JP2012203868A - Skim reading support system, skim reading support method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To support skim reading of a large number of documents.SOLUTION: According to an embodiment, a system comprises a document storage, a display unit, an input unit, a classified information storage, an extractor and a specifying unit. The document storage stores a plurality of documents together with identifying information. A user accesses a document on the display unit, and designates the classification type to be assigned to the document from the input unit. The classified information storage stores the classification type of the document designated by the user. The extractor extracts, out of one or more of the documents assigned the same classification type, one or more words or phrases to be highlighted in display regarding the classification type. The specifying unit specifies, regarding each of documents to which the user has assigned no classification type, a section in which the word(s) or phrase(s) to be highlighted in the document. The display unit, in displaying a document, highlights the word(s) or phrase(s) in the section.

Description

  Embodiments described herein relate generally to a browsing support system, method, and program for supporting browsing of a large number of documents.

  With the widespread use of computers, with the advancement of hardware such as higher communication speed and processing speed, and increased capacity of hard disks and memories, the digitization of documents has progressed and a large amount of information has been handled on a daily basis. . For general users, along with technological innovations of various software such as information retrieval, it has become possible to receive support for using information that they want to know from a large amount of information.

  However, in business scenes where content must be thoroughly investigated, such as patent research, literature research, and market research, for example, users need to read a large number of documents. It is difficult to reduce the number of documents to be read with high accuracy to a small amount. In addition, there are generally time restrictions, and browsing is usually performed such as selecting a portion to be read carefully, manually or unconsciously. Alternatively, as a pre-processing for detailed reading so that a large number of documents can be shared by a plurality of people, allocation is performed by browsing based on each person's specialized field.

JP 2008-171164 A

  No technology has been known to assist in browsing a large number of documents.

  An object of the present embodiment is to provide a browsing support system, a browsing support method, and a program that can support browsing of a large number of documents.

  According to the embodiment, a document storage unit, a display unit, an input unit, a classification information storage unit, an extraction unit, and a specifying unit are provided. The document storage unit stores a plurality of documents associated with identification information. The display unit displays all or a part of the plurality of documents with or without highlight display of words or phrases. The input unit inputs, from the user, an instruction of a specific document among the displayed documents and an instruction of a specific classification type to be given to the specific document among a plurality of predetermined classification types. The classification information storage unit stores classification information in which the identification information is associated with the specific classification type. The extraction unit extracts one or a plurality of words or phrases to be highlighted from the one or a plurality of the documents associated with the same classification type. The specifying unit specifies a portion to be highlighted in the document when each word or phrase extracted by the extracting unit is present in the document for each of all or a part of the document.

The figure which shows the function structural example of the browsing assistance system which concerns on 1st Embodiment. The figure which shows an example of the document data memorize | stored in a document data memory | storage part. The figure which shows an example of the document classification information (initial state) memorize | stored in a classification information storage part. The figure which shows an example of the scored word list (initial state) produced by the browsing word extraction part. The figure for demonstrating the system screen example and operation example of a browsing assistance system. The flowchart which shows an example of the process sequence of the browsing assistance system of 1st Embodiment. The figure which shows the example of a document display without a highlight display. The figure which shows an example of the updated document classification information. The figure which shows an example of the updated scored word list. The flowchart which shows an example of the process sequence of the browsing word extraction part. The flowchart which shows an example of the process sequence of a highlight location specific | specification part. The flowchart which shows an example of the process sequence of a text display part. The figure which shows the example of a document display accompanied with a highlight display. The figure which shows the function structural example of the browsing assistance system which concerns on 2nd Embodiment. The flowchart which shows an example of the process sequence of the browsing word extraction part. The figure for demonstrating an example of the word substitution scored and the word substitution between different languages. The flowchart which shows the other example of the process sequence of a text display part. The figure for demonstrating an example of the word substitution scored and the word substitution between different languages. The figure which shows the example of a document display accompanied with the highlight display by a different language. The figure which shows the function structural example of the browsing assistance system which concerns on 3rd Embodiment. The figure which shows the function structural example of the browsing assistance system which concerns on 4th Embodiment.

  Hereinafter, a browsing support system according to an embodiment of the present invention will be described in detail with reference to the drawings. Note that, in the following embodiments, the same numbered portions are assumed to perform the same operation, and repeated description is omitted.

(First embodiment)
Conventionally, when browsing a large number of documents, such as patent research, literature research, market research and other document research, etc., the number of documents to be read with high accuracy is reduced to a small amount only by mechanical processing such as search and classification. It was difficult, and the user himself scanned the word with his eyes and browsed it.

  In the first embodiment, based on information (classification type) indicating classifications for some documents input by the user (for example, after the user arbitrarily browses the displayed document and determines the classification type). An example in which browsing keyword extraction is automatically performed and the user supports browsing by highlighting keywords in an uncategorized document will be described as an example.

Hereinafter, a specific example in which at least the following two types of document types are provided as the classification type to be given to a document by the user depending on whether the document is a necessary document will be mainly described.
(A) a classification type (hereinafter referred to as a required document type) indicating a document that the user considers necessary;
(B) A classification type (hereinafter referred to as “unnecessary document type”) indicating a document that the user does not need.

  In the case of this example, the user can input either “required document type” or “unnecessary document type” as a classification type for a desired document.

  The above is an example, and various other classification methods can be used.

For example, it is possible to classify according to which person in charge the document is related to. For example, when five persons A to E are set, at least the following five document types are provided.
A classification type (person A document type) indicating a document related to person A
A classification type (person B document type) indicating a document related to person B
A classification type (person C document type) indicating a document related to the person C
A classification type (person in charge D document type) indicating a document related to person in charge D;
A classification type indicating a document related to the person in charge E (person in charge E document type).

  In the case of this example, the user can input any one of “person A document type” to “person E document type” as a classification type for a desired document.

In addition, for example, at least the following six document types can be provided in combination with the above two examples.
A classification type (person A document type) indicating a document related to person A
A classification type (person B document type) indicating a document related to person B
A classification type (person C document type) indicating a document related to the person C
A classification type (person in charge D document type) indicating a document related to person in charge D;
-Classification type (person E document type) indicating the document related to person E
A classification type (unnecessary document type) indicating a document that the user does not need.

  As one of the classification types, a type (herein referred to as an unread document type) indicating that the user has not yet input a classification type can be provided. For example, in the above specific example, when the unread document type is used further, the user inputs the required document type for some documents and the unnecessary document type for some other documents. In addition, the “unread document type” is automatically assigned to the remaining documents.

  FIG. 1 shows a functional configuration example of the browsing support system according to the first embodiment.

  As shown in FIG. 1, the browsing support system according to the present embodiment includes a browsing support system including a browsing word extraction unit 101, a highlight location specifying unit 102, a text display unit 103, a user input unit 104, and a document data storage unit. 105, a classification information storage unit 106 is provided.

  The document data storage unit 105 stores data of a plurality of documents.

  The document may be anything. For example, the document may be any document body or a summary sentence for the document body. For example, when the document text is a patent specification and each document stored in the document data storage unit 105 is a summary sentence corresponding to each patent specification, the user can select each patent document corresponding to each summary sentence. Each summary sentence may be browsed to determine whether or not it is actually read. In the present embodiment, a summary sentence of a patent specification will be described as an example of a document.

  FIG. 2 shows an example of document data stored in the document data storage unit 105. In the example of FIG. 2, a document identifier (hereinafter, document ID) (0001 to 0006 in the figure) is assigned to each document data. The document data storage format is not limited to that shown in FIG.

  The document data stored in the document data storage unit 105 may be input from a recording medium, may be downloaded via a network such as the Internet, or may be input by key input. May be included, or may be obtained by any other method.

  Further, the document data stored in the document data storage unit 105 may be sorted by, for example, the document ID or some other standard, or may not be sorted in particular.

  The text display unit 103 displays a desired number of documents d at a time for a plurality of documents (the number of documents is N) stored in the document data storage unit 105. In addition, although it demonstrates as 2 <= d <= N here, there may exist a display state of d = 1.

  At this time, there is no limitation on the display method. For example, for each document to be displayed, a method of displaying the entire document is also possible. For example, a part of the document is displayed (for example, A method is also possible in which the document is displayed up to a predetermined upper limit number of characters, and the remaining portion is displayed by the user performing a predetermined operation. Further, the user may be able to specify the number n of documents to be displayed simultaneously on one screen, the user may be able to specify the upper limit number of characters to be displayed for one document, and various other displays. A method is possible.

  In addition, when displaying a part of a plurality of documents, there is no limitation on a method for selecting a document to be displayed. For example, the document to be displayed may be selected based on the document ID or some other criterion, or the user may indicate the document to be displayed.

  The user input unit 104 inputs a document designation designated by the user and a classification type designation designated by the user for the document. For example, the user may select a desired document from a document displayed on the text display unit 103 by a predetermined method and select a desired classification type for the document by a predetermined method. There are no particular restrictions on the document selection method and the classification type selection method.

  As for the classification type, the classification method may be determined in advance as one type (for example, “necessary document type” or “unnecessary document type”), and the classification methods are determined in advance as a plurality of types. The user may select from among them (for example, “necessary document type”, “unnecessary document type”, or “person A document type”) to “person E document type”. ), The user can arbitrarily set a free classification type at any time, or a combination thereof.

  The classification information storage unit 106 stores document classification information indicating correspondence between each document ID and a classification type for the document ID.

  FIG. 3A shows an example of document classification information stored in the classification information storage unit 106. The example of FIG. 3A shows an initial state in which no classification type is input. Hereinafter, as an example, in the document classification information, it is assumed that a value “A” is stored for “unnecessary document type” and a value “B” is stored for “required document type”.

  If the above-mentioned “unread document type” is not used, the example of FIG. 3A may be used. When “unread document type” is used, values other than the classification type input by the user (for example, value “C”) are set for all documents as the initial state of the document classification information as shown in FIG. A document that is stored and the value “C” is stored may be handled as “unread document type”, or any value is stored using the example of FIG. A document that is not (or has a value of “null”) may be handled as an “unread document type”.

  In any case, when the number of documents stored in the document data storage unit 105 is N, the number c in which the value corresponding to the classification type input by the user in the document classification information is stored is: 0 ≦ c ≦ N.

  In the following, when c = 0, no highlight display is performed, and when c = N is reached, it means that all documents have been classified by the user. The processing of the browsing word extraction unit 101 and the highlight location specifying unit 102 is not performed. However, when the “unread document type” is used, it is possible to perform highlight display when c = 0.

  In the following description, it is assumed that only (Nc) documents that have not yet been assigned a classification type by the user are to be highlighted. However, all N documents can be highlighted.

  The browsing word extraction unit 101 corresponds to the document classification information when the classification type input by the user is stored for at least one document in the document classification information (that is, when c ≧ 1). For each classification type having a value to be generated, a list of words characteristic of the classification type is created from a document corresponding to the classification type. At that time, for each word, a score characteristic to the classification type and representing a certain level is calculated.

  FIG. 4 shows an example of a scored word list (scored word list) created by the browsing word extraction unit 101 in the process. When “necessary document type” and “unnecessary document type” are used as in this specific example, a scored word list is created for each document type. The examples of FIGS. 4A and 4B show initial states of the scored word list for “unnecessary document type” and the scored word list for “necessary document type”, respectively. When “unread document type” is used, a scored word list for “unread document type” is also provided.

  For each classification type, the words in the scored word list are selected as highlighted. In addition, for example, when the number of words in the scored word list exceeds a predetermined number k, a method of selecting the top k scores in the scored word list, the score is equal to or higher than a predetermined threshold A method using only these words or a combination of them is also possible, and various other methods are possible.

  Note that a word list with scores is not created for a classification type for which there is no corresponding value in the document classification information. For example, in the above specific example, in the document classification information, only the value “B” exists for FIG. 3A, and only “B” and “C” exist for FIG. 3B. For the “unnecessary document type” corresponding to the value “A”, a scored word list is not created.

  Hereinafter, the word (or term or phrase) selected to be highlighted for each classification type will be referred to as a “browsing word candidate”.

  For each document for which the user has not entered a classification type, the highlighted location specifying unit 102 searches for a location where each “browsing word candidate” appears in the document (note that all of the “browsing word candidates” or A part to be present may be absent), and a portion to be highlighted is selected from each “browsing word candidate” appearing in the document. For example, in a single document, if only one location is highlighted for the same “browsing word candidate”, when there are multiple “browsing word candidates” in a document, any location is highlighted. Select whether to display light.

  After the user inputs the classification type for at least one document, the text display unit 103 classifies the document for which the user has not input the classification type by the browsing word extraction unit 101 among the words in the document. A word extracted for each type and highlighted at a location specified by the highlighted location specifying unit 102 is highlighted.

  At this time, when the “browsing word candidate” is highlighted, it may be highlighted in a highlight form corresponding to the classification type. For example, when the browsing word candidate is shown by changing the font, the font such as the character color may be changed for each classification type, or when the browsing word candidate is surrounded by a frame and shown for each classification type, The shape, line type, color, presence / absence of hatching in the frame, the type of hatching, etc. may be changed, or combinations thereof, and various other highlight forms are possible.

  Further, for example, a method is possible in which the necessary document type is the most noticeable highlight form, the unnecessary document type is the next most noticeable highlight form, and the unread document type is the next most noticeable highlight form.

  When an input from the user occurs in the user input unit 104, the classification information storage unit 106 is updated, and a series of processes of the browsing word extraction unit 101, the highlight location specifying unit 102, and the text display unit 103 are performed.

  The series of processes may be executed every time document classification information related to one document is input from the user input unit 104, or the document classification information is received from the user input unit 104. The above-described series of processing is not executed only by the input. When a predetermined instruction for executing the above-described series of processing is input (for example, from the user input unit 104), the above-described series of processing is triggered by this. It is also possible to execute.

  Here, an outline of an overall operation example of the present embodiment will be described with reference to the system screen example of FIG. In FIG. 5, a case where each document is a summary corresponding to a patent specification will be described as an example.

  In the following, the necessary document type is also represented by “◯” and the unnecessary document type is represented by “×”.

  First, the text display unit 103 displays each document (see 122 in the figure) inside the system screen (121) of FIG. In the initial state, the classification type is not input by the user. Actually, unlike FIG. 5, “◯” and “x” are not displayed and no highlight display is performed.

  The specific example of FIG. 5 displays each document in five rows and two columns inside the system screen (121).

  Next, the user arbitrarily reads the document displayed on the system screen (121) of FIG.

  In the example of displaying in 5 rows and 2 columns in FIG. 5, when there are 11 or more documents, the user may display other documents by scrolling or page updating, for example. Of course, the display of the document is not limited to the display of 5 rows and 2 columns.

  Then, the user inputs the document classification information {document, classification type} by selecting the document determined to input the classification type and selecting the classification type determined by the user for the document. That is, the system side accepts input of {document, classification type} selected by the user in the user input unit 104.

  For example, when the user reads the document indicated by 124 and determines that this document is unnecessary (for example, when it is determined that it is not necessary to read the full text of the patent specification corresponding to this summary sentence), the document data ( 124), the user selects the unnecessary document type symbol “x”.

  Similarly, when the user reads the document indicated by 125 and determines that this document is necessary (for example, when it is determined that the full text of the patent specification corresponding to this summary sentence needs to be read), this document data For (124), the user selects the required document type symbol “◯”.

  It is assumed that the user has not assigned a classification to other documents.

  In this case, as shown in FIG. 5, “x” and “◯” may be displayed indicating the classification type assigned to each of the documents 124 and 125 to which the classification is assigned. Of course, other assigned classification types may be identifiable.

  Now, when a classification type is assigned to the document data 124, 125, in the system, the document classification information indicating the document ID of the document data 124 and the assigned classification type, and the document ID of the document data 125 and the assignment are given. Document classification information indicating the classified classification type is received.

  Then, based on the above two document classification information, a series of processes including the process of the browsing word extraction unit 101, the process of the highlight location specifying unit 102, and the process of the text display unit 103 is performed, and is illustrated in FIG. As shown, the browsing words are highlighted on the documents other than the document data 124 and 125.

  The browsing word extraction unit 101 calculates a score for each classification type.

  For example, as shown in the type classification legend (123) in FIG. 5, a word highlighting method is prepared for each type by preparing words with high scores for each of the required document type, unnecessary document type, and unread document type. Can be changed.

  As an example of applying different highlight methods, for example, the highlight color may be changed as described above. For example, each type of browsing word may be highlighted with the required document type pink, the unnecessary document type yellow, and the unread document type green.

  In FIG. 5, in order to exemplify how the word highlighting method is changed for each type, an unread document is displayed in a cross-hatching frame for a portion of a word to be highlighted in a necessary document type and in a hatched hatch frame for an unnecessary document type. An example of highlighting each type with a frame without hatching is shown.

  In the display state illustrated in FIG. 5, the user can browse the highlighted word while browsing the document group with the highlighted word, and further to the unclassified document. Classification can be given. At that time, for example, if necessary, peripheral words of the highlighted word can be read together.

  FIG. 6 shows an example of the processing procedure of the browsing support system of this embodiment.

  In step S1, the text display unit 103 initially displays a document.

  FIG. 7 shows an example in which the document illustrated in FIG. 2 is displayed.

  In this initial stage, the document classification information stored in the classification information storage unit 106 is as illustrated in FIG. 3A or FIG. The scored word list is as illustrated in FIG.

  In step S2, the user input unit 104 receives input of document classification information {document ID, classification type} from the user.

  In step S 3, the input document classification information {document ID, classification type} is recorded in the classification information storage unit 106.

  Here, it is assumed that “unnecessary document type”, “unnecessary document type”, and “necessary document type” are input by the user for each document ID = 0001 to 0003 in FIG. Examples of document classification information stored in the unit 106 are illustrated in FIG. 8A for FIG. 3A and illustrated in FIG. 8B for FIG. 3B.

  In step S4, if the end condition is satisfied, the process is ended. If the end condition is not satisfied, the process proceeds to the next step S5.

  Various termination conditions can be considered. For example, when the number of documents stored in the document data storage unit 105 is N, the above-described document classification information {document ID, classification type} has been input by the user for all N documents. The condition may be used, or the number of documents in which the document classification information {document ID, classification type} is input by the user is c, and the value of (N−c) (that is, the classification type is still assigned by the user) The termination condition may be that the number of undocumented documents) falls below a predetermined threshold. Of course, it is not limited to these.

  In step S4, if the end condition is not satisfied, the following series of processing is performed.

  In step S5, the browsing word extraction unit 101 performs processing to create a scored word list.

  For example, a scored word list for “unnecessary document type” is illustrated in FIG. 9A, and a scored word list for “necessary document type” is illustrated in FIG. 9B. Here, a1, a2, b1, and b2 indicate the respective score values. Each scored word list may be sorted in the order of score.

  In step S6, the highlight part specifying unit 102 performs processing to specify a highlight part in each document.

  In step S7, the text display unit 103 is processed to update the text display (highlighted state thereof) (see, for example, FIG. 13 described later).

  Hereafter, each process of the browsing word extraction part 101, the highlight location specific | specification part 102, and the text display part 103 is demonstrated in detail in order.

  First, the browsing word extraction unit 101 will be described.

  The browsing word extraction unit 101 is a module that extracts words suitable for browsing from document data and known classifications.

  FIG. 10 shows an example of the processing procedure of the browsing word extraction unit 101.

  In step S <b> 11, the browsing word extraction unit 101 reads the document data and the classification assigned by the user to the document data from the document data storage unit 105 and the classification information storage unit 106.

  In step S12, a score as a browsing word is calculated from the classification of each document and the word of each document.

  In step S13, the words are sorted in the order of the scores, and the upper word is used as a reading word candidate. In step S14, the reading word candidate and its score are output.

As the score calculation of the browsing word, for example, the calculation may be performed by the following formula. Here, the quickness (t, C) is introduced as a score “whether it is possible to immediately determine that the document D is the document type C when the word t appearing in the document D is seen”. .
Speed (t, C) = reading cost (t) × discrimination (t, C) (1)
Here, the reading cost refers to the cost required for a person to recognize a word, and depends on the character length and the complexity of the character. For example, the reciprocal of the number of characters (eg, 1 / (number of characters)) is used. Considering that the character types such as hiragana and katakana have fewer strokes than the kanji and are easily noticeable, it is possible to increase the score (eg, k / (number of characters)).

  The degree of discrimination indicates the degree of whether a word can be used for discrimination of a certain document type C. If a certain word is a word that seems to be a document type C or a word that does not seem to be a document type C, the degree of discrimination is high. For example, the calculation is performed using an expression such as tf (t) * log (df (t | C) / df (t)) * score_pos (t). Here, tf (t) is the word frequency of the word t in the document D, df (t | C) is the number of documents in which the word t appears in documents classified into the document type C, and df (t) Is the number of documents in which the word t appears, and score_pos (t) is the score of the part of speech of the word t, for example, the noun score when the word t is a noun. At this time, the score of a word having a specific part of speech can be increased by setting the score of the noun high.

  Next, the highlight location specifying unit 102 will be described.

  The highlight location specifying unit 102 is a module that searches each document for a location where a browsing word candidate selected by the browsing word extraction unit 101 appears in the document and selects a location to be highlighted in the document. is there.

  FIG. 11 shows an example of the processing procedure of the highlight location specifying unit 104.

  Here, L is the upper limit of the number of highlight words, N is the number of documents, M is the total number of highlight word candidates, and X is the number of words stored as highlight words in document i.

  First, since there is no browsing when there are many highlights, an upper limit L of the number of highlighted words is set.

  In step S <b> 21, the highlight location specifying unit 104 reads the document data from the document data storage unit 105 and also reads the word list and score output by the browsing word extraction unit 101.

  In step S22, 1 is substituted for i.

  In steps S23 to S25, it is confirmed whether each browsing word candidate appears for each document.

  However, the total number of documents in the document data storage unit 105 is N, and the total number of read word candidates is M.

  If the word j appears p times in the document i, the word position to be highlighted is selected from the p words j in step S27. For example, selecting a first appearance, selecting by neighborhood word importance, selecting a word not included in the compound word, and the like. Here, the neighborhood word importance refers to the importance of a word that appears in the vicinity of the word t. Since the highlights are eye-catching, those who have important words in the vicinity also have a higher score. The important words can be calculated using a calculation formula generally known as keyword extraction. For example, the neighboring word importance (ti) is scored using the importance of the neighboring word tj and the distance to ti, and Σj {tf (tj) * log (N / df (tj)) * (1 / | J−i |)}. Here, N represents the total number of documents, and j−i represents the distance between the words ti and tj. In addition, when a single word is included in a compound word, the score may be lowered because only one word may not make sense.

  In step S28, the word to be highlighted with respect to the document i and its appearance position are stored, and the same operation is repeated while the number of highlighted words X is not more than the upper limit L of the number of highlighted words.

  If the condition match is No in step S25, the process returns to step S23, proceeds to the next document i + 1, and repeats the same operation for steps S24 to S28.

  When the highlight location has been specified for all the documents read in step S21, the highlighted word and its position of each document are output in step S29.

  Next, the text display unit 103 will be described.

  The text display unit 103 is a module that highlights and displays the highlighted word position specified by the highlighted part specifying unit 102 for each document data in the document data storage unit 105.

  FIG. 12 shows an example of the processing procedure of the text display unit 103.

  In step S <b> 31, the text display unit 103 reads the document data in the document data storage unit 105, and also reads the highlight word output from the highlight location specifying unit 102 and its position.

  In step S32, the highlighted portion in the document data is highlighted and each document data is output.

  FIG. 5 is an example of highlighting of each document data.

  FIG. 13 shows another highlight example of each document data. FIG. 13 is a display example of another document when a classification is assigned to the first three documents in the document group illustrated in FIG. Further, as in FIG. 5, an example is shown in which the required document type is highlighted with a cross hatched frame, the unnecessary document type is highlighted with a hatched hatched frame, and the unread document type is highlighted with a frame without hatching.

  Now, after the highlight display as described above, the user can further assign a classification to an unclassified document.

  The user input unit 104 is a module that receives a classification given by the user to each document displayed on the text display unit 103. The user assigns a classification to each document in FIG. 5 (example: “◯” “×”). At this time, the user input unit 104 receives this classification information. When the user input unit 104 receives the classification assigned by the user, the processing of 101 to 103 is subsequently performed to update the highlight word.

  Next, a specific example of processing will be shown.

  Consider a case in which there are six documents 1201 to 1206 as shown in FIG. 13 and classification is given to the first three documents 1201 to 1203. In the classification information storage unit 106, “x” is given to the document 1201, “x” is given to the document 1202, and “◯” is given to the document 1203.

  Based on the three documents and the classification, the browsing word extraction unit 101 performs word extraction. In score_pos of equation (1), if the noun is 1 and the others are 0, the noun is extracted from the case document. From the document 1201, nouns such as “document”, “speech”, “instruction”, “input”, “frequency”, “situation”, “synthesis”, “sales”, “style”, “output”, and “device” are extracted. Similarly, nouns are extracted from the documents 1202 and 1203. Then, the expression (1) is calculated for each extracted noun. In this example, since the number of documents is small and the score hardly changes between words, it becomes a browsing word candidate corresponding to the classification of the document in which each noun appears. The nouns that appear in the three documents in the document 1204 are “operation”, “voice”, and “output”. Among these, since “output” appears only in compound words in the document 1201, it is assumed that the score is lowered and “operation” and “voice” are highlighted.

  Subsequently, a highlighted word in the document 1204 is selected by the highlight location specifying unit. Since “operation” appears only once, it is selected as it is. In addition, “voice” appears in two places, “voice output processing method” and “voice output”. In this case, for example, since both are compound words, the shorter “speech output” “speech” is highlighted, and the second “speech” is taken as the highlighted part.

  In the above description, Japanese has been described as a specific example. However, the present embodiment can be similarly applied to other languages.

  According to the present embodiment, in a case where it is desired to comprehensively investigate the contents of a large number of documents without omissions, such as patent research, literature research, market research, etc., the user has a large number of documents (for example, the document text or its summary). While reading some of the documents, for example, the classification of the documents that should be read and confirmed or the documents that should not be read and the documents that are not, and the classification to divide the person in charge to share and give, etc. Based on a document with classification, for example, it extracts words suitable for browsing for classification judgment from unclassified documents and keywords that are likely to serve as a basis for discriminating unclassified documents. By presenting with highlight display or marker addition, it is possible to assist the user in efficiently performing browsing performed by the user himself / herself.

(Second Embodiment)
In the second embodiment, when a classification is already given in the first language, a keyword for browsing is highlighted in a document in a second language different from the first language using the classification. An example is shown.

  For example, the first language may be the user's native language, the second language may be the foreign language, and conversely, the second language may be the user's native language and the first language may be the foreign language.

  Here, the case where the first language is English and the second language is Japanese will be described as an example.

  FIG. 14 shows a functional configuration example of the support system according to the second embodiment.

  As shown in FIG. 14, the browsing support system includes a browsing word extraction unit 101, a replacement unit 201, a bilingual dictionary 202, a highlight location specifying unit 102, a text display unit 103, a user input unit 104, a first A language document data storage unit 203, a first language classification information storage unit 204, a second language document data storage unit 205, and a second language classification information storage unit 206 are provided.

  The first language document data storage unit 203 stores document data in the first language. In addition, a classification corresponding to each document is already assigned, and the classification information is stored in the first language classification information storage unit 204.

  The browsing word extraction unit 101 basically reads the information in the first language document data storage unit 203 and the first language classification information storage unit 204, and performs browsing word extraction in the same manner as in the first embodiment. Since the contents of the first language classification information storage unit 204 do not change, the extraction here needs to be performed only once.

  On the other hand, the second language document data storage unit 205 corresponds to the document data storage unit 105 of the first embodiment, and the second language classification information storage unit 206 corresponds to the classification information storage unit 106 of the first embodiment. To do. The browsing word extraction unit 101 repeatedly performs the same processing as in the first embodiment for the second language document data storage unit 205 and the second language classification information storage unit 206.

  The browsing word candidate and the score output by the browsing word extraction unit 101 are input to the replacement unit 201.

  The replacement unit 201 searches the first language document data storage unit 203 and the first language classification information storage unit 204 for the first language word extracted by the read word extraction unit 101 (in the second language on the document data). This is a module for replacing with words in the second language (to make correspondence with words).

  The replacement unit 201 searches for a translation word into the second language using the bilingual dictionary 202 from the browsing word candidates described in the first language, and creates a translation word in the second language. At this time, when there are a plurality of translated words in the second language, or when there are other words in the first language that become translated words in the second language, it is ambiguous if translated from the first language to the second language. In such a case, priority is given to other words having no ambiguity by, for example, lowering the reading word score of this word.

  A series of processes of the highlight location specifying unit 102, the text display unit 103, and the user input unit 104 are basically the same as those in the first embodiment. However, the document data read by the highlight location specifying unit 102 and the text display unit 103 are the second language document data storage unit 205, and the classification information given by the user input unit 104 is stored in the second language classification information storage unit. 206.

  In the second embodiment, even in a situation where the classification is not yet given to the second language document data that is the document data that the user wants to classify, the first language data that the user has already given the classification, that is, the first language document The browsing word extraction unit 101 can perform processing from the data storage unit 203 and the first language classification information storage unit 204.

  In addition, after the data is added to the second language classification information storage unit 206, the processing of the word extraction unit 101 may be performed by browsing the data in the second language document data storage unit 205 and the second language classification information storage unit 206. it can.

  In the latter case, the replacement unit 201 is not required (switching to the configuration of FIG. 1).

  FIG. 15 shows an example of the processing procedure of the browsing word extraction unit 101.

  When processing is performed in an initial state in which no classification is assigned to the document by the user (Yes in step S41), the browsing word extraction unit 101 includes the first language document data storage unit 203 and the first language. A scored word list in the first language is created using the information in the classification information storage unit 204 (step S42). FIG. 16 shows an example of the scored word list. In the scored word list of FIG. 16A, the replacement unit 201 and the bilingual dictionary 202 replace the words from the first language to the second language. An example is shown in FIG. In addition, s1 has shown the score value.

  When processing is performed after classification is given to at least one document by the user (No in step S41), the browsing word extraction unit 101 stores the second language document data storage unit 205 and the second language classification information storage. A scored word list of the second language is created using the information of the unit 206 (step S43). In this specific example, for example, as shown in FIG. Then, the scored word list in the first language already created in step S42 (word replacement) and the scored word list in the second language created in step S43 are merged.

  When a predetermined condition is satisfied, only the second word scored word list may be used without performing step S43.

  The predetermined condition is, for example, a value in which c / N is predetermined, where N is the total number of documents in the second language, and c is the number of documents to which the classification type is assigned by the user to the documents in the second language. When c exceeds a predetermined value, or when a predetermined time has elapsed since the classification type was first assigned to the second language document by the user, etc. Various things are possible.

(Modification 1 of 2nd Embodiment)
In the above description, the first language can be the same as the second language. In this case, the replacement unit 201 and the bilingual dictionary 202 are not necessary.

(Modification 2 of the second embodiment)
In the second modification of the second embodiment, an example is shown in which a word is translated and displayed in the first language instead of highlighting the word in the document in the second language in the second embodiment.

  A functional configuration example of the modification of the second embodiment may be the same as that of FIG.

  However, the output from the replacement unit 201 is not only a translation of the first language word output from the browsing word extraction unit 101 into the second language, but also a set of translations into the first language and the second language. It is passed to the write location specifying unit 102.

  The processing of the highlight location specifying unit 102 performs the same processing as in the second embodiment, but the output at that time is passed to the text display unit 103 as a set of translation into the first language and the second language. .

  The processing of the text display unit 103 is different from that of the second embodiment.

  FIG. 17 shows an example of the processing procedure of the text display unit 103.

  In step S <b> 51, the text display unit 103 inputs the second language document data, the highlighted word that is the output of the highlighted part specifying unit 102, the translated word of the word into the first language, and the highlighted word position.

  In step S52, the second language document data is displayed, and the highlighted word is converted into a translated word into the first language and displayed in the sentence.

  FIG. 18 illustrates an example of a word list with scores according to the second modification of the second embodiment.

  An example of a scored word list in the first language generated by the browsing word extraction unit 101 is shown in (a). An example in which a word in the first language is replaced with a second language and this is further added to the scored word list is shown in (b). In addition, e1 has shown the score value.

  An example of the second language scored word list generated by the browsing word extraction unit 101 is shown in FIG. An example in which a word in the second language is replaced with the first language and this is further added to the scored word list is shown in (d). J1 indicates a score value.

  FIG. 19 shows an example in which words extracted for each classification type are highlighted in the first language (English) when a document in the second language (Japanese) is displayed. Further, in FIG. 19, the necessary document type is a frame, the unnecessary document type is a hatched hatched frame, and further highlighting is performed, and the unread document type is only highlighted in the first language (English). Show.

(Third embodiment)
In the third embodiment, an example in which the user himself selects a word used for browsing is shown.

  FIG. 20 shows a functional configuration example of the support system according to the third embodiment.

  FIG. 20 shows an example in which the user first inputs some words used for browsing and starts an initial operation based on the words.

  The word input unit 301 receives an input of a word used for browsing by the user. In this case, a word input by the user is directly read and used as a word candidate, and a highlight location is specified by the highlight location specifying unit 102 based on the candidate word. Other processes are the same as those in the first embodiment. In addition, after a user input by the user input unit 104, the word input is performed when a series of processes of the browsing word extraction unit 101, the highlight location specifying unit 102, and the text display unit 103 are performed in accordance with the update of the classification information storage unit 106. A new word may be input from the unit 301.

(Fourth embodiment)
In the fourth embodiment, an example in which the user himself selects a word used for browsing is shown.

  FIG. 21 shows a functional configuration example of the support system according to the second embodiment.

  FIG. 21 shows an example in which the user selects a word that he / she wants to use for browsing from the browsing word candidates calculated by the browsing word extraction unit 101.

  The processes of the browsing word extraction unit 101, the highlight location specifying unit 102, the text display unit 103, and the user input unit 104 are the same as those in the first embodiment.

  The browsing word selection unit 401 presents the browsing word candidates extracted by the browsing word extraction unit 101 to the user. The user selects a browsing word that the user wants to use or does not want to use, and the browsing word selection unit 401 accepts the selection. As a result of the selection, the priority of the word that the user wants to use as a browsing word is made high, and it is input to the highlight location specifying unit 102 as a word list with scores.

  It should be noted that the embodiments and modifications described so far can be implemented in any combination.

  In addition, the present embodiment (any of the embodiments and modifications described so far, or any combination thereof) can assist in browsing a large number of documents.

The instructions shown in the processing procedure shown in the above embodiment can be executed based on a program that is software. A general-purpose computer system stores this program in advance and reads this program, so that the same effect as that obtained by the browsing support system of the above-described embodiment can be obtained. The instructions described in the above-described embodiments are, as programs that can be executed by a computer, magnetic disks (flexible disks, hard disks, etc.), optical disks (CD-ROM, CD-R, CD-RW, DVD-ROM, DVD). ± R, DVD ± RW, etc.), semiconductor memory, or a similar recording medium. As long as the recording medium is readable by the computer or the embedded system, the storage format may be any form. If the computer reads the program from the recording medium and causes the CPU to execute instructions described in the program based on the program, the same operation as the browsing support system of the above-described embodiment can be realized. Of course, when the computer acquires or reads the program, it may be acquired or read through a network.
In addition, the OS (operating system), database management software, MW (middleware) such as a network, etc. running on the computer based on the instructions of the program installed in the computer or embedded system from the recording medium implement this embodiment. A part of each process for performing may be executed.
Furthermore, the recording medium in the present embodiment is not limited to a medium independent of a computer or an embedded system, but also includes a recording medium in which a program transmitted via a LAN, the Internet, or the like is downloaded and stored or temporarily stored.
Further, the number of recording media is not limited to one, and when the processing in this embodiment is executed from a plurality of media, it is included in the recording medium in this embodiment, and the configuration of the media may be any configuration.

The computer or the embedded system in the present embodiment is for executing each process in the present embodiment based on a program stored in a recording medium. The computer or the embedded system includes a single device such as a personal computer or a microcomputer. The system may be any configuration such as a system connected to the network.
In addition, the computer in this embodiment is not limited to a personal computer, but includes an arithmetic processing device, a microcomputer, and the like included in an information processing device, and is a generic term for devices and devices that can realize the functions in this embodiment by a program. ing.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 101 ... Browsing word extraction part, 102 ... Highlight location specific part, 103 ... Text display part, 104 ... User input part, 105 ... Document data storage part, 106 ... Classification information storage part, 201 ... Replacement part, 202 ... Two languages Inter-dictionary, 203 ... first language document data storage unit, 204 ... first language classification information storage unit, 205 ... second language document data storage unit, 206 ... second language classification information storage unit, 301 ... word input unit, 401 … Browsing word selection part.

Claims (16)

A document storage unit for storing a plurality of documents associated with identification information;
A display unit for displaying all or a part of the plurality of documents with or without highlighting of words or phrases;
An input unit for inputting from a user an instruction of a specific document of the displayed document and an instruction of a specific classification type to be given to the specific document among a plurality of predetermined classification types;
A classification information storage unit that stores classification information in which the identification information is associated with the specific classification type;
An extraction unit that extracts one or more words or phrases that should be highlighted for the classification type from one or more of the documents associated with the same classification type;
A specifying unit that specifies a portion to be highlighted in the document when each word or phrase extracted by the extraction unit exists in the document for each of all or a part of the document A browsing support system characterized by that.   The browsing support system according to claim 1, wherein the specifying unit performs the specification for each of the documents other than the specific document.   The display unit displays all or a part of the plurality of documents without the highlight display before the place to be highlighted is specified, and the part to be highlighted is displayed. After the word is specified, when displaying all or a part of the plurality of documents, the word or phrase in which the portion to be highlighted in the document is specified is selected as the highlight corresponding to the classification type. The browsing support system according to claim 1, wherein highlight display is performed in a light form.   The said extraction part performs the said extraction by the said extraction part, the said specification by the said specific part, and the update of the display content by the said display part triggered by the content of the said classification information having been updated. 1. A browsing support system according to 1.   The extraction unit performs the extraction by the extraction unit, the specification by the specification unit, and the update of display content by the display unit when an update instruction is input from a user. 1. A browsing support system according to 1.   The display unit displays, together with the information indicating the classification type assigned to the document, for each of the specific documents when displaying all or part of the plurality of documents. The browsing support system according to claim 1. The document stored in the document storage unit is described in a specific language,
The browsing support system is
A different language document storage unit for storing a plurality of documents described in a language different from the specific language associated with the identification information;
A different language classification information storage unit that stores classification information in which the identification information is associated with the classification type given in advance by the user for all documents stored in the different language document storage unit;
The extraction unit includes the one or more specific languages to be highlighted based on the document stored in the different language document storage unit and the classification information stored in the different language classification information storage unit. Also extracts words or phrases in different languages,
The browsing support system is
The browsing support system according to claim 1, further comprising a replacement unit that replaces words or phrases in a language different from the specific language with the specific language.   The extraction unit is extracted based on the document stored in the different language document storage unit and the classification information stored in the different language classification information storage unit until a predetermined condition is satisfied, and is extracted by the replacement unit. Extracted based on only the word or phrase replaced with the specific language, or only the classification information stored in the different language classification information storage unit and the document stored in the different language document storage unit The specific language extracted based on the word or phrase replaced by the specific language by the replacement unit, the document stored in the document storage unit, and the classification information stored in the classification information storage unit As a result of the extraction, the merged word or phrase according to is used as a result of the extraction, and after a predetermined condition is established, the document stored in the document storage unit and the document Browsing support system according to claim 7 in which only the word or phrase extracted the specific language on the basis of the classification information stored in the information storage unit, characterized in that the result of the extraction. The replacement unit refers to a bilingual dictionary in which correspondence between the specific language and a language different from the specific language is registered for a plurality of words or phrases, and performs the replacement.
The extraction unit preferentially extracts words or phrases in which correspondence between the specific language and a language different from the specific language is one-to-one correspondence in the bilingual dictionary. 8. A browsing support system according to 8.   The browsing support system according to claim 1, wherein the classification type indicated through the input unit includes at least a classification type related to a conforming document or a classification type related to a nonconforming document.   The browsing support system according to claim 1, wherein the classification type instructed via the input unit includes each classification type related to a document assigned to each of at least a plurality of persons in charge. A word input unit for inputting a word or phrase from a user;
The browsing support system according to claim 1, wherein the word or phrase input via the word input unit is used in addition to or instead of the word or phrase extracted by the extraction unit. . The extraction unit extracts candidates for the word or phrase,
The browsing support system further includes a word selection unit for displaying the word or phrase candidates and inputting a selection of the word or phrase from the user among the displayed candidates.
The said extraction part makes it easier to select the said word or phrase selected via the said word selection part, when selecting the candidate of the said word or phrase. Browsing support system.   The browsing support system according to claim 1, wherein the document is a target document text or a summary thereof. A browsing support method of a browsing support system including a document storage unit for storing a plurality of documents associated with identification information,
Displaying all or part of the plurality of documents with or without highlighting of words or phrases;
An instruction from a specific document among the displayed documents and a specific classification type instruction to be given to the specific document among a plurality of predetermined classification types are input from the user,
Storing classification information in which the identification information is associated with the specific classification type;
Extracting one or more words or phrases to be highlighted for the classification type from one or more of the documents associated with the same classification type;
For each of all or a part of the document, when each word or phrase extracted by the extraction unit is present in the document, the location to be highlighted in the document is specified. Browsing support method. A program for causing a computer to function as a browsing support system including a document storage unit for storing a plurality of documents associated with identification information,
A document storage unit for storing a plurality of documents associated with identification information;
A display unit for displaying all or a part of the plurality of documents with or without highlighting of words or phrases;
An input unit for inputting from a user an instruction of a specific document of the displayed document and an instruction of a specific classification type to be given to the specific document among a plurality of predetermined classification types;
A classification information storage unit that stores classification information in which the identification information is associated with the specific classification type;
An extraction unit that extracts one or more words or phrases that should be highlighted for the classification type from one or more of the documents associated with the same classification type;
For each of all or part of the document, when each word or phrase extracted by the extraction unit exists in the document, a computer is used as a specifying unit that specifies a place to be highlighted in the document. A program to make it work.

Images