Published Feb 16, 2017 by Grapple Data

We are pleased to announce the latest release of AKIN HyperSearch It brings with it a newly designed user interface layout and styling. We expect to continue to make improvements in the area of UI design. However, our primary focus remains on continuing to build a faster, more efficient, accurate, and intelligent search engine, expanding the functionality provided by the application as a whole, and working to integrate it with upcoming business and enterprise related product releases. This is an exciting time for our company as we continue working to deliver greater value to our customers.

Published on May 17, 2015 by Grapple Data

We are all intimately acquainted with the challenges of finding the correct document fast when we need it. That’s why so many people look for better search tools like AKIN HyperSearch for their desktop.

However, once we actually find the document(s) we are looking for, our overall sense of well-being and productivity can be significantly impacted when we have to hunt through large documents we've found to get at the actual actual content we need. Yes, most document editors or word processors like MS Word or Adobe Acrobat give us basic tools to search for text, but:

1) They rely on exact textual matches

2) They generally have no sense of semantics or textual meaning

This means that if you don't know EXACTLY what you are looking for, you will have to try one search word or phrase, try another, try again, over and over again, hopping around in the document to hopefully find the content you want, or you might resort to just scanning the document manually which is obviously less than optimum. To help make our customers lives easier and hopefully grease the skids of productivity, in our latest release we have focused a great amount of attention toward helping to get you to the document content you are looking for faster and with significantly fewer iterations than ever possible before.

NEW FEATURE 1 - Deep Document Content Search

This feature itself provides two new sub-features:

1) Whenever you do a search for documents in the main search window, and click on the "Content Results" section tab, like before, you will be shown all documents that were found to have that content However, now when you click on one of those documents in that view, the document will immediately open up in our new Deep Document Content Search form and immediately initiate a deeper fuzzy content scan of the document for you and provide you with a list of all the results right beside the raw (unformatted) document text itself. The results are sorted by similarity and by word match proximity, so that the most likely relevant items are presented at the top of the results. Clicking on an item in the list takes you to that text found in the document.

2) In the main application window there is now a new button near the filters button called "DS". This button opens a blank Deep Content Search form from which you can either "Browse" to any supported file and search its contents, or you can also right click into the text area and paste the contents off the copy clipboard. This is really handy when you want to search through content from something like an unsupported file. If you can open it in a native editor and copy its text, you can then paste it into the new DS form in AKIN HyperSearch. You can obviously also do this with things like large web page content as well. As long as you can get the raw text to the clipboard you can use this tool to search through the content using AKIN's powerful fuzzy content search logic and pattern recognition engine.

NEW FEATURE 2- Semantics-Aware Early-Discovery

"Early Discovery" refers to the idea that AKIN can do a lot of work to help you figure out if the content you are looking for is in your document even before you hit the search button and/or start to navigate and scan the document text itself. While you are typing AKIN looks at what you are typing and tells you if it found the word you are typing within the document by showing you a dropdown list. If the word was not found in the document, it will not show up in the list. However, if the word is not found, AKIN goes further and looks for similar text found in your document using fuzzy logic and pattern recognition.

"Semantics-Aware" refers to AKIN's new ability to try and understand WHAT YOU MEAN and look for that meaning. It does this by combining its fuzzy search with linguistic knowledge-bases. As you type your searches in the Deep Content Search form, AKIN looks into its Knowledge-Base (currently only English is supported), and looks for the exact word you typed as well as similar words, and if found, looks for words associated with the word found that may also be in your document. For example, if you typed "Surpised" (misspelled on purpose) and that word was not in your document, AKIN might find the related words "Startled" or "Alarmed" in your document if they existed there. Words that were found in the Knowledge-Base but not in your document have a lighter font than words found in your document. However, that's just the start.

Although it is helpful that AKIN is showing you all these other possible words that actually exist in your document, which will help you choose the more relevant search words and phrases, there may be so many of these suggested words that trying out all the possible combinations to find the most relevant set of search terms could take as much time as manually scanning your document. So, AKIN also does this work for you. If you type in more than one word into your search and both of those words or their related words are found in your document within close proximity, AKIN will show you a list of the word combinations or phrases it found within close proximity. We call this whole process "Early-Discovery" because it goes a long way to helping you figure out if the content you are looking for exists in your document, even when you don't know exactly how it might be written, before you actually hit the search button and begin to visually scan the actual paragraph contents of the document itself.

These are new features, so we expect to release many improvements to them over the course of the next several months. They are already very powerful and we only anticipate them getting even better. We invite you to come download the free trial and see what you may have been missing in your productivity suite:

AKIN HyperSearch Deep Content Search and Semantics-Aware Early Discovery

Published on January 2, 2015 by Grapple Data

The first update of 2015 is here and we are bringing current and future users a new tool as part of the AKIN HyperSearch experience. A good number of users have complained about the general lack of useful file directory tools, so we've decided to help fill that need. The new AKIN Directory Analysis Tool (or ADAT) allows a user to compare the directories and files between any two directories in some unique ways not found in most directory compare tools. It also allows users to perform some useful operations between the two directories in a way that makes managing file directories a lot easier. To learn more, click the link AKIN Directory Analysis Tool. We sincerely wish all our users an extremely wonderful new year!

Published on July 31, 2014 by Grapple Data

If you are a power user, or you speak or deal with information in languages other than English, we hope you've had a chance to see how powerful AKIN can be in multi-lingual environments. On our features page, you'll note that while AKIN is only currently certified for the English language, we've added additional support for Chinese, Hiragana/Katakana, Hangul/Korean, Hebrew, Arabic, and various other European languages.

Published on August 8, 2013 by Grapple Data
  • Now Certified for 1,000,000 (1M) + items indexed
  • 90% + reduction in startup time at 500,000 + indexed items
  • 60% + reduction in standing RAM(memory) usage at 500,000 + indexed items
  • 90% + decrease in time to search for word phrases that occur very frequently in large volume indexes
  • at 500,000 + indexed items

    This is one of the most significant updates we've released this year. As most of you know, for the first half of this year since releasing AKIN we've been primarily focused on improving UI and features for general users who don't usually have more than 100,000 items to index on their systems (max 300K); the everyday business user who just wants to find their most important information quickly. The great community support and overwhelming positive feedback we've received to date has really propelled AKIN forward.

    However, this update has shifted focus to optimizing the engine for hard-core power users, as well as preparing it for release as a small/medium size business or department level file server search service for businesses to be released in the near future. Based on the statistics above, we can truly say this release is turbo charged, and makes the "Hyper" in HyperSearch even more emphatic.

    Additionally, we've improved stability and taken care of a few bugs.
    Also included: The free download now includes a 30 day trial without registration with ALL features unlocked and unlimited
    However, you can still register for the free version (which still has limitations) in order to continue using AKIN past the 30 day period if you don't need the PRO version.

    We are all super excited by this released and hope you will be as happy with it as we are.

    Published on April 21, 2013 by Grapple Data

    The tendency for information to flow towards an individual in a particular context, without the individual needing to explicitly or consciously search for that information.

    A system that reacts to the activities and informational context of the user and proactively brings relevant related information to them.

    Published on April 13, 2013 by Grapple Data

    Applied specifically to textual information, Fuzzy and Fault-Tolerant Textual Search and similarity assessment are monikers that have popularly stuck to the concept of a computer being able to assess the similarity of textual information that is non-exact. For example, these two strings mean the same thing but are written or represented differently:

    A. November Payroll Records

    B. Nov Pyrll Recs

    We can debate all we want about whether string B should ever exist in the first place, but the harsh reality is that the world is a place full of informational diversity. Every person is different, has their own way of doing things, and no matter how much we have tried to standardize, these differences have invaded the data and information individuals and organizations have on their computer systems. Even worse than individual difference is the fact that we as individuals also change our perspectives, opinions, habits, and ways of doing things over time so that it becomes much less likely we can faithfully replicate (from memory) our own writing the farther we are from the date of its creation. Three years from now I will not remember what I named any particular document, let alone everything I wrote in it. I will remember bits and pieces or aspects, but not in its entirety, and I won’t remember exactly how I wrote it. Additionally, it’s quite likely I made a few typos along the way.

    Contrary to what some movies have portrayed, computers in and of themselves are not intelligent. They deal in very simple exact bits of data. By default, if you give a computer two strings of data and ask them if they are the same, if even one bit is different they will tell you they are not the same, although to a human being the two strings of text may look completely semantically identical. Thus, it was necessary for human beings to devise both algorithms and heuristics to try and build more intelligence into computers in order that they could more effectively “see” lexical and semantic similarities in textual data. Today some very advanced systems exist within very large companies like Google, Microsoft, IBM, etc. that are much more effective at accomplishing that task.

    Fuzzy and Fault-Tolerant are possibly misnomers for this technology. These might imply that the information itself is somehow fuzzy and faulty.  However, in reality many times the textual information simply varies in completely valid and rational ways, and when a human being looks at it, it makes complete sense to them what the information represents. What can be said to be fuzzy or faulty then is the computer’s own ability to actually see features and characteristics of textual information with a high degree of fidelity. Without this fidelity, the computer is essentially blind and cannot see textual similarities well, let alone determine semantic associations effectively. By default, a computer’s vision is in very low definition so that small differences create very large distortions such that when comparing textual information the computer will automatically decide they are different if even the slightest difference exists. For this reason, I personally don’t think of more intelligent search as fuzzy search, but rather more like clear and high resolution/fidelity search. It is as if we as programmers are giving computers better sets of optics to see the individual aspects and semantics of textual information more clearly, and from there derive both lexical and semantic associations.

    There are some very standard well-known fuzzy algorithms and methods known to the general public, but the really effective and intelligent ones that power the most effective search engines in the world are proprietary, and are what make up the "secret sauce" of many search engines. AKIN is no different. Our algorithms are uniquely tailored to provide computers with the ability to see your information and queries more clearly than ever.

    As a human you can get a sense for how computers see textual information by looking at the following image of two strings of text that start from a very pixelated view of them to a very high definition view of them. Here we represent 8 different levels of definition. Through a normal computers default eyes with no additional intelligence programmed in, these strings look like level 1 to the computer in terms of similarity. There is no way it can determine their similarity and simply states they are different.
    What is fuzzy search

    Over time, various algorithms have been developed to give computers gradually improved clarity. However, in large part this technology is not available to the individual consumer in the form of local desktop information management and discovery. Sadly, it remains in use primarily by large corporations internally to drive their enterprise data. Most of these enterprise systems are operating somewhere between level three and five respectively, whereas local desktop information search in large part continues to remain suck at or between levels one and three. That being said, the systems that have been created to power the big web search engines a la Google, Bing, Yahoo, Ask, etc., out of competitive necessity, and the huge resources thrown at them, have achieved levels more like level five or six. A large part of the reason web search engines haven't yet moved past level 6 is simply the enormous volume and heterogeneity of information they must process. Contrast that with the relatively miniscule amount of information the average person has on their desktop/laptop, and it makes little sense why desktop search is so under-powered.

    The differences between the results a computer can generate at various levels are seen not just in the result items that are actually returned to the user, but specifically where in the results they are presented (top, middle, bottom). An algorithm that just makes a lot of uneducated guesses about similarity will generate results that to the user do not seem sorted in any particular order of actual similarity or relevance. This necessitates scanning far down the list of results to find and make sure we don't miss the desired item(s). Whereas, a computer that can “see” more clearly and discern both lexical and semantic similarity with accuracy may still produce large lists, but users can be confident that the the most similar and relevant items will generally be within the first 5-15 items at the top of the results. This confidence means it is usually unnecessary to scan the entire list of results unless one is curious.

    The algorithmic technology we've created for AKIN HyperSearch not only brings this higher power computer vision to your desktop/laptop computer for your own local personal information, but in some aspects its algorithms push beyond certain limitations of the web search engines into level 7 territory, and every major update we release also brings it sharper vision. Unfortunately, achieving level 8 will be many more years out and also require continuing advances in hardware to power the software in ways that don’t overburden the system resources. Yet, even now, it stands as a giant leap forward in local desktop information search power, and translates directly into improved personal efficiencies, insights, and significant reductions in stress for information centric business professionals and academic researchers.

    Published on April 11, 2013 by Grapple Data

    I'm pleased to announce we've released our first patch for AKIN HyperSearch 2.0 ( rather than just fix bugs, we thought we'd go one further and also add Evernote fuzzy content search to the list of features as requested by several. Enjoy!

    Published on April 9, 2013 by Grapple Data

    Have you ever wished you could search the content of documents in a way that was fuzzy/non-exact, and didn't require you to remember exactly how something was written? Wouldn't it be great if you could do this across Word, Excel, PowerPoint, OneNote, Pdf's, Rtf's, HTML, Outlook Email, Tasks, Notes, and Calendar Items?

    Now you Can! It's our pleasure to announce that AKIN HyperSearch 2.0 has released with a brand new UI (User Interface) and new Fuzzy Content Search with Phrase Proximity Detection.

    Additionally, with this release we have included a completely FREE Edition, which only requires quick registration to receive an activation code.

    We hope you enjoy the new software, and it helps you make your lives significantly easier, more stress free, and enjoyable.

    Published on April 4, 2013 by Grapple Data

    There has been a great disparity between the ever increasing power and effectiveness of internet search engines to find relevant information across the giant World Wide Web, and the relatively static inability of search technology to do the same for the comparatively small cache of local information the general consumer has stored on their desktop and/or laptop hard-drive.

    Fundamental and important deficiencies stubbornly persist in today’s desktop information search software. A good example of this is the fact that for internet search engines, generally, the more information you type in your search query, the more relevant the results are that they provide. However, on the desktop, the situation mostly seems reversed. Often and consistently, the results we get on the desktop discourage us from typing too many words into the search field; because the more words you type in, the more likely it is that you will NOT find ANY results, let alone results that are more relevant.

    There is a persistent intolerance for textual variations. If any of the search terms or character combinations one enters don’t explicitly and exactly exist in the target item’s name or content, even if some or part of it does exist, the target item won’t show up in the results. For example, using the available search tools, I can’t find the document “Comparisons of taxable income.doc” because when I searched for it I only remembered to write “Income Taxes Compared”, because I didn’t remember exactly how the document was named. I also couldn’t find “Revenue projections” because my brain immediately looked for “2013 Projected Revenues”. This especially happens frequently when you are looking for emails or other documents that have been named by other people who don’t think exactly the same way you do.

    If I want to find the document, I must play a tiresome game of trying to guess how little information needs to be entered in order to at least get the item I’m looking for to show up in the results. In this case I might type only “Income Tax”, or just “tax”, and then have to wade through a long list of results that are not sorted in any meaningful way. In the end I might find it, but at the cost of more energy, time, interrupted workflow, and frustration. Intelligent software should be able to do more with more information, not less.

    At best, deficiencies like this discourage efforts to effectively curate and make use of valuable information, and at worst they cause us headaches, stress, delays, broken workflows, and even at times notable losses in income, reputation, and success; even if sometimes we are not always conscious of it.

    Why has there been so little innovation in such an important informational environment? In the end, our local information is where we composite our thoughts, research, plans, and projects. The sum of that stored information represents many of our perspectives at any given point in time, and should form the basis for deep insights and wisdom for the future; if only we could recall it effectively. Why has such an important environment been so badly neglected over the last decade?

    Imagine how different our lives would be if our technology worked harder to automatically and proactively associate and recall information for us in the context of any of our activities or processes, bringing information to us when we needed it, rather than needing to constantly go digging and hunting for it? How much easier would our lives be if a computer could “see” the similarities in textual information in a way that was more similar to how a human being sees?

    To some extent, an evolution is occurring in some mobile devices and applications with regard to smarter data integration, although local search on mobile devices is still quite poor. They must be connected to powerful search providers in order to harness their power. The desktop and/or laptop environment, where most professionals spend a significant portion of their day, still remains a place where various types of important information remain highly disconnected and non-integrated, having no fabric to resolve their heterogeneity. This requires the user to bridge the gaps, which is becoming more difficult as the amount of information we must manage increases exponentially. Although our desktop computers are much more powerful today than ten years ago, modern desktop search software has left this computational power vastly underutilized.

    It is our mission at Grapple Data Technology to effect significant changes in this regard. We are dedicated to creating a new paradigm that dramatically empowers users to more effectively harness the information they curate. Smarter search technology having more relevant results makes it less necessary to have better ways to immediately categorize the information you receive and create. Additionally, powerful new integration fabrics need to make it easier than ever to cluster, organize, composite, recall and curate information in ways that save us time, and tend to produce stronger and more frequent moments of insight.

    Our first product, AKIN HyperSearch, is the first major step in that journey, and a key technological foundation for its success. This product has a number of significant innovations to contemplate and experience, but in a nutshell it is the most powerful and intelligent federated search software for the desktop in existence today. It contains the kind of intelligence only found (and rarely so) in advanced data discovery systems implemented in the world’s largest and most successful fortune 50 companies, and we have now brought it to the average desktop consumer. It solves many of the problems mentioned in the paragraphs above, and more.