JPH0399366A - English word search device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to an English word search device.

[Prior Art] When searching for an English word in a dictionary, the dictionary is normally searched sequentially starting from the first character of the input English word and checking for matches and mismatches. If there are a large number of English words in the dictionary, it will take a lot of time to go through them one by one from the beginning of the dictionary. ) are prepared.

Using this table, you can grasp the dictionary position of the English word you want to search for, and then search for each word in detail. The size and structure of this table changes to reduce the dictionary used and search time.

If the input English word is misspelled or does not exist in the dictionary, the search result is the English word that has the most characters that match the headword of the English word in the dictionary.

[Problem to be solved by the invention] If the input English word is not slow in spelling, and the English word is not in the dictionary, the only option is to use the closest English word in the dictionary as the search result. do not have.

However, even if the English word you are trying to input is in the dictionary, it is difficult to search for the correct English word if it is spelled incorrectly. If a character near the end of an English word is entered incorrectly, searching for the English word that has the most characters that match the headword of the English word in the dictionary may accidentally yield overwhelming results. be. However, if there is an error in a character near the beginning of an English word, the search itself in the table will not go well, so it is almost impossible to correct spelling errors when inputting an English word by search. In particular, misspellings when inputting English words are completely coincidental and there are no rules, so it is not possible to correct them based on the way they are input.

In principle, you can find the correct word by comparing the headword in the dictionary with the English word you entered, replacing the missing letters with other letters, and searching sequentially.

However, this method requires searching for all alphabets, which takes a very long time. In some cases, it may be more convenient for the user to perform a somewhat general search and then look at the contents of the dictionary and search backwards and forwards to find it.

The English word search device of the present invention solves the above-mentioned problems, and its purpose is to provide an English word search device in which even if an incorrectly spelled English word is input, the correct English word can be searched from the dictionary by searching the dictionary contents. An object of the present invention is to provide a word search device.

[Means for Solving the Problems] (1) The English word search device of the present invention includes a search item input means for inputting an English word to be searched, a dictionary containing English words, their meanings, usage examples, etc., and the above-mentioned search. In an English word search device comprising a search means for searching the dictionary for the position of an English word input by the item input means and a display means for displaying the search results of the search means, the input English word may contain a misspelling, etc. If the search cannot be performed correctly, try again by replacing the characters that do not match the entered English word with the headword of the English word in the dictionary with the character that has a high probability of occurring next to the character before the unmatched character. It is characterized by searching.

(2) The English word search device is characterized in that the dictionary contains data about the character that follows a certain character and its probability.

(3) The English word search device is characterized in that data about the character that follows a certain character in the dictionary and its probability is transferred from the dictionary to the RAM and stored therein before the search.

[Operation] If you are searching for an English word and encounter a character that does not match a dictionary entry, replace that character with another character and perform the search.

As for the character to be replaced, pay attention to the character immediately before the character that did not match, and consider the character that should come after that character. It is possible to find out probabilistically what letter comes after a certain letter in a word. This can be determined by taking statistics from the arrangement of each character for all the words in the dictionary you are trying to search.

By doing this, you can determine which character comes after a certain character.
Candidates for upcoming characters can be arranged in descending order of probability. The candidate characters thus created are replaced with the unmatched characters and the search is repeated to obtain the correct English word from the dictionary.

[Example] Hereinafter, the English word search device of the present invention will be described in detail based on an example.

FIG. 1 is a diagram clearly showing the basic configuration of the English word search device of the present invention.

The English word to be searched for is inputted and encoded through the search item input means 1 using a keyboard or the like.

For the input and encoded English word, a search means 2 searches a dictionary 3 for a position on the dictionary containing the target English word and its meaning. Based on the located position, the display means 4 displays headwords, meanings, usage examples, and the status of each means.

FIG. 2 is a block diagram showing an example of the circuit configuration of the English word search device of the present invention.

All controls are managed by the CPU 5. The display is performed by a display 6 (CRT or liquid crystal display board).

A keyboard 7 is used as the human power device. The contents of the keyboard 7 are read by the keyboard control unit 8 and then
Sent to PU5. In addition to the input device, a mouse, OCR, etc. can also be used.

The dictionary 9 uses various types of devices depending on the content capacity, search speed, etc. For example, if the capacity is small, a ROM is used, and if the capacity is large, a CD-ROM is used. The dictionary control unit 10 manages access to each dictionary, and arbitrary data can be retrieved by commands from the CPU 5.

Various programs including search are available in program ROM11.
It is stored in Further, a table for improving search efficiency is provided in the search data table ROM 12. In addition, the search data table ROM 12 also stores probabilistic data regarding what character will come next after a certain character.

If the dictionary 9 is a ROM or the like and has a limited capacity, a dedicated ROM is prepared. Data for various processing is stored in RAM.
13 temporarily memorized.

WI about what kind of character comes after a certain character
The rate data differs depending on the dictionary used, so Dictionary 9 is R.
If the data is fixed as in the case of OM, the data may be stored in the search data table ROM 12. However, if the dictionary 9 is intended to be replaced with a CD-ROM, there is no need to replace the search data table ROM 12 if the data is stored in the CD-ROM.

Further, after replacing the CD-ROM or the like, before performing a search, probabilistic data regarding what character will come next after a certain character is transferred to the RAM 13. By doing so, there is no need to search for data from the CD-ROM one by one during a search, and since a RAM with a low access speed is used, the search can be made faster.

FIG. 3 is a diagram explaining a basic search method for English words.

Figure 3 shows the search English word 20 (apple) in the dictionary 25.
This shows the procedure for finding the location.

Two tables 21 and 22 are used during the search.

First, the first character of "rapple" is "a", so "a"
If you look for an English word that starts with ", you'll find it in Table 21.
This is region 23. Note that the numbers in each area indicate the position within the dictionary.

Next, focusing on the second character "p" in rapple4, the position of the word starting with 'apl' is investigated. Use table 22 to find the desired value in area 24. Only after the search range is limited in this way is the headword in the dictionary 25 sequentially searched.

Area 26 corresponding to the value r1029J shown in area 24
The search area 27 is searched one by one starting from the English word 20 (apple) to reach an area 28 containing a word that matches the searched English word 20 (apple). The numerical value "1061" in the area 28 has a meaning, an example, etc. stored in the position indicated by the numerical value.

In the current example, two tables are used, but if the dictionary has a large capacity and the search range is wide, it is necessary to increase the number of tables and reduce the search range as much as possible.

FIG. 4 is a diagram showing how to create a new search candidate English word when the input English word is misspelled.

For example, when the input search English word 30 is 'all)pie', if the search continues as it is, there will be no match at the fourth character 'i'.

So, let's focus on the letter "p" that comes before "i".

Suppose that the statistical results of what kind of letter often comes after "p" are as follows.

Probability that "e" comes next after "p"... N1% Probability that "a" comes next after pJ... N2% rl comes after "p"
Probability that J will come... N3% Probability that "m" will come next to "p" ... N4% Probability that "o" will come next to "p" -
・Assume that N5% and arranged in descending order of probability are as follows.

NI>N3>N2>Na>Ns Search candidate English words 31 to 35 are created in accordance with the order of this inequality. The search candidate English word 31 is the candidate with the highest probability, and the search candidate English word 35 is the candidate with the lowest probability. The search is performed by comparing the search candidate English words 31 to 35 with the dictionary 36 in descending order of probability.

In this way, only words with relatively high probabilities are selected based on the statistically investigated appearance probabilities, so it is possible to significantly narrow down the search candidate English words.

FIG. 5 is a flowchart showing an example of software for re-searching using a newly created search candidate English word when the input English word is misspelled.

Before the first search, refer to the table and check whether the input English word can be searched in the table (step 40). If it cannot be searched (step 41), create a new search candidate English word and search again.

First, search in the table and find the 1st character that does not match.
Find statistical data about the character that comes after the previous character. Data is search data table ROM12 or R
Reference from AM13. (Step 42) Search candidate English words are created based on the data. (Step 43
) This process is as explained in FIG.

Using the newly created search candidate English words, the search is performed again starting with English words with high probability (step 44), and it is checked whether new English words can be searched in the table. If you can't find it (step 45)
) Next, it is checked whether there is a search candidate English word with a low probability (step 46), and if there is a search candidate English word, the search is performed again; otherwise, the process returns to the normal dictionary search. If a new English word is found in the table, move on to dictionary search.

In the dictionary search (step 47), the degree of matching with the headword is actually checked in the dictionary to find at least the word closest to the manually created English word. and display the results. (Step 48) As described above, the correct search English word is predicted based on the input English word and the data on the probability of what character will appear next to a certain character, and the search is performed again.

[Effects of the Invention] As described above, in the English word search device of the present invention, when there is a spelling error in the input English word, data on the probability of what character will appear next to a certain character can be collected. It is possible to predict the correct English word to search for, perform a re-search, and find the correct English word from the dictionary. Since newly created search English words are created based on the appearance probabilities of characters, it is possible to narrow down the search English word candidates and shorten the search time.

In addition, by preparing and storing data for each dictionary about the probability of what character will appear next to a certain character, and storing it in advance by moving it to RAM before use, it is possible to improve search accuracy. High-speed search becomes possible.

Since the prediction of the correct English word to be searched is performed based only on the data that the English word search device has uniquely, it is possible to perform a correct search regardless of the influence of the input method.

[Brief explanation of drawings]

FIG. 1 is a diagram clearly showing the basic configuration of the English word search device of the present invention, FIG. 2 is a block diagram showing an example of the circuit configuration of the English word search device of the present invention, and FIG. Figure 4 is a diagram explaining the basic search method. Figure 4 is a diagram showing how to create a new search candidate English word when the input English word is misspelled. Figure 5 is a diagram showing how to create a new search candidate English word when the input English word is misspelled. 12 is a flowchart showing an example of software that performs a re-search using a newly created search candidate English word when the spelling is incorrect. 1 ・ 2 ・ 3 ・ 4 ・ 5 ・ 6 ・ 7 ・ 8 ・ 9 ・ 0 1 2 3 ・Search item input means ・Search means ・Dictionary ・Display means ・CPU ・Display ・Keyboard ・Keyboard control unit ・Dictionary ・Dictionary Control unit/Program ROM/Search data table ROM/RAM 20...Search English words 21.22...Tables 23.24, 26, 27.28 25...Dictionary 30...Search English words 31, 32, 33, 34.35 36... Dictionary/area/search candidate English words or more

Claims (3)

[Claims] (1) A search item input means for inputting an English word to be searched, a dictionary containing English words and their meanings and examples, and a search means for searching the dictionary for the position of the English word input by the search item input means. and a display means for displaying the search results of the search means,
If the entered English word has a spelling mistake and cannot be searched correctly, the character that does not match the entered English word and the headword of the English word in the dictionary will be placed next to the character before the non-matching character. An English word search device that performs a re-search by replacing characters with characters that have a high probability of appearing. (2) The English word search device according to claim 1, wherein the dictionary contains data about the character that follows a certain character and its probability. (3) In the English word search device, data regarding the character that follows a certain character in the dictionary and its probability is transferred from the dictionary to the RAM and stored therein before the search. English word search device.