JPH02230370A - System and device for morpheme analysis - Google Patents
System and device for morpheme analysisInfo
- Publication number
- JPH02230370A JPH02230370A JP1051114A JP5111489A JPH02230370A JP H02230370 A JPH02230370 A JP H02230370A JP 1051114 A JP1051114 A JP 1051114A JP 5111489 A JP5111489 A JP 5111489A JP H02230370 A JPH02230370 A JP H02230370A
- Authority
- JP
- Japan
- Prior art keywords
- word
- words
- adjacent
- candidate
- candidate word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 33
- 230000000877 morphologic effect Effects 0.000 claims description 20
- 230000001186 cumulative effect Effects 0.000 claims description 11
- 238000000034 method Methods 0.000 claims description 9
- WURBVZBTWMNKQT-UHFFFAOYSA-N 1-(4-chlorophenoxy)-3,3-dimethyl-1-(1,2,4-triazol-1-yl)butan-2-one Chemical compound C1=NC=NN1C(C(=O)C(C)(C)C)OC1=CC=C(Cl)C=C1 WURBVZBTWMNKQT-UHFFFAOYSA-N 0.000 abstract 1
- 230000011218 segmentation Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Landscapes
- Machine Translation (AREA)
Abstract
Description
〔産業上の利用分野〕
本発明は日英機械翻訳システム、日本語テキスト音声合
成システム等の必須構成要素である日本語の形態素解析
方式およびその装置に間するものである。
〔従来の技術〕
従来、単語の境界に空白などの切れ目がないという特徴
がある日本語テキストの解析を行なうために、単語の境
界を決定する形態素分割の種々の方式が提案されている
。これらには、たとえば「情報処理」第２７巻第８号９
５１ページに記載されているように、最長一致法、二文
節最長一致法、文節数最小法，拡張文節モデル上のコス
ト最小法等の日本語形態素解析の技術が知られている．
〔発明が解決しようとする課題〕
しかしながら、従来の形態素解析方式においては、単語
が有する、ある種の性質をもつ語とは隣接しやすいとい
う語禽的な情報を用いて多義解消を行なうことがなかっ
た．
本発明の目的は、このような欠点を改良した高精度の形
態素解析方式を提供することにある。
〔課題を解決するための手段〕
本発明の形態素解析方式は、単語の構文・意味的な諸性
質を単語の属性として有する辞書を用いる形態素解析方
式において、文中で隣接しやすい語がもつ属性を単語の
辞書情報として登録しておき、入力文中の各単語区間に
おいて、前記隣接しやすい語がもつ属性を実際に満足す
るような単語の隣接数が最大となるような候補単語を選
択することを特徴としている。
また、本発明の形態素解析装置は、文中で隣接しやすい
語がもつ属性が登録された辞書と、入力文を辞書引きに
よって候補単語区間に分割する手段と、前記各候補単語
区間における複数の候補単語を保持する手段と、前記各
候補単語区間の各候補単語に対して前記隣接しやすい語
がもつ属性を実際に満足するような隣接候補単語の数を
保持する手段と、各候補単語区間において、前記属性を
実際に満足するような隣接候補単語の数が最大であるよ
うな候補単語を選択する手段とを含んで構成されること
を特徴としている。
また、本発明の形態素解析方式は、単語の構文・意味的
な諸性質を単語の属性として有する辞書を用いる形態素
解析方式において、文中で隣接しやすい語がもつ属性を
単語の辞書情報として登録しておき、前記隣接しやすい
語がもつ属性を実際に隣接語が有するということの出現
数が最大となるような単語の組合せを選択することを特
徴としている．
また、本発明の形態素解析装置は、文中で隣接しやすい
語がもつ属性が登録された辞書と、入力文を辞書引きに
よって候補単語区間に分割する手段と、前記各候補単語
区間における複数の候補単語を保持する手段と、前記各
候補単語区間の各候補単語に対して、前記隣接しやすい
語がもつ属性を実際に満足するような隣接候補単語の識
別子を保持する手段と、前記各候補単語区間の各候補単
詰に対して、その単語までの文頭からの前記隣接しやす
い語がもつ属性を実際に満足するような隣接候補単語の
累積数が最大となるような前接候補単語の識別子を保持
する手段と、前記各候補単語区間の各候補単語に対して
、前記隣接しやすい語がもつ属性を実際に満足するよう
な隣接候補単語の数の文頭からの最大累積数を保持する
手段とを含んで構成されることを特徴としている。
〔作用〕
日本語の語禦には、同じ表記で別の意味をあらわす語・
辞が複数存在し得るという多義性の問題があり、形態素
解析処理においては、この多義性をできるだけ解消せね
ばならない。例えば、「高」という表記には、（１）ｒ
［高】いく形容詞語幹）Ｊ．（２）ｒ生産[Industrial Application Field] The present invention relates to a Japanese morphological analysis method and an apparatus thereof, which are essential components of a Japanese-English machine translation system, a Japanese text-to-speech synthesis system, and the like. [Prior Art] Various methods of morpheme segmentation for determining word boundaries have been proposed in the past in order to analyze Japanese text, which is characterized by the absence of breaks such as blanks at word boundaries. For example, "Information Processing" Vol. 27 No. 8 9
As described on page 51, techniques for Japanese morphological analysis are known, such as the longest match method, the two-clause longest match method, the minimum number of clauses method, and the minimum cost method on extended clause models. [Problem to be solved by the invention] However, in the conventional morphological analysis method, it is difficult to resolve ambiguity by using word-like information that words with certain properties tend to be adjacent to each other. There wasn't. An object of the present invention is to provide a highly accurate morphological analysis method that improves these drawbacks. [Means for Solving the Problems] The morphological analysis method of the present invention is a morphological analysis method that uses a dictionary that has various syntactic and semantic properties of words as attributes of words. It is registered as word dictionary information, and in each word section in the input sentence, candidate words are selected that have the maximum number of adjacent words that actually satisfy the attributes of the words that are likely to be adjacent. It is a feature. Further, the morphological analysis device of the present invention includes a dictionary in which attributes of words that are likely to be adjacent in a sentence are registered, a means for dividing an input sentence into candidate word sections by dictionary lookup, and a plurality of candidates in each candidate word section. means for retaining words; means for retaining a number of adjacent candidate words that actually satisfy the attributes of the words that are likely to be adjacent to each candidate word in each candidate word interval; , and means for selecting a candidate word for which the number of adjacent candidate words that actually satisfy the attributes is the largest. Furthermore, the morphological analysis method of the present invention is a morphological analysis method that uses a dictionary that has various syntactic and semantic properties of words as word attributes, and registers the attributes of words that are likely to be adjacent in a sentence as word dictionary information. The method is characterized in that a combination of words is selected such that the number of occurrences of adjacent words actually having the attribute of the words that are likely to be adjacent is maximized. Further, the morphological analysis device of the present invention includes a dictionary in which attributes of words that are likely to be adjacent in a sentence are registered, means for dividing an input sentence into candidate word sections by dictionary lookup, and a plurality of candidates in each candidate word section. means for holding words; means for holding, for each candidate word in each candidate word section, an identifier of an adjacent candidate word that actually satisfies the attributes of the words that are likely to be adjacent; and each candidate word. For each candidate simple in the interval, the identifier of the prefix candidate word that maximizes the cumulative number of adjacent candidate words that actually satisfy the attributes of the words that are likely to be adjacent from the beginning of the sentence up to that word. and means for maintaining, for each candidate word in each candidate word section, the maximum cumulative number of adjacent candidate words from the beginning of the sentence that actually satisfy the attributes of the words that are likely to be adjacent. It is characterized by being comprised of the following. [Effect] Japanese words include words that have the same meaning but have different meanings.
There is a problem of ambiguity in that a plurality of words may exist, and this ambiguity must be resolved as much as possible in the morphological analysis process. For example, the notation "high" includes (1) r
[High] Iku adjective stem) J. (2) r production
【高】 （接
尾辞・量をあらわす）Ｊ，（３）’立川[High] (suffix/representing amount) J, (3)'Tachikawa
【高】 （接尾
辞・高校をあらわす）Ｊ、（４）’ドル[High] (suffix indicating high school) J, (4)' dollar
次に、第２図を参照しつつ、本発明の第一の実施例につ
いて詳細に説明する．
第２図は本発明の原理を実現するための一実施例をあら
わすブロック図である．
入力文は、形慧素分割部２０１の処理の結果、第１図に
みられるような単語区間の系列となり、分割結果保持部
２０４に出力される．ひとつの単語区間には、複数の候
補単語が存在する可能性がある。解析制御部２０３は、
形態素分割部２０１の処理終了信号を受取ったら、分割
結果保持部２０４中の各単語区間の各候補単語に対して
、次の処理を行なう．
（１）辞書情報中にＰＲＥが存在するならば、そこに記
述されている情報をもつ候補単語が左側の単語区間に存
在するか否かをチェックし、もし存在するならば、Ｌ＝
１とする．そうでなければ、Ｌ＝Ｏとする．
（２》辞書情報中にＰＯＳＴが存在するならば、そこに
記述されている情報をもつ候補単語が右側の単語区間に
存在するか否かをチェックし、もし存在するならば、Ｒ
＝１とする．そうでなければ、Ｒ＝Ｏとする。
（３）Ｓ＝Ｌ＋Ｒを求め、これを満足条件数保持部２０
５の現在の候補単語に対するエントリーとする．
解析制御部２０３の処理終了信号を受け取ったら、最大
候補選択部２０６は、満足条件数保持部２０５に保持さ
れている各単語区間において、Ｓの値が最大であるよう
な候補単語を選択し、その区間の単語として出力する．
以上に説明した実施例の形態素解析装置により、各単語
区間において条件満足数が最大となる解析結果が得られ
る．
次に、本発明の第２の実施例について、第３図を参照し
ながら説明する．第３図は、本発明の原理を実現するた
めのブロック図である．入力文は、形態素分割部３０１
の処理の結果、第１図にみられるような単語区間の系列
となり、分割結果保持部３０４に出力される．解析制御
部３０３は、形態素分割部３０１の処理終了信号を受取
ったら、分割結果保持部３０４中の各単語区間の各候補
単語Ｗに対して、次の処理を行なう。
（１−１）辞書情報中にＰＲＥが存在するならば、そこ
に記述されている情報をもつ候補単語が左側の単語区間
に存在するか否かをチェックし、もし存在するならば、
その識別子の集合をｗ，Ｌとする．
（１−２）辞書情報中にＰＯＳＴが存在するならば、そ
こに記述されている情報をもつ候補単語が右側の単語区
間に存在するか否かをチェックし、もし存在するならば
、その識別子の集合をｗ，Ｒとする。
（１−３） ２つ組（ｗ，Ｌ，ｗ，Ｒ）を、条件満足単
語識別子保持部３０５の現在の候補単語に対するエント
リーとする。
次に、解析制御部３０３は、分割結果保持部３０４に保
持されている各単語区間の各候補単語Ｗに対して、次の
処理を行なう。
（２−１）左側の単語区間の各候補単語ｗｊに対して以
下の処理を行なう。
もし、ｗｊ，ＲにＷの識別子が含まれているならば、ｐ
ｉ （ｗｊ，ｗ）＝１、そうでないならば、ｐｉ （ｗ
ｊ，ｗ）＝Ｏとする。
もし、ｗ，Ｌにｗｊの識別子が含まれているならば、ｐ
２ （ｗｊ ，ｗ＞＝１、そうでないならば、ｐ２ （
ｗｊ ，ｗ＞＝Ｏとする。
（２−２） ｗに対する最大累積満足条件数Ｓを次式に
よって求め、累積満足条件数保持部３０６に格納する．
ここで、Ｓｊは、ｗｊにおける最大累積満足条件数であ
る。
Ｓ ＝ ｍ３ｘ（Ｓｊ ＋Ｐ１ （ｗｊ ． ｗ）＋ｐ
２（ｗｊ ． ｗ））また、この右辺の最大値を与える
候補単語Ｗｊ＊の識別子を前接単語識別子保持部３０７
の現在の候補単語Ｗに対するエントリーｂ（ｗ）として
格納する．
（３）最後に、解析制御部３０３は、累積満足条件数保
持部３０６より、最も右側の単語区間に対する最大累積
満足条件数Ｓが最大となる候補単語Ｗ＊を同定し、前接
単語識別子保持部３０７を参照して、ｂ（ｗ＊）から順
に、前接単語をつぎつぎと左端まで決定して行く．これ
によって得られる単語列を、解析結果として出力する。
以上説明した本実施例の形態素解析装置により、文全体
にわたって条件満足数の累積が最大となる解析結果が得
られる．
において、辞書に保持された隣接に関する選択情報を利
用して、強力な多義解消機能を実現することが可能とな
る。Next, a first embodiment of the present invention will be described in detail with reference to FIG. Figure 2 is a block diagram showing an embodiment for realizing the principle of the present invention. As a result of processing by the shape segment division unit 201, the input sentence becomes a series of word sections as shown in FIG. 1, and is output to the division result holding unit 204. There is a possibility that a plurality of candidate words exist in one word section. The analysis control unit 203
When the processing end signal of the morpheme segmentation unit 201 is received, the following processing is performed on each candidate word in each word section in the segmentation result holding unit 204. (1) If PRE exists in the dictionary information, check whether a candidate word with the information described there exists in the word section on the left, and if it exists, L=
Set it to 1. Otherwise, let L=O. (2) If POST exists in the dictionary information, check whether a candidate word with the information described there exists in the word section on the right, and if it exists, R
=1. Otherwise, let R=O. (3) Find S=L+R and store it in the satisfaction condition number holding unit 20
This is the entry for the current candidate word of 5. Upon receiving the processing end signal from the analysis control unit 203, the maximum candidate selection unit 206 selects the candidate word with the maximum value of S in each word section held in the satisfaction condition number holding unit 205, Output as words for that interval. The morphological analysis device of the embodiment described above can obtain an analysis result that maximizes the number of condition satisfactions in each word section. Next, a second embodiment of the present invention will be described with reference to FIG. FIG. 3 is a block diagram for realizing the principle of the present invention. The input sentence is processed by the morpheme dividing unit 301
As a result of the processing, a series of word sections as shown in FIG. 1 is obtained, which is output to the division result holding unit 304. When the analysis control unit 303 receives the processing end signal from the morpheme division unit 301, it performs the following processing on each candidate word W in each word section in the division result holding unit 304. (1-1) If PRE exists in the dictionary information, check whether a candidate word with the information described there exists in the word section on the left, and if so,
Let w and L be the set of identifiers. (1-2) If POST exists in the dictionary information, check whether a candidate word with the information described there exists in the word section on the right, and if it exists, its identifier Let the set of w and R be w and R. (1-3) Let the pair (w, L, w, R) be the entry for the current candidate word in the condition satisfying word identifier holding unit 305. Next, the analysis control unit 303 performs the following processing on each candidate word W in each word section held in the division result holding unit 304. (2-1) The following processing is performed for each candidate word wj in the left word section. If wj,R contains the identifier of W, then p
i (wj, w)=1, otherwise pi (w
j, w)=O. If w and L contain the identifier of wj, then p
2 (wj , w>=1, otherwise p2 (
Let wj , w>=O. (2-2) The maximum cumulative number of satisfied conditions S for w is determined by the following equation and stored in the cumulative number of satisfied conditions holding unit 306.
Here, Sj is the maximum cumulative number of satisfied conditions in wj. S = m3x(Sj +P1 (wj.w)+p
2(wj.w)) Also, the identifier of the candidate word Wj* that gives the maximum value on the right side is stored in the prefix word identifier holding unit 307.
is stored as an entry b(w) for the current candidate word W. (3) Finally, the analysis control unit 303 identifies a candidate word W* with the maximum cumulative number of satisfaction conditions S for the rightmost word section from the cumulative number of satisfaction conditions holding unit 306, and holds the preceding word identifier. Referring to section 307, prefix words are determined one after another starting from b(w*) until the left end. The word string obtained by this is output as the analysis result. The morphological analysis device of this embodiment described above can obtain an analysis result that maximizes the cumulative number of condition satisfactions over the entire sentence. In this case, it is possible to implement a powerful disambiguation function by using the selection information regarding adjacencies held in the dictionary.
第１図は形態素分割部から解析制御部へ出力される形態
素分割結果の一例をあらわす図、第２図及び第３図は本
発明の第１及び第２の実施例を示すブロック図である．
２０１・・・形態素分割部、２０２・・・辞書、２０３
・・・解析制御部、２０４・・・分割結果保持部、２０
５・・・満足条件数保持部、２０６・・・最大候補選択
部、３０１・・・形態素分割部、３０２・・・辞書、３
０３・・・解析制御部、３０４・・・分割結果保持部、
３０５・・・条件満足単語識別子保持部、３０６・・・
累積満足条件数保持部、３０７・・・前接単語識別子保
持部。FIG. 1 is a diagram showing an example of the morpheme segmentation results output from the morpheme segmentation unit to the analysis control unit, and FIGS. 2 and 3 are block diagrams showing first and second embodiments of the present invention. 201...Morpheme division unit, 202...Dictionary, 203
...Analysis control unit, 204...Division result holding unit, 20
5... Satisfied condition number holding unit, 206... Maximum candidate selection unit, 301... Morphological division unit, 302... Dictionary, 3
03... Analysis control unit, 304... Division result holding unit,
305...Condition satisfying word identifier holding unit, 306...
Cumulative satisfaction condition number holding unit, 307 . . . Prefix word identifier holding unit.
Claims (4)
有する辞書を用いる形態素解析方式において、文中で隣
接しやすい語がもつ属性を単語の辞書情報として登録し
ておき、入力文中の各単語区間において、前記隣接しや
すい語がもつ属性を実際に満足するような単語の隣接数
が最大となるような候補単語を選択することを特徴とす
る形態素解析方式。(1) In a morphological analysis method that uses a dictionary that has syntactic and semantic properties of words as word attributes, the attributes of words that are likely to be adjacent in a sentence are registered as word dictionary information, and each A morphological analysis method characterized in that, in a word interval, a candidate word is selected such that the number of adjacent words that actually satisfy the attributes of the words that are likely to be adjacent is maximized.
書と、入力文を辞書引きによって候補単語区間に分割す
る手段と、前記各候補単語区間における複数の候補単語
を保持する手段と、前記各候補単語区間の各候補単語に
対して前記隣接しやすい語がもつ属性を実際に満足する
ような隣接候補単語の数を保持する手段と、各候補単語
区間において、前記属性を実際に満足するような隣接候
補単語の数が最大であるような候補単語を選択する手段
とを含むことを特徴とする形態素解析装置。(2) a dictionary in which attributes of words that are likely to be adjacent in a sentence are registered; means for dividing an input sentence into candidate word sections by dictionary lookup; and means for holding a plurality of candidate words in each candidate word section; means for maintaining a number of adjacent candidate words that actually satisfy the attributes of the words that are likely to be adjacent to each candidate word in each candidate word section; and means that actually satisfy the attributes in each candidate word section. A morphological analysis device comprising means for selecting a candidate word having the maximum number of adjacent candidate words.
有する辞書を用いる形態素解析方式において、文中で隣
接しやすい語がもつ属性を単語の辞書情報として登録し
ておき、前記隣接しやすい語がもつ属性を実際に隣接語
が有するということの出現数が最大となるような単語の
組合せを選択することを特徴とする形態素解析方式。(3) In a morphological analysis method that uses a dictionary that has syntactic and semantic properties of words as word attributes, the attributes of words that are likely to be adjacent in a sentence are registered as word dictionary information, and the A morphological analysis method characterized by selecting a combination of words that maximizes the number of occurrences of adjacent words actually having the attribute that the word has.
書と、入力文を辞書引きによって候補単語区間に分割す
る手段と、前記各候補単語区間における複数の候補単語
を保持する手段と、前記各候補単語区間の各候補単語に
対して、前記隣接しやすい語がもつ属性を実際に満足す
るような隣接候補単語の識別子を保持する手段と、前記
各候補単語区間の各候補単語に対して、その単語までの
文頭からの前記隣接しやすい語がもつ属性を実際に満足
するような隣接候補単語の累積数が最大となるような前
接候補単語の識別子を保持する手段と、前記各候補単語
区間の各候補単語に対して、前記隣接しやすい語がもつ
属性を実際に満足するような隣接候補単語の数の文頭か
らの最大累積数を保持する手段とを含むことを特徴とす
る形態素解析装置。(4) a dictionary in which attributes of words that are likely to be adjacent in a sentence are registered; means for dividing an input sentence into candidate word sections by dictionary lookup; and means for holding a plurality of candidate words in each candidate word section; means for holding, for each candidate word in each of the candidate word sections, an identifier of an adjacent candidate word that actually satisfies the attributes of the words that are likely to be adjacent; and for each candidate word in each of the candidate word sections; means for holding an identifier of a preceding candidate word that maximizes the cumulative number of adjacent candidate words that actually satisfy the attributes of the words that are likely to be adjacent from the beginning of sentences up to the word; For each candidate word in the candidate word section, the method further comprises means for maintaining, for each candidate word in the candidate word section, the maximum cumulative number of adjacent candidate words from the beginning of the sentence that actually satisfy the attributes of the words that are likely to be adjacent. Morphological analysis device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP1051114A JP2526657B2 (en) | 1989-03-02 | 1989-03-02 | Morphological analyzer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP1051114A JP2526657B2 (en) | 1989-03-02 | 1989-03-02 | Morphological analyzer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPH02230370A true JPH02230370A (en) | 1990-09-12 |
| JP2526657B2 JP2526657B2 (en) | 1996-08-21 |
Family
ID=12877780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP1051114A Expired - Lifetime JP2526657B2 (en) | 1989-03-02 | 1989-03-02 | Morphological analyzer |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP2526657B2 (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5990167A (en) * | 1982-11-12 | 1984-05-24 | Fujitsu Ltd | Sentence analyzing system |
| JPS61187077A (en) * | 1985-02-14 | 1986-08-20 | Ricoh Co Ltd | Japanese language analysis device |
-
1989
- 1989-03-02 JP JP1051114A patent/JP2526657B2/en not_active Expired - Lifetime
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5990167A (en) * | 1982-11-12 | 1984-05-24 | Fujitsu Ltd | Sentence analyzing system |
| JPS61187077A (en) * | 1985-02-14 | 1986-08-20 | Ricoh Co Ltd | Japanese language analysis device |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2526657B2 (en) | 1996-08-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5794177A (en) | Method and apparatus for morphological analysis and generation of natural language text | |
| CN104679850B (en) | Address structure method and device | |
| WO1997004405A9 (en) | Method and apparatus for automated search and retrieval processing | |
| JPH02297195A (en) | Morpheme analysis system | |
| JP3992348B2 (en) | Morphological analysis method and apparatus, and Japanese morphological analysis method and apparatus | |
| JPH06149887A (en) | Text type database device | |
| KR100282610B1 (en) | Korean morphological analyzer based on partial string discrimination and its method | |
| KR100374114B1 (en) | Hyperlink generator for korean language terminology based HTML | |
| KR970049752A (en) | Korean Natural Language Query Information Retrieval Using Verb Information | |
| Uchimoto et al. | Morphological Annotation of a Large Spontaneous Speech Corpus in Japanese. | |
| JP2526657B2 (en) | Morphological analyzer | |
| JP2655711B2 (en) | Homomorphic reading system | |
| JPS63228326A (en) | Automatic key word extracting system | |
| JP2897942B2 (en) | Japanese morphological analysis system and morphological analysis method | |
| JPH0785040A (en) | Inconsistent notation detection method and kana-kanji conversion method | |
| JPH02140869A (en) | Sentence structure analyzing method | |
| JPH10269228A (en) | Important part extracting device, and document retrieval device | |
| JP2786211B2 (en) | Post-editing device in machine translation equipment | |
| JPH0756937A (en) | Word extraction system | |
| JPS6395570A (en) | Language analysis system | |
| JPH08212225A (en) | Language judgment device | |
| JPH0855121A (en) | Sentence correspondence device and corresponding method between original and translated sentence | |
| JPH02257363A (en) | Automatic translation device | |
| JPS62203276A (en) | Form element analysis device | |
| JPH09319746A (en) | Document analysis method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20080614 Year of fee payment: 12 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20090614 Year of fee payment: 13 |
|
| EXPY | Cancellation because of completion of term | ||
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20090614 Year of fee payment: 13 |