative performance value, multiply each quantitative performance value by the corresponding weighting coefficient, and sum the products of the multiplied quantitative performance value and the weighting coefficient to determine the vendor expenditure analysis. The weighting coefficients may be based on user input received through an input interface. The weighting coefficients may be adjusted so that the vendor expenditure analysis focuses on aspects of performance criteria that the user deems most important and / or does not emphasize criteria that the user deems unimportant. In one use case, the user may set the performance coefficients to indicate that a particular performance criterion is extremely important. At a later point, the user may adjust the weighting coefficients to reflect that other performance criteria have become more important.

[0039] The display module may receive data generated by the quantification module and the vendor spending analysis module and generate data for displaying the received data. The display data may be communicated by the display module to a printer driver and a printer, and the printer may create a printable version of the display data. The display data may also be communicated by the display module to a graphics subsystem for the display device 140, and the display device 140 may render the display data for the user. Visual examples of display data generated by the display module and which can be printed by the printer or rendered by the display device 140 are described in detail below.

[0040] The display data generated by the display module may take various forms depending on the various implementations of the display module and the computing system 100. For example, the generated display may represent the content of an image, one or more pages in a document, or other visual artifacts. The generated display data may be compressed or uncompressed image data conforming to formats such as Joint Photographic Experts Group (JPEG), Tagged Image File Format (TIFF), Scalable Vector Graphics (SVG), bitmap, or other formats. The generated display data may be data or instructions that can be interpreted by a window manager, desktop manager, graphical user interface, printer driver, graphics driver, word processor, document display or editing program, or other software or hardware components relating to the implementation of a user interface and / or graphics.

[0041] The results analysis module may receive data generated by the vendor expenditure analysis module and determine whether to modify the parameters used by the accounting system 120 and the vendor expenditure analysis module. For example, the results analysis module may compare the generated vendor expenditure analysis values ​​for multiple vendors. If the majority of vendors have vendor expenditure analysis values ​​above a threshold, the results analysis module may adjust the parameters used by other modules so that the generated vendor expenditure analysis values ​​decrease. Alternatively, if the majority of vendors have vendor expenditure analysis values ​​below a threshold, the results analysis module may adjust the parameters used by other modules so that the generated vendor expenditure analysis values ​​increase. Parameters that can be modified by the results analysis module include performance value criteria, performance value sub-criteria, and weighting coefficients used by other modules.

[0042] Furthermore, the accounting system 120 may perform electronic commerce or banking functions based on data received from the quantification module and / or the vendor expenditure analysis module. Data showing past payments to service providers may be available to the results analysis module in a memory device and / or via a network interface. Payments to service providers may be conditional on subsequent vendor expenditure analysis, performance criteria and / or performance sub-criteria, and rules linking performance to compensation may be stored in the memory device. Based on the compensation rules and data received from the quantification module and / or the vendor expenditure analysis module, the accounting system 120 may determine that a debit or credit should occur to the service provider's account. The accounting system 120 may then initiate an electronic transaction to generate the determined debit or credit. For example, if the results analysis module determines that the service provider has exceeded the required vendor expenditure analysis or service performance criteria or sub-criteria, the results analysis module may initiate a bonus payment to the service provider. If the results analysis module determines that it has overpaid the service provider for the work performed, the results analysis module may initiate a debit to the service provider's account. The results analysis module may initiate credit or debit transactions via a network interface using technologies such as Electronic Funds Transfer (EFT) or other appropriate services or protocols.

[0043] Furthermore, the accounting system 120 may analyze data received from the quantification module and / or vendor spending analysis module based on triggers, and notify the user when a triggered event occurs. For example, a trigger may be set based on whether a service provider exceeds the required vendor spending analysis, one or more performance thresholds, or one or more performance sub-thresholds. Alternatively, a trigger may be set based on whether multiple service providers have the required vendor spending analysis, one or more performance thresholds, or one or more performance sub-thresholds. If a trigger condition exists, the results analysis module may generate a notification email to send to the user indicating that the trigger condition has been met. Alternatively, the accounting system 120 may communicate with the display module to generate display data indicating that the trigger condition has been met. The corresponding notification may then be displayed on the display device 140 for the user. The user may use the notification, for example, as a basis for renegotiating SLA terms with one or more service providers.

[0044] The computing system 100 may be implemented, for example, as a computer, a system-on-a-chip, or other computing or data processing device or apparatus. The processor 165 can be configured to execute instructions that specify the function of the module, as described above with reference to Figure 1. The instructions may be stored in the memory device 170 and / or one or more additional computer-readable storage media (not depicted).

[0045] Figure 2 shows an exemplary screenshot 200 of System 100 from Figure 1, illustrating vendor spending analysis by providing categorized spending 210. Screenshot 200 shows that the company spent $149.9 million 220 in fiscal year 2019. This spending data is then divided into specific spending categories, and the color and size of the display area are used to indicate the direction and magnitude of the amount spent in each category. As shown in screenshot 200, Application 230 has the largest spending, corresponding to the largest depicted box and an amount of $66.9 million spent. Big Data 240 is the second largest spending, depicted using the second largest box, representing an amount of $27.1 million spent. Calculation 250 is the third largest spending, representing $26.5 million, and is slightly smaller in size than Big Data, representing a slightly smaller spending. Further expenditures are also shown, including non-IT 260, platform 270, end-user 280, delivery 285, security 290, and IT management 295. Each box is sized according to the category and the amount of expenditure it corresponds to.

[0046] For example, the entire display could be used to represent the total expenditure. From there, the size of each category could be determined according to the percentage that the category's expenditure represents of the total expenditure.

[0047] In addition to the size of the depicted boxes, the color of the boxes may also be used to provide additional information. This information may include spending trends for each category.

[0048] Figure 3 shows an exemplary screenshot 300 of the system 100 from Figure 1, illustrating vendor spending analysis by providing vendor-specific spending 310. Similar to the depiction in Figure 2, the screenshot 300 in Figure 3 again shows that the company spent $149.9 million in fiscal year 2019 320. This spending data is then broken down into specific vendors and depicted using the color and size of the display area to indicate the direction and magnitude of the amount spent by each vendor. As shown in the screenshot 300, the first vendor 330 has the largest spending, corresponding to the largest depicted box and an amount of $36.3 million spent. The second vendor 340 has the second largest spending, depicted using the second largest box representing an amount of $22.6 million spent. The third vendor 350 has the third largest spending, representing $18.2 million spent, and is slightly smaller in size than the depiction of the second vendor 340, representing a slightly smaller spending. Further spending is also shown for other vendors, including vendors 4 through x 360. Each box is sized according to the vendor and the corresponding expenditure amount.

[0049] For example, the entire display could be used to represent the total expenditure. From there, the size of each vendor could be determined according to the percentage that the vendor's expenditure represents of the total expenditure.

[0050] In addition to the size of the depicted box, the box's color may also be used to provide additional information. This information may include spending trends regarding vendors.

[0051] Figure 4 shows the vendor rollup flow 400 corresponding to the system 100 in Figure 1. In the vendor rollup shown in Figure 4, vendor data is extracted from the database 41 of the system 100 in Figure 1. This extracted vendor data makes it possible to illustrate, for example, in Figure 3, how much a company spends in total on multiple vendors rolled up into one company through corporate acquisitions such as mergers and acquisitions, showing this to users such as CIOs, CMOs, and CFOs. Vendor rollup may also be called parentage.

[0052] Vendor rollups allow for the identification of all manufacturers of purchased goods through OEM disclosure, and further enable linking to OEM owners. The purchasing company may be automatically and in real-time or near real-time if some vendors have been acquired by other companies. This feature also allows for notifications regarding potentially redundant, wasteful, and inadequate deal negotiations.

[0053] For example, a major financial institution spends $30 million annually on VMware, $5 million on Pivotal, $25 million on EMC, and $20 million on Dell. Purchasing managers in the company's procurement department may not be aware that Dell acquired VMware, Pivotal, and EMC. Therefore, the financial institution could view this as a single, much larger expenditure of $70 million on Dell, rather than negotiating each contract individually for small amounts. This larger expenditure could provide a better perspective for negotiating pricing with each of the commonly owned vendors, potentially resulting in significant savings of tens of millions of dollars.

[0054] Identifying actual companies requires normalization of supplier and OEM company names. The method in Figure 4 allows for abbreviations of company names to initials, synonyms, common names (DBA), or alternative names (AKA) when importing them. Further improvements to the method allow for considering companies located in different states or countries as the same entity. Additionally, names may be partially or entirely stemmed to create the normalized names. These normalized names are stored in the database in Figure 1 and reused in association with the appropriate companies.

[0055] Furthermore, multi-stage fuzzy matching algorithms 4301 and 4302 are used for further normalization. The algorithms use statistically tested cutoff scores to produce accurate matching results. Figure 4 shows a detailed diagram of the process flow.

[0056] Specifically, Figure 4 shows the organization database of Figure 1, which provides inputs for new companies and vendors to be searched to a first step 420 in which a propensity score matching algorithm identifies the nearest neighbor and finds a perfect match. The output of this test is provided to an algorithm 450 that determines the parent company of a vendor or company. The algorithm also outputs a perfect match to second algorithms 4301 and 4302 that use fuzzy matching as described herein. The output of fuzzy matching 4301 is provided to the second fuzzy matching algorithm 4302, where items 440 are filtered by a predetermined cutoff score. These values ​​are then input to the algorithm 450 that determines the parent company of a vendor or company. The results are then fed back to the organization database 470, which outputs a final dataset 460 for vendor OEM rollups.

[0057] The system intelligently determines and reveals vendor parent-child relationships resulting from mergers and acquisitions. This solution can reveal critical relationship data in less time than any existing process or solution because current methods fail to normalize company names. This failure leads to inaccurate associations, even if they are established. The system improves parent-child information to include all known historical parent organizations and the dates on which parent-child relationships changed, enabling the ability to analyze spending over specific time periods. This is particularly important when we consider benchmarks and trends.

[0058] As described above for Figure 4, accounts payable, invoice, and purchase order data are used to determine a company's spending on OEMs and vendors at the parent level. By using artificial intelligence and machine learning in the vendor taxonomy, the original OEM can be determined regardless of who the goods or services were purchased from or whether OEM data is included in the customer data. The combination of artificial intelligence and machine learning creates "reseller spending visibility." This visibility reveals and visualizes a company's actual spending on original equipment manufacturers ("OEMs"), which is often invisible to most systems due to purchases from or through multiple resellers.

[0059] For example, a company might spend over $2 million per month ($24 million per year) on Cisco products, but this isn't apparent from any of the company's financial records. This is because purchase invoices, orders, and resellers indicate that Cisco products were purchased through numerous VARs (Value-Added Resellers) and resellers. Large organizations can negotiate better terms by presenting where their vendors are located on the company's vendor list. In this example, if Cisco is one of the top five vendors, leverage could be applied based on this relationship. This kind of information is not revealed in existing financial systems when an OEM requires an organization to purchase its products through a VAR or reseller.

[0060] This system performs vendor aggregation similar to that described above. This includes not only what occurs at the OEM level but also the integration of “non-OEM” vendor spending. The need for vendor aggregation arises from multiple vendor instances in the source data (inconsistent trading partner vendor tables) and corporate dependencies that exist outside the source data. Spending amounts may be categorized into standardized technology business management categories to promote transparency in showing how money paid to vendors is allocated or used within the organization.

[0061] In the database shown in Figure 1, a core OEM hierarchy table is constructed. The tables link to themselves, forming a parent-child hierarchy. Customer spending data is ingested and matched with potential naming by machine learning, which may be unique to the supplier name (vendor in the purchase record), OEM name, manufacturing part number, model number, and trading partner data. This method uses enhanced NLP (natural language processing) to process the spending description. The results of the spending description are then passed to a deep learning model that includes an inference engine with heuristics that provide speed and scale. Linear support vector machines and random forests also process the final output. The final output 460 is used to determine whether the input OEM name matches an existing OEM name. If system 100 does not have sufficient confidence in the matching, the combination of purchase data and the input OEM name is passed through a neural network. If there is still not sufficient confidence in the matching, the input OEM name is passed to a human curator for analysis. The human curator attempts to match it with an existing OEM name or points out that it is a new OEM. The system creates new OEMs as needed. Once a match is indicated, either manually or by a neural network, the entered OEM name is added as an alias to the existing OEM, eliminating the need for further manual curation. Once the spending data is analyzed and matched, the spending amounts can be aggregated by parent OEM, appropriate to the OEM ownership hierarchy at the time of spending.

[0062] The OEM hierarchy tables are kept up-to-date through a combination of human curation and automated ingestion of organizational hierarchy data from companies and open ownership sources such as LexisNexis, Dun & Bradstreet, and Moody's. The OEM hierarchy is also tracked by date range, as corporate ownership of organizations changes over time.

[0063] Most companies fail to accurately report vendor and OEM expenses. This is because accounts payable and purchase order (PO) systems are not designed to report with this level of detail, manual data entry when creating purchase orders, such as product information, is often not systematized, and financial reporting systems are focused on financial reporting. Accounts payable systems do not necessarily need to include detailed account items as part of payment records for product information. Typically, financial reporting systems do not include vendor details for expenses recorded and processed through fixed asset systems. Even financial system reporting may aggregate data based on general ledger accounts, but generally does not maintain records of sufficient vendor detail to provide accurate vendor reporting. This system 10 intelligently identifies OEMs through a combination of technologies to achieve a higher matching rate in a shorter time than any existing process or solution.

[0064] Accurate vendor spending aggregation is a daunting task. In large companies, vendors often overlap in accounts payable and purchase order systems, due to factors such as payments to multiple addresses, global regional instances of a given vendor, user biases like the use of inconsistent abbreviations in entered vendor names, and the maintenance of generally inconsistent vendor tables. This system normalizes and organizes inconsistent vendor names by "compressing names." For example, in the accounts payable vendor table in Figure 1, IBM appears in multiple forms. These may include, for example, IBM, IBM (I.B.M.), International Business Machine, and International Business Machine SA. Training the system to recognize each name as the same organization is essential for efficient and accurate vendor aggregation.

[0065] When processing a transaction, the first items to be resolved are the supplier's name and the manufacturer's name. These may vary in spelling accuracy, use of kana, and abbreviations. For example, Dell, Dell Technologies, Inc., and Dell Corp ltd should all refer to Dell. For example, HP and Hewlett Packard are the same, but HPS is different. System 100 matches the entered vendor name to an existing hierarchy by interpreting and normalizing the various variations of company names. System 100 also simplifies the dataset through a visualization engine.

[0066] Figure 5 shows Flow 500 for searching for OEMs from products in the database in Figure 1. Item descriptions are used to train a model to recognize words for classifying OEMs. Manufacturers use phrasing created to differentiate their products. This phrasing is not necessarily captured at the purchase order account item level, but is often captured at the invoice account item level. In topic modeling, the concept of word vectors may be used to search for topics of interest, and in this system, it is used to search for OEMs from product descriptions. This method requires a large corpus of data from OEMs to lead to a high degree of certainty.

[0067] Flow 500, which is implemented, employs an incremental approach to learn the OEM. The OEM is validated, the model is updated, and the training iterations improve.

[0068] Flow 500 may be able to further leverage online learning by forming a robust foundation that gradually improves, taking into account the fact that the words describing a category may change over time. Each word is taken as a random variable, and multiple combinations are the conditional probability that those words are a given OEM. The set of features, which are the words used, may change over time, and this set of features is defined by Equation 1.

number

number

[0069] While other organizations use catalogs of products manufactured by OEMs, this system uses a combination of artificial intelligence and machine learning, along with purchase item data, reseller information, and expenditure classifications, to identify OEMs based solely on product descriptions. Furthermore, the system differentiates its use of product-level information through a visualization engine.

[0070] Specifically, Figure 5 shows that in step 510, the data is in PDF format. The flow is shown in the form of input in CSV or other file formats. In step 520, the text description of the invoice PO account item is determined. A random forest method may be used to learn the categories, and variable importance may be used to prune the dictionary of terms used. Alternatively, in step 530, the system may convert the text into terms and then use term frequency and inverse document frequency to remove N-grams that do not support class separation.

[0071] In step 540, tolerance may be performed from the text description. This may include removing stop words and creating N-grams spanning entire words, ranging in length from one to three words, while maintaining order. The enriched data from step 530 may be further supplied for tolerance.

[0072] In step 550, the system may learn which N-grams match the categories. From there, in step 570, it may learn the OEMs, and in step 590, it may learn the taxonomy categories to identify the OEMs from the descriptions.

[0073] In addition to providing transparency about "who" a company is spending money on, an equally important data point is providing transparency about "what" a company is buying. By using predefined and recognized taxonomies, combined with the ability to display datasets through taxonomies, systems can provide a standardized context for spending aggregations. Standardized taxonomies enable analyses such as benchmark testing and anonymous peer-to-peer comparisons. Custom taxonomies via customer portals may be configured to allow customers to create further data visualizations.

[0074] For businesses, the basic principle is to define a predetermined number of levels for their taxonomy and render them as a single row in the database. This system offers maximum flexibility by allowing the use of any number of levels to accommodate all trading partners. This system may be used to encompass both IT and non-IT spending.

[0075] The taxonomy to be set may be defined at the beginning. The model must then learn to automatically classify expense account items into appropriate categories. This automatic classification is achieved using a hybrid model that combines a support vector machine (SVM), a random forest (RF), a neural network (NN), and an expert system (ES). The random forest may be used to detect N-grams correlated with categories. This process reduces many variables from 6,000 to between 200 and 500. These variables may be used to construct the SVM and NN. In the SVM and NN, the results are adjusted to show the probability that an item actually belongs to a class. This probability is used to determine which items need to be reviewed manually for appropriate class assignment. Using both SVM and NN is equivalent to obtaining expert agreement by having two experts verify that an item belongs to a particular class. If this method produces inconsistent results, the class for that item is selected or sent to human review based on the certainty of the method and the difference in that certainty.

[0076] Expert Systems (ES) are for items with excellent inference rules for determining the correct class label. This is a preliminary process until SVMs, RFs, and NNs have sufficient capacity to learn categories and perform automatic classification. RFs are time-consuming but closely match the way experts infer class labels.

[0077] This approach utilizes both hybrid methods and methods where data is prepared through learning to successfully implement the taxonomy. A trade-off with expert system rules is essential to incorporate inference rules from experts and feed results into the engine for training. The initial data volume is small, requiring a means to acquire high-quality labeled data. Expert systems provide this bridge.

[0078] Figure 6 shows a screen rendering 600 of the enterprise system 100 from Figure 1, which displays a navigation window for selecting vendor actual spending 610, peer comparison 620, market insights 630, and normative considerations 640. This screen may also be used to generate and utilize aggregated data from the database in Figure 1. The system initiates actual spending analysis by selecting the actual spending navigation selection button.

[0079] Figure 7 shows a screen rendering 700 of the enterprise system 100 in Figure 1, illustrating a landing page for the category spending 610 described herein. This uses the taxonomy described above in Figure 2, which includes applications 230, big data 240, computing 250, platforms 270, end users 280, delivery 285, security and compliance 290, IT management 295, and non-IT (not shown). An IT pattern is provided to present, and in this example, annual spending 710 is provided. The presentation may be modified, for example, by changing each of the variables, using different colors for trends, or using different display timeframes.

[0080] Figures 8 and 9 show a screen rendering 800 of the enterprise system 100 in Figure 1, illustrating vendor spending after selecting vendor spending 610 in window 700 of Figure 7. Figure 8 represents a partially modified screen rendering 800 from screen 700 in Figure 7, and Figure 9 represents a more completely modified screen rendering 800 from screen 700 in Figure 7. The screen rendering 800 provides vendor-based spending data within the taxonomy of Figure 2, again including applications 230, big data 240, computing 250, platforms 270, end users 280, delivery 285, security and compliance 290, IT management 295, and non-IT (not shown) as shown in Figure 7. Specifically, the size 810 of each vendor box is allocated based on the spending level within the taxonomy. Red and green may be used to indicate increased (red) and decreased (green) expenditures, respectively, with the shade of each color indicating the degree of increase or decrease. Furthermore, the identified vendors may describe the resellers that may be involved in the chain. As is evident in comparison with the depictions in Figures 8 and 9, the vendor boxes replace the expenditure values ​​depicted in Figure 7. Each vendor that makes up part of the taxonomy in Figure 2 is represented by size and color, which provide meaning to the data, as described herein.

[0081] Figure 10 shows a screen rendering 1000 of the enterprise system 100 in Figure 1, showing vendor spending for a specific category selected in Figure 7 or Figures 8 and 9. In this rendering, compute 250, which is part of the taxonomy, is selected and shown. Screen rendering 1000 provides vendor spending for the selected compute 250. As is evident from a comparison with Figure 9, for example, this expands the display of the compute taxonomy 250 and the underlying vendors corresponding to the compute 250 category. Screen rendering 1000 highlights details about the benefits of vendor integration that can be achieved.

[0082] Figure 11 shows additional details and exemplary calculation 250 taxonomy selections from Figure 10, and further provides details by selecting a specific vendor 1110 in the depiction of Figure 10, and shows a screen depiction 1100 of the enterprise system 100 in Figure 10. For example, by selecting VMware 1110, additional details are provided regarding the top spenders within the company identified by name in this category and the total spending in the category. Such information may also provide information about the person in charge within the company for a given vendor.

[0083] Figure 12 shows a screen rendering 1200 of the enterprise system 100 in Figure 1, illustrating additional details provided by selecting vendor 1110 identified in category 250 of Figure 10. This makes it possible to display a specific order 1210 in category 250 and vendor 1110. Furthermore, through interaction with screen 1220, additional information 1230 regarding the specific order, such as order number, date, cost, and product, may also be provided to the user.

[0084] Figures 13 and 14 show a screen rendering 1300 of the enterprise system 100 in Figure 1, which shows a user-selectable toggle 1305 that allows for the visualization of spending trends 1310, strategic value 1320, vendor risk 1330, preferred vendors 1340, and exchange rates 1350. The toggle 1305 allows for the selection and provision of other metadata to the user. This visualization system may provide output renderings of any other metadata found within system 100. For example, vendor risk 1330 in the taxonomy of Figure 2, which again includes applications 230, big data 240, computing 250, platforms 270, end users 280, delivery 285, security and compliance 290, IT management 295, and non-IT (not shown), as shown in Figure 7, may be displayed and color-identified by the metadata selection via this additional toggle 1305.

[0085] Figure 15 shows a screen rendering 1500 of the enterprise system 100 of Figure 1, showing the actual expenditures 1520 of a specific vendor 1510, regardless of category. As depicted, each vendor 1510 is identified and presented based on the amount of expenditure 1520, taking into account the underlying related organizations as described herein. The amount of expenditure 1520 is represented by the size of the box, and the trend direction of expenditure 1520 is represented by color identification.

[0086] Figure 16 shows a screen rendering of the enterprise system 100 from Figure 1, illustrating the process when moving to actual spending on a specific vendor 1610. This illustrates the dependent spending incurred by the enterprise towards vendors.

[0087] While features and elements have been described above in specific combinations, those skilled in the art will understand that each feature or element can be used individually or in any combination with other features or elements. Furthermore, the methods described herein may be implemented in computer programs, software, or firmware embedded in a computer-readable medium for execution by a computer or processor. Examples of computer-readable mediums include electronic signals (transmitted via wired or wireless connections) and computer-readable storage media. Examples of computer-readable storage media include, but are not limited to, read-only memory (ROM), random-access memory (RAM), registers, cache memory, semiconductor memory devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROMs and digital multi-purpose discs (DVDs). Software and associated processors may be used to implement the systems and methods described herein.

Claims

1. An intelligent enterprise system for a company, comprising: an input / output (IO) interface configured to receive information relating to accounting information; a processor interconnected with the memory for storing received information from the IO interface and configured to analyze the received information to provide a timely vendor spend analysis; a display device configured to display the provided timely vendor spend analysis; A system comprising: the received information includes expenditure records of the business; the timely vendor spend analysis is comprised of spend records of the enterprise, identifies a plurality of spend categories with associated direction and magnitude of spend for each category, and indicates vendor performance including vendor metadata; The processor, in analyzing the received information, using the expenditure records to categorize vendor expenditures of the enterprise into a plurality of expenditure categories; aggregating said company's vendor spending by spending category; calculating a spend analysis for said vendor based on the value of said vendor's performance criteria; A system characterized by:

2. the received information includes expenditures by category; 2. The system of claim 1 .

3. the received information includes expenditures by vendor; 2. The system of claim 1 .

4. the received information includes expenditures over a period of time; 2. The system of claim 1 .

5. the period is a selected fiscal year; 5. The system of claim 4.

6. processing, by the processor, a vendor name contained in the expenditure records of the received information using nearest neighbor propensity scoring to find an exact match for the vendor name in an organization database; 2. The system of claim 1 .

7. processing, by said processor, a vendor name contained in the expenditure records of said received information using a fuzzy matching algorithm to identify vendors in an organizational database that match said vendor name; 2. The system of claim 1 .

8. processing, by the processor, a vendor name contained in the expenditure records of the received information using at least two fuzzy matching algorithms to identify vendors in an organization database that match the vendor name; 2. The system of claim 1 .

9. The matching is based on a cutoff score.

9. The system according to claim 7 or 8.

10. The system of claim 1, a database for storing the hierarchical table; the hierarchical table being linked to itself to form a parent-child hierarchy; using the hierarchy table to identify parent company information for at least one vendor; A system characterized by:

11. Text tokenization is performed, 2. The system of claim 1 .

12. learning an OEM based on line items of the company's expenditure records contained in the received information; In the learning, the item descriptions of the account items are used to train a model for determining the OEM.

2. The system of claim 1 .

13. performing an automatic classification of the account items into appropriate taxonomy categories to determine taxonomy categories based on the account items of the expenditure records of the enterprise contained in the received information; 2. The system of claim 1 .

14. In the automatic classification, the appropriate taxonomy category is determined from the text description of the account item using at least one hybrid model that combines a support vector machine (SVM), a random forest (RF), a neural network (NN), and an expert system (ES).

14. The system of claim 13.

15. the display device is configured to select vendor actual spend, peer comparison, market insight, and normative insight aspects; 2. The system of claim 1 .

16. A method for providing an analysis of vendor spend of a company, comprising: receiving information regarding accounting information via an input / output (IO) interface; storing the received information via a processor interconnected with a memory; and analyzing the received information to provide a timely vendor spend analysis. displaying, via a display device, the provided timely vendor spend analysis. A method comprising: the received information includes expenditure records of the business; the timely vendor spend analysis is comprised of spend records of the enterprise, identifies a plurality of spend categories with associated direction and magnitude of spend for each category, and indicates vendor performance including vendor metadata; In analyzing the received information, using the expenditure records to categorize vendor expenditures of the enterprise into a plurality of expenditure categories; aggregating said company's vendor spending by spending category; calculating a spend analysis for said vendor based on the value of said vendor's performance criteria; A method comprising:

17. the received information includes at least one of spending by category and spending by vendor; 17. The method of claim 16.

18. analyzing the received information, said processor using nearest neighbor propensity scoring to find vendors in an organization database that have exact matches to vendor names contained in expenditure records of said received information; and at least one of: using a fuzzy matching algorithm to identify vendors in the organization database that match the vendor name; 17. The method of claim 16.

19. learning an OEM based on line items of the expense records of the enterprise contained in the received information, using line item descriptions of the line items to train a model for determining the OEM; determining a taxonomy category based on an item of an expenditure record of the company contained in the received information, and automatically classifying the item of the expenditure record into an appropriate taxonomy category; 17. The method of claim 16.

20. In the analysis, at least one hybrid model combining a support vector machine (SVM), a random forest (RF), a neural network (NN), and an expert system (ES) is used to determine a taxonomy category of an item of account from a text description of the item of account in the expense record of the enterprise.

17. The method of claim 16.

Images