KR20170005428A - A method of searching text based on two computer hardware processing properties: indirect memory addressing and ascii encoding - Google Patents

Abstract

A method and process for retrieving and inserting a word or set of words in a large data set for a real-time data-intensive search application using a memory bank is disclosed. Traditional search methods optimize time and space by pre-sorting the data to achieve faster searches. Unfortunately, in a real-time search intensive application, it is almost impossible to take a snapshot of a data set in real time while a transaction is taking place to sort and search for a word or set of words. The method and process is an innovative approach to using an unordered list to retrieve data in real time without pre-sorting the data. The time complexity of the proposed method is very fast. Additionally, the proposed method utilizes indirect addressing in a memory bank to perform insertion and retrieval that reduces code complexity and time.

Description

TECHNICAL FIELD The present invention relates to a text retrieval method based on two computer hardware processing properties of indirect memory addressing and ASCII encoding. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a text retrieval method based on two computer hardware processing properties of indirect memory addressing and ASCII encoding.

The present invention relates to a method and a process for inserting, storing and retrieving a list of words in a high speed memory bank using two computer hardware processing properties - indirect memory addressing and ASCII encoding.

Search is a very basic operation in computing, and is used to find work attributes within a large collection of attributes. For example, a single number may be retrieved within a set of numbers, or a single word may be retrieved within a set of words within the database.

The search method generally needs to be fast because the applications that retrieve the attributes are gated by the answers that are retrieved again through the search. For example, if the search is successful, the logic follows a course of predetermined actions rather than when the search is unsuccessful.

Popular searches include binary searches. However, binary retrieval requires pre-processing of the data, typically sorting. The same method as the previous search is very good when working with complete and stable data. When data (of size N) are ordered, they realize the log (N) rate. However, many data-intensive applications are evolving and it is very difficult to take a snapshot of the data to draw lines and operate them for retrieval purposes.

Some data intensive applications are in areas such as banking, airline and hotel reservations. Banking transactions appear continuously, and searches for specific information appear in parallel. For example, in applications such as credit card companies and universities, new records are always added. It is cumbersome to sort and resort to operate by the search method depending on the sorted data.

Most industries sort data in a predetermined time slot (eg, night) by taking a snapshot so that no interruption occurs during the transaction. Unfortunately, sorted lists are not up-to-date lists.

In contrast, it is difficult to keep the list always sorted, because it takes time to insert the new item into the sorted list. When applications have millions of transactions, it is very difficult to invest the time to keep the information just sorted for optimal retrieval purposes. Different search methods are needed for time, power, and computer memory minimization.

In the present disclosure, novel methods and processes for inserting and retrieving words in a list of words stored in a memory bank located in a processor or computer hardware are described. In one embodiment, a method of inserting words into an index-based array using a memory bank is provided. In another embodiment, a method of retrieving words from the same index-based array is realized, thus eliminating the need to perform sorting operations prior to array retrieval.

In one embodiment, methods and processes for utilizing ASCII encoding and indirect memory access for insertion and retrieval are described. The minimum requirement is to have access to a memory bank of at least 256 bytes in size. In one embodiment, we divide the memory bank into two memory banks, for example we call it memory bank A and memory bank B. In one embodiment, we initialize memory A and B by setting all entries to zero. The inserting method is the memory method when the memory method is used for memory retrieval.

In this insertion method and process, in one embodiment, the first step is realized by initiating the recording of all single characters in memory bank B. Then, the next step is to read the ASCII code of the first character in the memory bank B. [ Adjacent addresses in memory bank A represent all ASCII codes in a particular order. As the second step, the data is written in the memory bank A at the position set by the ASCII code of the first character with the value 1 indicating that the first character is to occur. The ASCII code of the second character is found in the memory bank B again and the value 2 is written to the memory bank A at the memory position set by the ASCII code. The reading of all the characters in the memory bank B and the corresponding position in the memory bank A are written to the address set by the ASCII code.

In a third step, with the insertion method and the process, all two-letter syllables present in the word are inserted into memory banks B and A, respectively. As an example, in the case of an address write operation, in the case of a two letter syllable, we need the number of memory blocks A that can be equal to the number of characters in memory block B. As a further step, a reading of the first two letter syllables is made, and then the determination of the ASCII code of the first character is performed. These ASCII codes determine the memory block number so that the positions of these two letter syllables are located. To obtain the ASCII value of the second character, this value gives the address in the selected memory block. Here, the addresses of two character syllables are recorded in the present memory block.

In a fourth step, all three letter syllables in a number of memory banks are also inserted and recorded. For address recording of these three letter syllables, we need a number of memory blocks equal to the number of all two letter syllables. The first two characters are read, and then the address of the two-letter syllable is obtained as in the preceding paragraph. Using the values read to determine the number of memory banks, the addresses of the three spelling syllables are located. The ASCII value determines the location in the selected memory bank, and the address of the three letter syllable will be placed.

This reading and writing process continues this reading and writing process until all syllables of all sizes have been read and their positions recorded and written to the specified location. In the case of a syllable of length N, the address will be recorded in a memory block equal to the address of the syllable of length N-1. ASCII of the letter N determines the position in the selected memory block so that the address of the syllable having N characters is located.

In one embodiment, a search is performed, and the search is a reverse insertion operation. For example, given an N-letter syllable, we first use the ASCII code in memory block A to arrive at the location in memory block B to obtain the position of the first two-letter syllable. These two position values are used to read the position of the two-character syllable. The position of the three-letter syllable can be obtained by using the position of the two-letter syllable with the position of the third character. Again, with the position of the fourth character, using the position of the three character syllables, we obtain the position of the four-letter syllable. The search continues using an indirect memory addressing and ASCII encoding method until a syllable position of length N is reached. If the value is not zero, this syllable is found and its position is given; otherwise, the value is zero and syllables are not found.

The generic search method uses aligned insertion techniques, and insertion takes time to properly place words. The technology takes time to build applications and data structures that involve a greater number of transactions so as not to be able to provide this delay during the search process.

The global search method searches for a word or a group of words, and assumes that such data can be sorted through sorting to enable fast searching. For example, a binary search method of already sorted or sorted data groups is efficient. However, it takes time to sort the data through sorting. Costly search applications such as search engines can not always be provided to keep the latest sorted data.

The global search method has an array of words or a single sorted data structure using a memory bank when a set of words is to be searched for a particular word. In our proposed method, a plurality of arrays stored in an efficient memory bank are used for the retrieval process. One array through the memory bank is used to store the words to be searched, and a multidimensional array is used to store the positions of all words, syllables, and characters.

In the proposed method, the insertion of a word into a data structure and the retrieval of a word in the data structure are handled. A word is classified as length 1, length 2, or less than the maximum distinguishable value (MaxDist). MaxDist is the longest matching requirement of an already existing search word.

Figure 1 illustrates a high-level architecture of search operations in which one search method is used to search for one attribute among a set of attributes.
Figure 2 shows a generic search method in which the input from an application is first inserted into an unordered data structure. The data structure is pre-sorted before the generic search method operates on the data for property discovery.
Figure 3 shows a proposed fast index based search method in which words are stored in the main array at syllable and character positions based on memory banks.
4 shows a proposed method for charging a main array memory bank.
Figure 5 shows the process of finding the maximum distinguishable length of a word before insertion into the main array.
Figure 6 shows a method of charging the main array after finding the maximum distinguishable length.
FIG. 7 shows a process for inserting words of various sizes.
FIG. 8 shows a process for inserting a single character syllable.
FIG. 9 shows a syllable inserting process with a length greater than one.
Fig. 10 shows the extraction of syllables after the first syllable.
11 shows a final syllable extraction process.
12 shows a data structure for storing various syllables and words.
FIG. 13 shows an example of a process for inserting a single character syllable.
14 shows an example of insertion of a new word into the main array memory bank.
FIG. 15 shows an example of insertion of new words that are not made up of single letters.
Fig. 16 shows an example of a search process of a new word.
17 is a schematic system diagram of a computer device diagram capable of performing the embodiments disclosed herein, in accordance with one embodiment.
18 shows an overall system for processing an insertion and retrieval method, in accordance with one or more embodiments.

The present disclosure relates to methods and processes for quickly performing insertion and retrieval of a word or group of words. In particular, a memory bank is used to efficiently record and assign ASCII values for each character for a word or spelling group or group of words. There is no sorting of data during insertion and retrieval methods and processes.

In the present state of the art, sorting is introduced as a precursor to minimize memory. However, memory has become cheaper in recent years, and such constraints can be mitigated for realizing high-speed and real-time computation.

Data-Intensive Application: The present invention relates to an insertion and retrieval method and process in which a word or a set of words is inserted into a large list of data and retrieved within a large list of data. Examples of such data-intensive applications in which the present methods and processes for insertion and retrieval may be used include banking, hotel management, search engines, university registries, and warehouse inventories.

In the present method and process, a description of a method of fast insertion in an unordered data list is presented with the search being performed in real time quickly. This is primarily a quick and immediate retrieval using the preceding memory constraints of unaligned data.

Processor speed has increased more than a thousand times from 2.4MHz in 1980 to 2.4GHz in 2013. During the same period, main memory grew more than a million times from 4KB in 1980 to 4GB in 2013. In addition, a portable memory space has become available to operate the data abundantly. For example, the simple memory stack today reaches 128 GB. Thus, the method and process utilize memory for speed realization. Traditional search methods optimize time and space by using the old real-time operation. Therefore, it takes more time to work on stable data or to optimize space. However, using the proposed method, we can operate at full speed and in real time by alleviating memory space requirements. The proposed method works well for large database applications such as space-intensive applications such as cloud computing, university records, airlines, search engines and banks.

A typical search engine has 2 million hits a day for tens of thousands of search operations per day. It is practically impossible to pre-sort and maintain data on a daily basis. Periodic snapshots of the sorted data are possible, but not the latest sorted data. Thus, methods such as binary search are difficult and complex to operate for such applications.

We present a search method that uses unsorted data for word search. Briefly, given a list of words to be searched, we create a plurality of string arrays to store the set of words, syllables, and words to be searched. The data structure is constructed using a word array of a plurality of dimensions so that the retrieval is performed in the best case of a single operation, that is, 0 (1) and the worst case of the length of the word, Time.

The method is fast because of the small number of large words in the dictionary. In practice, many of the words in the dictionary are less than 16 characters. As a method, all words over 16 characters can be trimmed to less than 16 characters and still be in a distinct and distinguishable state. Large words are trimmed to the size we call the maximum distinguishable value. Finally, the average word length is about 4.5 characters. Moreover, English Wikipedia has more than 25 trillion words, and the average word size is 5 characters per word. This indicates that our method can realize approximately 0 (5), or can realize an average of 5 operations per search.

Applications with real-time data: Applications such as search engines require real-time data manipulation. This means that there is no "free" time for data to be searched and inserted continuously and the data can be taken off-line and sorted. Such applications require smart search and insertion methods that can be done in a rapid manner. To speed things up, arrays are stored in memory banks in memory hardware for faster access and transmission. Memory banks may be located in the processor, computer hardware, and the cloud.

Applications search requirements: Applications such as search engines typically require a word or word set (sentence) to be retrieved with a large data set of collected data over a period of time. Data-intensive applications typically perform searches and inserts with speed requirements. The application requires a word or word set to be placed in the set of opposing data collections. The traditional mechanism expects to realize this speed by keeping such large data collection sorted, and thus becomes easier each time a search requirement exists. However, if an application is continuously updated and the data changes in real time, it is very difficult to stop the process to keep the data in a sorted state. Therefore, there is a need to find efficient insertion and retrieval methods and processes in performing a function at high speed.

Application insertion requirements: Applications such as search engines not only want to retrieve data, but also want to insert a word or word set being searched. The knowledge base is kept up to date by learning new words being searched. Additionally, other applications, such as banking, require specific functions to search for a word (name) or word set (address), and the insertion function is used to specifically insert new words into the data. Real-time applications require inserting the insertion function quickly into a data structure that can be used for retrieval. In the proposed method, we show that the insertion method is a subset of the search. In order to insert new words, we search and insert them so that future searches of the words can be done quickly. The proposed method utilizes a non-aligned list array to store words stored in a memory bank for fast access and transmission.

Data sorting: The generic sorting technique uses a complex data structure such as a tree for data sorting. To achieve this, we expect full data sets to be a priori available. A sorting technique, such as quick sort, is used to sort the data where the search is made. Although this method can be fast (although it can be as low as 0 (N log N) for a data set consisting of N elements), the expectation is that the data set should be available for a period of time when no transactions are made. In a real-time application, finding such a pause is very difficult due to the unpredictable nature of the transaction time. Therefore, there is a need to find a fast search method that works on an unordered list of data rather than sorting by pre-condition.

A comprehensive search method using sorted data: A comprehensive search technique uses a sorting process on the data structure to sort the data. Once the data is sorted, the search is easy and quick. For a data set with N elements, the search can be realized at 0 (log N) after sorting the data using at least 0 (N log N). The problem with sorted data is that the transaction must be stopped to complete the sorting mechanism separately when the data needs to be stable. If the transaction continues and a copy of the data is sorted, the retrieval of the sorted data is done using stable data. The proposed method does not require separate sorting, and the data is composed using an unordered list. A fast search without pre-sorting is an indexing-based array search, and the data is stored in a memory bank for fast access. This process is used for insertion to induce dual use.

Complexity of the global search method: It is as low as 0 (log N) to realize the search for sorted data in relation to the data set of N elements. However, this does not take complexity into account for data a priori. The proposed method is very fast and has the worst case of 0 (1) and 0 (word length) for unaligned data. It is important to note that English has a larger number of words with fewer characters.

Data storage - Data structures: The data structures used to store data are typically based on abstract data types. However, using a data type, such as a tree for pre-storing, requires a snapshot of the data that real-time applications can not provide. The proposed method utilizes an unordered list of data types, implemented using an array structure for convenience and speed.

Data storage - Index-based arrays: The index-based arrays used in the proposed method provide the flexibility to address certain steps in a single step and keep them logically continuous. Due to the a priori declaration of size, the high utilization of memory in the array is a threat. However, as previously described, memory is becoming cheaper and can be compromised for speed and accuracy. In addition, we propose the use of high-speed memory banks for data operation and index-based array storage to achieve the transfer rate.

Time-to-space complexity: The comprehensive search method has optimized both time and space, while trying to work offline as needed. However, over the past decade, such offline computations, such as sorting, were not possible when real-time data-intensive search applications were fairly sensitive to response time. Thus, the time and space optimization of the search and sorting algorithms alone was not sufficient. Optimization of algorithm speed and response is important. In our approach, we introduce a fast search mechanism that provides instantaneous response without off-line sorting requirements.

1 is a diagram showing a high-level structure of a search operation. The search operation consists of a search method 102, an attribute 104 that needs to be searched, and a predetermined attribute array 106 that is linked in a special way. For example, in the present application, we introduce a new search method that efficiently retrieves one attribute 104 or attribute set from a set of already learned attributes 106 that are arranged in an unordered list of arrays.

2 is an illustration of a generic search method corresponding to the case of an application 202 contributing to a word set that is inserted as part of an unaligned data structure for a data foundation build. It goes through another sorting stage 206 for readying the search. Whenever a word or word set needs to be searched, the global search method 210, such as a binary search, operates on the sorted data 208 to find out if a word or word set is present. Thus, two stages of processing are involved, the first being to insert words, the second is to sort the completed data structure to keep it up-to-date and to realize fast searching for the sorted data. Unfortunately, the two stages take more computation time and it is difficult or impossible to keep the sorting data up-to-date for real-time input applications, such as search engines or stock market data.

3 is a diagram of a proposed fast index based search method. This is done by using a memory bank for data storage and operation on the data. The application input 202 inputs a word or word list. Our proposed method provides both an insert and search function 304 by inserting a new word into an unordered index-based data structure array 306. The main array 306 consists of a set of words to be searched. Upon insertion, a new word is placed in the Main array memory bank at location Z. The final character of this new word is placed in the One Character Syllable at position Y. The remaining syllables are located in a syllable store at location X. The value Z is then placed in the multidimensional memory bank array FindWordLodcation 308 at a location indexed by X and Y. [

When searching for the last character of a word that needs to be searched, the first syllable is searched for (found at position X), and then the remaining syllables are searched in syllable store (at position Y). X and Y are used as indices in the FindWordLocation memory bank 308 to obtain the position (or index Z) of the word in the main array memory bank.

Proposed retrieval method with unaligned data: The fast response of this method is realized by using an unaligned data structure, and a data set is generated as new words are inserted, which is helpful for subsequent retrieval. In addition, the data is stored in a high-speed memory bank for quick access. Fast response is also realized by using common functions for both searching and inserting. The unaligned data consists of just three sets of word arrays 306 - a main array memory bank for word storage, one character syllables for storing single character occurrences, and a syllable store for syllables storage. These three arrays are manipulated to store all the words and retrieve them quickly using the proposed method.

Optimization of the search process: The proposed method optimizes the search process by knowing the maximum distinguishable length of the words 404 in the data set. Accordingly, the search process limits the search to the maximum distinguishable length of the total length of the words to be searched. All the words that need to be inserted are filled into the main array memory bank, which is the word array 306. [

Figure 4 shows the proposed method of charging (406) the main array memory bank with the words to be searched. Charging the array with full words and searching the database increases the search time for long words. Thus, we find 404 the maximum distinguishable length of a word-that is, the length of a subset of words that can be distinguished to trim a word.

FIG. 5 illustrates a process for finding the maximum distinguishable length of a word before inserting 406 into the main array memory bank. The idea here is to find the length of every long word that is distinct when trimming to this size. No other word matches these trimmed words. The output of the next process is all MaxDist lengths and is an array of individual, trimmed words. The steps in this process are as follows:

Initialization of the first MaxDist value: Obtain an array of new words to be searched and initialize all arrays (502). The next new word is successively read from the list of new words (504). If the word length is greater than or equal to the observed maximum distance, i.e., if the word length is far (506), a new word is inserted (508). If the word is found to be already there (510), the maximum distance is incremented and the process begins all over again and proceeds with the insertion again. Otherwise (516), proceed to the next word (514). If all the words are checked, the final maximum distance found will be the maximum distinguishable length (518).

Figure 6 illustrates a method of charging a main array memory bank after finding the maximum distinguishable length. This method reads the word to be searched 604 and charges the main array memory bank until all the words are inserted 608 using the insertion process 606. [

FIG. 7 shows an inserting process 606 for a new word. If the word is length 1, i. E., A short character word (702), the word is inserted 708 into the single character syllable data structure. If the word is a word of length greater than one, i. E. The word is not a short character word 704, then the word is inserted 710 into the data structure for the plural character syllables. If the word is greater than the maximum distinguishable length, it is inserted 706 after parsing with prefix and postfix and after trimming with MaxDist size.

The method used for new word insertion is new word insertion (606). The new word to be inserted is checked whether it is a short character word 702 or a longer word 704. New word search follows the same process as new word insertion. If a location for a new word insertion already exists, it will not be empty, leading to a positive search.

Figure 8 is a sequence of steps for inserting (708) a single letter word, i.e., a single letter syllable word. First, the position of the character is checked in the array (802). If the position flag is empty 804, the position flag is incremented 810 and the data structure is updated 812 to reflect the learning of the new short character word following the update of the main array memory bank for insertion reflection Updated data structure 812, and an update 814 of the main array memory bank to reflect subsequent insertion.

FIG. 9 illustrates the steps taken to insert 702 a word of length greater than one. The initial step reads the first character of the word (902) and checks if the first character is in the single character data structure (804). If not, a similar process is followed (810, 812) with the insertion of a single character. Then, when the word length is 2 or more, the remaining syllables are extracted (912). If the word length is 2, the remaining syllables are also single letters, in which case they are extracted as a separate process (914).

FIG. 10 shows the extraction process 912 of the remaining syllables. The first character of the remaining syllable is read (1002), the short character position flag is checked (804), the position is incremented (810), and the data structure is updated (812). If the first character position flag is not empty, the remaining syllables are configured in character units, and ultimately the final character is extracted (914).

Fig. 11 shows a final character extraction process 914. Fig. If the character is not located, follow steps similar to inserting a single character (804, 810, 812). Once the word position is determined and a word is found 1112, the main array memory bank is updated 1114 and the word position is returned to the new word insertion process 1116 to indicate that the word was found. If no word is found, the new word is inserted and the word position is returned.

Figure 12 shows a different string data structure array used as part of the search and insert process. The main array memory bank is configured with new words (1202).

The OneChSylbl array consists of a single character 1204 and the syllable store 1206 comprises the rest of the syllables. In addition, three integer arrays are used - OneChSylblLoc to flag the position of the single character syllable, SylblAdd to provide the address of the syllable, and a FindwordLocation memory bank 308 to flag the word position in the main array memory bank. SylblAdd < / RTI >

Figure 13 shows an example of a single character insertion process. OneChSylbl is an array of already learned intrinsic characters. For example, when four words, I, AM, A, and MAN, are inserted, the OneChsblbl array will have the letters I, A, M, and N that are distinct (1302). Likewise, the integer array OneChSylblLoc shows the number of times characters have ever appeared. This character is traced for each ASCII value (1304).

14 shows an example of inserting a new word into a part of a string array called the main array 1402. Fig. It is clear that I, AM, A, and MAN occupy the first four indices, and the zeroth index shows the next available free index.

A single character word is added to the single character syllable data structure (812). The position of the character in the data structure is captured at the one character syllable position 812. This position is flagged (1304) in an array of ASCII characters that determines how many of the related characters have been recorded. Syllables are stored in an array of characters (1302). The main array memory bank is also updated with words (814, 1402).

A word longer than 1 is divided into a first character and a remaining syllable. The first character is treated as a single character occurrence (812), and the remaining syllables are treated as character units (1002). Syllables are integrated and added to the data structure syllable store, and the position of the syllable is tracked in the syllable position data structure (1010).

Fig. 15 shows an example of insertion of a new word not a single character. SylblStore is a string array memory bank 1502 that stores syllables. For example, to insert "MAN", the syllable "MA" is stored in the SylblStore memory bank at the [1, 2] position, because it is the first met word with two characters. When inserted, position [0, 2] is incremented to 2 to show the next available free space for inserting the next two letter syllables. Similarly, integer array SylblAdd 1504 flags the location of [3, 2] for SylblAdd (2, position "M", position "A"), which means that "M" occupies the third position in the OneChsylbl array , "A" occupies the second position in the OneChsblbl array. If the word is "MANY" instead of "MAN ", then SylblStore will have positions of" MA "

Figure 16 shows a search for a new word. For example, if the word "MAN" is to be found using FindWordLocation (position 3, "MA "," N "), the position of MA is 1 and sylblAdd "). ≪ / RTI > The position of "N" is 4, and it is determined from OneChSylblLoc (ASCII (N)). Thus, the FindwordLocation memory bank returns the location of the word in the main array memory bank. If the index is 0, the word does not exist. Thus, it can be seen that the retrieval is extremely fast and the retrieval can be completed in a couple of steps by simply manipulating the array memory bank. This is considerably easier than the comprehension method that requires string sorting before searching.

The last character is extracted and a Find Word Location table is constructed which tracks the word. Find Word Location constitutes a set of tables for words of different lengths. The word position table 1602 shows that the positions of the three spelling words being searched, i.e., 4, are in the main array memory bank at the fourth position. If this entry is 0 rather than 4, it means that the retrieved word does not exist. At this point, the word may be inserted using the same procedure used for the search while updating the main array memory bank.

For example, "MAN" is grouped into "MA" in syllable storage data structure under two letter syllables 1502 followed by the final letter "N ". The positions of "M" and "A" are tracked in syllable address table 1504. The word "MAN " is tracked in a word locating process that clearly shows the location of the full word in the main array memory bank 1602. [ When we take "MANY" as an example, the positions of the syllables "MA" and "N" will be tracked at the syllable address with the last single word "Y". The word "MANY" will be found using the positions of "MAN " and" Y "

Figure 17 is a schematic diagram of an integrated system for data management 1700 that illustrates communication between a user and a server over a network, in accordance with one embodiment. In one embodiment, one user or a plurality of users may be connected to a server hosting a multimedia tool in the system. In another embodiment, user hardware such as a PDA, a mobile device, such as a tablet, etc., a computer or mobile phone or any wireless device, or an electronic book (e-book) , The user can use multimedia tools for training, learning, and / or interactive play games. Network 1701 may be a LAN, WAN, mobile, radiocommunication, Internet, intranet, WiFi, and / or ZigBee network, The user / individual 1705, 1704, 1703, the database 1702 for storing all information, etc. may be an individual, a parent, a scientist, or a writer, but is not limited to such a fork group. Users and individuals are used interchangeably and mean the same. The user can be any person who has access to the data management system for various activities as mentioned in the scenario of the different case in the figures. A cloud server may also be used for data storage and processing. Access data management tools for searching, creating content, uploading content, viewing content, using content, and / or deleting content. The server may be a stand-alone, cloud-based, or hosted service.

FIG. 18 is a schematic system diagram 1800 of a computer device diagram capable of performing the embodiments discussed herein, in accordance with one embodiment. In particular, computer system diagram 1800 includes a processor 1802, a main memory 1804, a static memory 1806, a bus 1812, a video display 1820, a numeric character input device 1822, A cursor control device 1824, a driver unit 1826, a signal generating device 1828, a network interface device 1808, a machine readable medium 1830, an instruction 1832, and a network 1701.

The computer system diagram 1800 may represent a personal computer and / or a data processing system (e.g., a server) that performs one or more of the operations described herein. The processor 1802 may be a microprocessor, state machine, application specific integrated circuit, field programmable gate array, or the like. Main memory 1804 may be a dynamic random access memory and / or main memory of a computer system. The static memory 1806 may be a hard drive, flash drive, and / or other memory information associated with the computer system. The bus 1812 may be an interconnect between various circuits and / or structures of a computer system. The video display 1820 may provide a graphical representation of the information for the data processing system. The numeric character input device 1822 may be a keypad, a keyboard, and / or any other text input device (e.g., a dedicated device for facilitating physical impairment). The cursor control device 1824 may be a pointing device such as a mouse.

Drive unit 1826 may be a hard drive, storage system, and / or other longer-term storage subsystem. The signal generating device 1828 may be a functional operating system and / or a bios of the data processing system. The network interface device 1808 may perform interface functions such as code conversion, protocol conversion, and / or buffering, which are required for communication into and out of the network 1801. The machine-readable medium 1830 can provide instructions that can perform one of the methods described herein. The instructions 1832 may provide the source code and / or data code to the processor 1802 to implement one or more of the operations described herein.

Using the present systems, methods, and processes, accurate information can be intelligently and securely updated, maintained, and dynamically reconnected between the database and the transport channel at the right time. Limitations and rules may be implemented to suit any user / user organization. Accordingly, the specification and figures are to be regarded in an illustrative rather than a restrictive sense. The invention is valid for a general purpose application that requires the insertion and retrieval of a single word or set of words.

Claims (16)

Receiving input from a user in word form;
Sequentially storing all the distinct characters in the word in order of occurrence in the hardware memory with the memory bank B,
Reassembling the location of the character in memory bank B to the location specified by the store of memory bank A and the ASCII code of the character,
Retrieving the characters sequentially in the word from the memory bank B by translating the characters into corresponding ASCII codes and retrieving the location of the characters from the hardware memory to the memory bank B using the corresponding ASCII code;
Storing an order of occurrence of characters in the word in an ASCII code corresponding to the address in memory bank A; The method according to claim 1,
Wherein the memory banks A and B are at least 256 bytes in size. The method according to claim 1,
Wherein UNICODE is used instead of ASCII code, and memory bank A is at least 16 KB and memory bank B is less than 16 KB. The method according to claim 1,
The storage of all the distinct first two or more character syllables of the input word is stored in a memory bank of a size corresponding to a multiple of memory bank B and the location of all two or more distinct character syllables is located in the memory location of memory bank A , And the number of memory banks of size A required for location storage for a first starting syllength of size N is equal to the number of distinct syllables of size (N-1). 5. The method of claim 4,
Wherein the storing of all the separate input words of two or more characters is stored in a memory bank of a size corresponding to a multiple of the memory banks B. 6. The method of claim 5,
Locating an input word of size N for an in-memory location is at a location determined by the position of the start syllable of size N-1 for that word and the position of the last character of the word. The method according to claim 6,
Retrieving a word of size N that requires a location to be stored in memory. 5. The method of claim 4,
Further comprising optimizing the time for insertion of an input word and its corresponding location based on an unaligned list stored in a plurality of memory banks of size A and B. The method according to claim 1,
Wherein the memory bank can be located in a database or any abstract data type or data structure. The method according to claim 1,
A common method that can be used for both fast access and insertion and retrieval to reduce code complexity is due to multiple memory banks of size A and B and indirect memory addressing. The method according to claim 1,
Further comprising optimizing time for a real-time search using an unordered list to generate a data set as it is searched and inserted. The method according to claim 1,
Wherein the retrieval and insertion of very large words in the order of occurrence of the letters in the word with the ASCII code corresponding to the addresses in the memory bank A and the memory bank B are made using the MaxDist operation. The method according to claim 1,
Further comprising performing a search based on an index-based array such that data is stored in memory bank A for fast access. Receiving input from a user in word form;
Sequentially storing words in the order of occurrence in the data structure with the memory bank B,
Sequentially accessing the characters in the word from the memory bank B,
Optimizing the time for insertion based on an unaligned list stored in a multiple of memory banks B,
Inserting a word or set of words that does not require taking an snapshot of the data and the order by interrupting the insert application or the search application so that the search complexity can be optimized. 15. The method of claim 14,
Converting the character into a corresponding ASCII code for insertion into memory memory A into the hardware memory,
Storing a generation order of characters in a word in an ASCII code corresponding to an address form in the memory bank A;
Retrieving a related word by matching the order of occurrence of characters in the word based on the address located in the memory bank A without a priori sorting;
Further comprising optimizing the time for a real-time search using an unaligned list such that the data set is generated as it is searched and inserted. 16. The method of claim 15,
In the memory bank A and the memory bank B
Wherein the step of retrieving and sorting the occurrence order of the characters in the word with the corresponding ASCII code in the address form is performed using the MaxDist operation.

Images