Top Posts Tagged with #auto-indexing

Auto-indexing – Essential for loan document classification

For modern-day loan package processing, manual document management just doesn’t cut it.

Mutually Reinforcing Challenges

With loan document packages often exceeding 500 pages and potentially including over 100 unique document types, the mortgage loan origination lifecycle is heavily dependent on accurate and efficient document classification. Loan documents come from multiple sources, including brokers, lenders, borrowers, employers, and online vendors, and the set of documents required varies from state to state and county to county. Inconsistent and potentially confusing nomenclature can make it difficult to categorize documents by hand. For example, “1003”, “mortgage application form”, “uniform residential loan application”, and “URLA” all refer to the same document.

Furthermore, the PDF file created by combining scanned images of these documents doesn’t adhere to a single, standardized document order. Over the course of the loan origination lifecycle, document management teams will typically encounter multiple versions of many of these documents, too. As you can see, a competitive title company or document preparation firm cannot rely on manual efforts to manage documents in an accurate, timely, and cost-effective manner. For these reasons, loan servicers, title companies, and other organizations in the mortgage industry must instead turn to automated IT solutions.

OCR to the Rescue

One fundamental tool used for automatic document indexing is optical character recognition (OCR), which converts the pixels that make up an image of a page of text into a stream of characters that computers can interpret and store. OCR allows indexing software to automatically categorize software, group multiple versions of the same document, and extract information to be used by other software systems.

While some OCR vendors like Kofax, Xerox, and Paradatec offer generic character recognition capabilities that lack any specific knowledge of loan documents, whereas Visionet Systems uses OCR engines that have been trained using the various types of loan documents, encountered while indexing a loan package. This cloud based accelerator platform hosts solution suite with a powerful combination of NLP (Natural Language Processing), analytics, domain specific knowledge and OCR technology elements to delight customers with superior accuracy rates as compared to their generic counterparts. Armed with this domain-specific knowledge of document layouts and the types of information to expect in each form field, expert systems for loan document indexing significantly improve Cost, Quality and Time parameters of document indexing process.

VisiLoanReview: Visionet’s Auto-Indexing Solution

Given the various challenges to loan document management mentioned above, Visionet has put a lot of thought into designing VisiLoanReview (VLR), a solution for indexing and validating loan documents. The embedded OCR engine separates loan package PDF files into individual documents, which are then automatically classified and organized. VLR maintains a complete revision history for each document, making it easy to view and use information from the most current version. Once automatic indexing is complete, VLR highlights the exceptions, which can be manually checked to ensure the exceptions are taken care of appropriately. This entire process significantly cuts down the processing cost of a mortgage loan. It is a win-win solution for all the stakeholders: borrowers can get faster loans, brokers and lenders can get higher accuracy in lesser time and can manage different types of documentations predictably.

Future Directions

The encouraging 90%+ accuracy rate achieved by VLR with loan documents has led Visionet to repurpose this auto-indexing engine so that it can be used with other document types across all kinds of business processes. In the near future, Visionet’s auto-indexing technology will be offered for managing invoices, requisition forms, insurance documents, contracts, and many other complex business documents. These advanced technologies will help organizations streamline operational processes, cut costs, improve accuracy, and save time, resulting in a substantial competitive advantage.

About the author:

Alok Bansal Alok Bansal is Managing Director and Country Head of Visionet Systems Inc. and a 21 years of experience in managing strategy, global operations, optimizing and leading growth of financial services, mortgage banking, and BPO firms.

Source: https://www.visionetsystems.com/blog/auto-indexing-essential-loan-document-classification

#Auto-Indexing Technology #Auto-Indexing #Loan Document Classification #Loan Documents #Mortgage Loan Origination #Mortgage

•18+ Adults Only

Watch Anya Live on Cam

Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.

✓ Live Streaming✓ Interactive Chat✓ Private Shows✓ HD Quality✓ Free Actions

Free to watch • No registration required • HD streaming

VI High 14 - How to Select Random Elements from an Array in LabVIEW

In this episode we are preparing to select the winners of our sweepstakes giveaway. What better way to make the random selection than a LabVIEW VI? None. Join us as we look at how to randomly select items from an array and how to avoid the pitfalls of this seemingly simple task. For more on learning LabVIEW, check out the Sixclear Lucid LabVIEW Fundamentals Training (formerly Sixclear LabVIEW Fundamentals) course offered by Sixclear: sixclear.com.

Experience Level: Basic

(start transcription) "Avid Sixclear paparazzi are already well aware that entry to our sweepstakes giveaway of six licenses of the Sixclear Lucid LabVIEW Fundamentals Training (formerly Sixclear LabVIEW Fundamentals) course is days away from coming to a close. As such, we need a way to randomly select the six lucky winners from the eager entrant pool. What better way than LabVIEW?

Our task is well defined. We will have an array of strings with names of our entrants. From that array, we will randomly select 6 with LabVIEW. There are several ways to implement this, we’ve chosen the most hopelessly interesting. We will implement a solution in the following manner: 1. Generate 6 random numbers 2. Scale the random numbers by multiplying them by the size of the name array, this will generate an array of indices that can be used to pull out random names from the name array 3. Check for and replace duplicate indices 4. Use the indices to generate an output array of 6 random names Let’s head to LabVIEW and do it. I’ve already created a multicultural array of names on the front panel that we’ll use. Generating the 6 random numbers is easy. Just put a Random Number function in a For Loop running 6 times and auto-index the output. That occurs by default. Now to scale those to the size of the name array, we’ll take an Array Size function and use the polymorphic attribute of the Multiply function to multiply the array and the scalar output together. Now these numbers coming out are extremely precise fractional numbers and we just want whole numbers corresponding to indices, so we’ll use the Round Toward -Infinity function. For now, we’ll skip step 3 and just use this array of indices to generate our output names array. We’ll auto-index this array into a For Loop and pass the index to an Index Array function. The array we’re indexing is the full Name Array, and we’ll want to disable indexing on this tunnel coming in because we want that full array. Each iteration of this loop will give us one of the selected names, so we’ll auto-index the output to generate the full list of 6 names. If I run this a few times we see that this procedure does indeed work except that we sometimes run into duplicate indices as a result of rounding down each element. You can imagine that multiple numbers in this array could be rounded down to the same number and so we have duplicates. Obviously we have a small Name Array and so we’re more likely to get duplicates, but we need a robust way to find and eliminate those duplicates for any size array coming in. Remember, robustness. And remember that until next time because we are out of it. Uh, that’s time, not robustness, still plenty of that. In the meantime, you can enter the sweepstakes until June 9, 2011, just check out the entry requirements on our news blog: news.sixclear.com." (end transcription)

#LabVIEW training #LabVIEW tutorial #array manipulation #arrays #auto-indexing #for loop #labview #polymorphism #random number #shift registers