Other Database Systems And TIE
Other Database Systems Compared to SpeedTrack's Technology for Information Engineering(TIE)
The TIE system differs in almost all respects from other current systems. We list here the most important differences.
It is important at the outset to realize that TIE always responds to each query with two results:
• First is the list of all matching items (records, joins of records, or documents),and
• Second is the resulting adjusted set of lists of available selectors (field values, words, or search terms).
The first is the least taxing on resources, but the second is considerably more taxing and so required special optimized methods to make the response fast. This second response is critical for all users because it allows true navigation through the available information. Therefore comparisons of the TIE and RDB response times to defined queries are meaningless. What may be meaningfully compared is the time and effort it takes to discover specific targeted information using the two systems
TIE Custom Metadata
The TIE system extracts all the information from the data and stores this in associative matrices as its custom metadata, independently of the location, manner of storage, and structure of the original data. The original data can be stored in any manner, in any structures, or totally unstructured, locally or distributed through multiple locations. TIE does not rely on any databases to store its metadata of associative matrices. The metadata contains all the information needed to perform any search, data mining, and analytics. The original data does not need to be accessed during searches
1. Search: Current RDB text search systems allow a search for anything at all, whether it is present in the data or not. The TIE system allows searches for only the available data and guides you to it. This means that you will never get the "nothing found" result.
2. Search Guidance: Users appreciate guidance in formulating queries.When an initial query matches too many items, I want to know my choices in narrowing the search. Current text search systems do not provide this.The TIE system presents the user with an automatically adjusting vocabulary of terms to choose from, in each face of the data, thus guiding the user to the formulation of queries, narrowing the search through point -and-click, or by typing. This feature is not practical to achieve in current RDB systems without the TIE technology.
3. Boolean Search Training: Current RDB systems require special knowledge and training of the user who formulates Boolean queries, In most cases, they require the intervention of specialists between those who need answers to questions and those executing the queries. Such intervention precludes the possibility of browsing and discovery by those who need to determine what questions would be worth asking. In many cases what questions to ask is not known. The TIE system lets anyone, after just a few minutes of demonstration, search use queries and perform analyses with just point-and-click which would require very complex Booleans in current RBD systems.
4. Text Navigation: In TIE and its user interface GIA, text search through narrative text in records or in documents is replaced by the much more powerful guided text navigation. This guides you to the available, dynamically adjusted, vocabulary of words and lets you narrow by choosing combinations of words (with optional alternatives) which must be present together in a sentence in any order. This method solves in a natural, intuitive way, most common search problems such as: not knowing which words or word forms were used; which misspellings were present; how to narrow too many matches because of common words; etc.
5. Merging RDBs: The TIE system does not rely on data storage structure. It can be used to manage both structured and unstructured data, both separately and in combination. Current systems rely on a very structured storage system, that structure being dependent on the nature of the data. As a result, under current RDB systems, it is very time consuming and hence expensive, to merge different databases, even those with the same kind of data. Under the TIE system, the principal task in merging databases is the merging of their vocabularies, a very much easier and shorter task, resulting in great cost savings and a very much more satisfying result.
6. Viewing the Data: Current RDB systems do not present the user with any view of the information within the data they contain. The TIE system, by its very nature, lets users "see" the information through the displayed vocabulary and its associated item counts. Very useful practical consequences result from this. For example, almost all databases have bad data, but current databases, without special applications, give you no idea of the kind or the extent of this bad data. The TIE system lets you see immediately, bad data such as-historical event dates going through the year 2020. It also shows you the Item count of all bad (and good) years. Another look and you see how many Items have no data in certain crucial fields, or how many entries match none of the meaningful values of a field. Just one click on any of these bad terms lists all the Items and so records, that contain the bad data.
7. Analyzing the Data: GIA, by its very nature, presents analysis of results in response to each user action. GIA displays all terms, or field values (selectors) associated with all matching items and the frequency of items associated with each selector. This provides a view of the associations between selectors. Those with the highest frequencies are most highly associated with each other. The association measures can be shown in a column. Just one click makes available to the user thousands of association measures, sorted from highest to lowest association. Columns can be displayed showing calculated and aggregated values associated with each selector. The result is a complete dynamic report containing the results of all useful analyses. Because it is dynamic, every new click on a selector adjusts the report appropriately. Any part of a report can be copied to a spread-sheet or a document for printing.
1. Storage Structures: Current RDB systems rely on storage structures to store relational information. Therefore changing relational information means re-structuring the data storage, which is almost always very time consuming, therefore expensive, and prone to error. The TIE system stores relationships (associations) between data elements in a set of universal matrices optimized for quick searches and analyses and enabled to quickly adjust the available vocabulary after each user choice. When the user sees the associations between the displayed selectors (i.e. field values, or words in a sentence, etc.) they see the true information in the data, not just the data itself.
2. The Data Cube Replaced: Current RDB systems require special procedures (called building the data cube) and applications for On Line Analytical Processing (OLAP) in order to be able to analyze data using a number of fields or dimensions. TIE is usually built to give a user this ability as a standard and with just a few mouse clicks. In the terminology of the data cube, TIE gives users the ability to see slices through the cube at will through simple mouse clicks, using any number of dimensions. The only limit on the number of dimensions is the practicality of presenting them all to the user in an uncluttered GUI.
3. Calculating Alternatives: Current RDB systems have no simple way to calculate and show users valid alternatives for fruitful choices during a search, or at the start of a search. TIE is built to do just that. This is enabled through a special matrix-like data structure holding all associations between Items, entities and selectors (field values, words,sentences, etc.).
The Developed Technology
TIE comprises several unique technologies, implemented in a query server and a client. In addition TIE supports several data servers for getting various kinds of data with associated display clients. The following are the two most important methodologies of TIE:
1. A methodology of storing, as metadata, the relationships between information elements in a data structure, optimized for very rapid, real-time responses to very complex queries-queries that are impractical to pose and even if posed, impractical to wait for a response under current RDB systems. This is implemented in software as a server, called the TIE Server. All searches are based on this metadata and so do not need to access the real data until a user requests it.
2. Several types of user interfaces, implemented as Guided Information Access (GIA) which allow very intuitive real-time complex queries by users with the minimum of training. One implementation of this is a Java application. It lets users perform a generalization of what is currently called faceted navigation through all the data, even narrative text data (text navigation). Text navigation is much more useful than ordinary text search, making text search unnecessary.
3. Web Page Embedded Client. A new client-server system using a minimal browser-driven client, is in development, with the first alpha versions being tested. This will allow the use of GIA on any desktop, laptop, or mobile device.
4. All these advantages form the heart of TIE.
TIE is covered under US patents 5,544,360; 6,826,566;7,236,972; 8,122,023. Other patents pending.