JP2019012382A - Document search apparatus, word presentation apparatus, and method and program thereof - Google Patents

Abstract

To provide a document search technique capable of retrieving documents such as picture books in consideration of not only some words but also words related to those words.SOLUTION: A document search apparatus comprises: a storage unit 1 for storing information on each word and an acquisition timing of each word as the information about the acquisition timing of the word which is the information about an acquisition timing that an infant acquires each word; a word selecting unit 21 for selecting, from the storage unit 1, a word which has the information on the acquisition timing of the word close to the information about the acquisition timing of the input word; and a document search unit 3 for retrieving a document using the input word and the selected word as a keyword.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for searching a document and a word presentation technique used for searching a document, for example.

  As a technique for searching for books, a search technique based on bibliographic information is known (see, for example, Non-Patent Document 1). In the search technology based on bibliographic information, bibliographic information such as author name and title for each book is registered in a database in advance. When a search keyword is input, a book having bibliographic information including the search keyword is searched with reference to the database. Then, the book found by the search is presented to the user.

National Diet Library, [Search May 16, 2017], Internet <URL: http://www.ndl.go.jp/>

  Until now, in order to teach a certain word to infants, etc., for example, using the book search technology of Non-Patent Document 1, search for a picture book, etc. using that word as a keyword, and read the picture book found by the search to the infant, etc. By telling them, teaching a certain word to infants was performed.

  By the way, it is known that when a certain word is slowly remembered, the words related to the word also tend to be remembered slowly. For this reason, when a language hearing person teaches a certain word to an infant or the like, it is expected that it is more effective to teach a word related to the certain word at the same time.

  However, no technology has been proposed that can search for textbooks such as picture books in consideration of not only a certain word but also a word related to the certain word.

  An object of the present invention is to provide a document search apparatus that can perform a document search in consideration of not only an input word but also a word related to the input word, for example, a word presentation apparatus that can be used in the document search apparatus. It is to provide these methods and programs.

  According to the word presentation device according to the aspect of the present invention, each word and information related to the acquisition time of each word are stored as information related to the acquisition time when the infant acquires the word. , A word selection unit for selecting a word related to the input word from the storage unit, information on the acquisition time of the input word and information on the acquisition time of the selected word And a weighting unit that assigns a weight that is the importance of the word to each of the selected words, and is assigned to each of the input word, the selected word, the input word, and the selected word A word presentation unit that outputs weights.

  According to an aspect of the present invention, there is provided a document search device that performs the above-described word presentation device and a document search using the input word and the selected word output by the word presentation device as keywords, and the input word and the selected word. And a document search unit that emphasizes a word having a greater weight.

  A document such as a picture book can be searched in consideration of not only a certain word but also a word related to the certain word. Or, a word related to a certain word as well as a certain word can be presented.

The block diagram for demonstrating the example of a word presentation apparatus and a document search apparatus. The flowchart for demonstrating the example of the word presentation method and a document search method. The figure which shows the example of the understanding curve and utterance curve of the word one one. The figure for demonstrating a web diary.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  As shown in FIG. 1, the document search apparatus includes a storage unit 1, a word presentation unit 2, and a document search unit 3, for example. The document search display method is realized, for example, when each unit of the document search apparatus executes the processing from step S21 to step S3 described later with reference to FIG.

  The word presentation device includes a storage unit 1 and a word presentation unit 2. A word presentation method is implement | achieved, for example, when each part of a word presentation apparatus performs the process of step S21 to step S22 demonstrated later by FIG.

<Storage unit 1>
In the infant vocabulary acquisition DB of the storage unit 1, a set of each word and information related to the acquisition time of each word is stored in association with an ID for identifying each word.

  The information regarding the acquisition time of a word is information regarding the acquisition time when the infant acquires the word. Here, acquiring a word means that the word can be understood or the word can be spoken.

  An example of the information related to the word acquisition time is at least one of information related to the word understanding curve and information related to the word utterance curve described below.

  The understanding curve of a certain word is a curve representing the proportion of infants who understand the certain word with respect to the age of the infant. The comprehension curve is created based on information inputted by the caregiver in the vocabulary checklist survey or information input by the web diary described later. Comprehension curves are calculated by categorizing the answers answered by determining whether the infant can understand the corresponding word (even if it cannot speak) at the time of the survey. For example, as shown in FIG. Is the age of the infant, and the vertical axis is the percentage of infants who can understand the corresponding word, plotted as Δ, and approximated by a logistic curve. FIG. 3 is a diagram illustrating an example of an understanding curve and an utterance curve of the word “One One”.

  The utterance curve of a certain word is a curve representing the percentage of infants who can utter a certain word with respect to the infant's age. The utterance curve is also created on the basis of information that the nurturer answered in the vocabulary checklist survey or information input by the web diary described later. The utterance curve is obtained by calculating whether the infant is able to utter the corresponding word at the time of the survey, for each age, and for example, as shown in FIG. The percentage of infants who can utter the corresponding word is taken as ◯ and plotted with a logistic curve.

The logistic function is expressed by the following formula (1). In equation (1), x represents the age of the month, f (x) represents the acquisition rate, and e represents the number of Napiers. a, b, and c are parameters that determine the shape of the logistic curve.
f (x) = (ae ^{(cx + b)} ) / (1 + e ^{(cx + b)} )… (1)

  If a, b, and c are determined, the shape of the logistic curve is determined. For this reason, for example, parameters a, b, and c of a logistic curve representing an understanding curve of a certain word are information on the understanding curve of the word. Further, for example, parameters a, b, and c of a logistic curve representing an utterance curve of a certain word are information on the utterance curve of the word.

  Another example of information on the acquisition time of a word is X% acquisition age at which X% of an infant acquires the word, where X is a predetermined number of 0 ≦ X ≦ 100. X is, for example, 50.

For example, the X% acquired age of a word is the age at which X% of infants understand the word. In this case, the X% acquired age of a word can be obtained from the understanding curve of the word. In other words, given the parameters a, b, and c of the comprehension curve of the word, x that satisfies X = (ae ^{(cx + b)} ) / (1 + e ^{(cx + b)} ) is obtained. The found x is the age at which X% of the word is acquired.

Also, the X% acquisition age of a word may be the age at which X% of the infant speaks the word. In this case, the X% acquired age of a word can be obtained from the utterance curve of the word. That is, given the utterance curve parameters a, b, and c of the word, x that satisfies X = (ae ^{(cx + b)} ) / (1 + e ^{(cx + b)} ) is obtained. The found x is the age at which X% of the word is acquired.

  Thus, since the X% acquisition age of a word can be obtained from the understanding curve or utterance curve of the word, information on the word understanding curve and information on the utterance curve of the word are substituted for the X% acquisition age of the word. May be stored in the storage unit 1. In this case, the weight assigning unit 22 refers to the storage unit 1 to obtain the X% acquisition age of the word from at least one of the information related to the word understanding curve and the information related to the word utterance curve, and acquires the obtained X% of the word The process described later is performed as the age of X% acquired month of the word read from the storage unit 1.

  Of course, in addition to at least one of the information related to the word understanding curve and the information related to the word utterance curve, the X% acquired age of the word may be stored in the storage unit 1. As a result, the calculation cost for obtaining the X% acquired age of the word can be reduced.

  The storage unit 1 may store a category of each word.

  Here, the word category means not a general part of speech such as “noun”, “verb”, “adjective”, but a word category classified according to the development of the child. Examples of word categories include categories such as specific animals, food, and parts of the body, categories of nouns and adjectives that represent time and space (hereinafter referred to as “spatial categories”), family members and relationships. At least one of a category of a noun or an adjective to be shown, a category of a verb or adjective representing an action, a category of a regular expression such as a daily greeting, or a category such as a color name of a number

  For example, it is assumed that the words “up”, “done”, “yoko”, “close”, and “far” are included in the space category. In this case, information indicating that the category of each word “up”, “do”, “yoko”, “near”, “far” is a spatial category is stored in the storage unit 1.

  In addition, when a category has an inclusion relationship, the category inclusion relationship may be stored in the storage unit 1. For example, if the animal category includes a carnivore category and a herbivore category, information to that effect is stored in the storage unit 1.

<Word selection unit 21>
A word is input to the word selection unit 21 of the word presentation unit 2.

  The word selection unit 21 selects a word related to the input word from the storage unit 1 (step S21). Information regarding the selected word is output to the weighting unit 22. The word selection unit 21 selects at least one word.

  Hereinafter, an example of word selection by the word selection unit 21 will be described.

<< Word selection example 1 >>
In example 1 of word selection, the word selection unit 21 selects words in the same category as the input word.

  Specifically, the word selection unit 21 obtains the category of the input word by referring to the storage unit 1. Then, the word selection unit 21 selects words in the same category as the obtained category from the storage unit 1.

  For example, when the word “up” is input, the word selection unit 21 refers to the storage unit 1 to obtain information that the category of the word “up” is a spatial category. Then, the word selection unit 21 selects the words “do”, “yoko”, “close”, and “far” included in the space category other than the word “up”.

<< Example 2 of word selection >>
In the example 2 of word selection, the word selection unit 21 selects words included in a parent category that is separated by a predetermined hierarchical distance D based on the input word category. D is a predetermined positive integer. The parent category that is separated by a predetermined hierarchical distance D with reference to a certain category is the D parent categories on the basis of the certain category. Hierarchy distance is the number of times a hierarchy needs to be moved from one category to another. The hierarchical distance is defined as a parent in a so-called family relationship, for example.

  For example, D = 1, the information that the animal category includes the carnivore category and the herbivore category is stored in the storage unit 1, and the input word is “lion” It is assumed that information indicating that the category of “lion” is a carnivore is stored in the storage unit 1. In this case, the word selection unit 21 refers to the storage unit 1 to obtain information that the category of the word “lion” is a carnivore category. Then, the word selection unit 21 obtains information that the parent category separated by the hierarchical distance 1 with respect to the carnivore category is an animal category. Then, the word selection unit 21 selects a word included in the parent category other than the word “lion”. More specifically, the word selection unit 21 selects words other than the word lion, which are words included in a lower category such as a carnivore and a herbivore belonging to an animal category and an animal category.

<< Word selection example 3 >>
In the example 3 of word selection, the word selection unit 21 selects a word having information on the acquisition time of the word that is close to the information on the input time of the word.

  When the information about the acquisition time of the word is the X% acquisition age at which X% of the infant described later acquires the word, the word selection unit 21 acquires X% near the X% acquisition age of the input word. Select a word with age. For example, the word selection unit 21 selects a word whose difference between the X% acquired age of the word and the X% acquired age of the input word is equal to or less than a predetermined value.

  When the information about the word acquisition time is at least one of the information about the understanding curve of the word and the information about the utterance curve, which will be described later, the word selection unit 21 sets the input understanding curve or utterance curve of the word. Select a word with a close comprehension curve or utterance curve. For example, the word selection unit 21 selects a word whose similarity is less than or equal to a predetermined value with respect to the word understanding curve or utterance curve and the input word understanding curve or utterance curve. As the similarity, the Batacharia distance of Expression (3) described later can be used.

<Weighting unit 22>
Information related to the word input to the word presentation unit 2 and the word selected by the word selection unit 21 is input to the weighting unit 22.

  The word assigning unit 22 outputs the words input to the word presenting unit 2 and the words selected by the word selecting unit 21 and the respective weights of these words by the following processing.

  The weighting unit 22 uses the information about the input time of the word and the information about the time of acquisition of the word selected by the word selection unit 21 to determine the input word and the word selected by the word selection unit 21. A weight that is the importance of the word is assigned to each (step S22).

  For example, the weight assigning unit 22 assigns a greater weight to a word having information related to a word acquisition time that is closer to the information related to the input word acquisition time.

  Hereinafter, an example of weighting by the weighting unit 22 will be described.

<< Example 1 for weighting >>
In weighting example 1, assuming that the information related to the word acquisition time is the X% acquisition age at which X% of the infant acquires the word, the weighting unit 22 is close to the information related to the input word acquisition time. A greater weight is given to a word having information on the acquisition time of the word.

Specifically, the weight assigning unit 22 reads from the storage unit 1 the X% acquisition age of the input word and the X% acquisition age of the selected word. Then, the weight assigning unit 22 obtains a predetermined function f ₁ (μ, μ) where X is the X% acquisition age of the input word obtained by reading, and AoA is the X% acquisition age of the word whose weight is to be obtained. The output value when these X% acquired ages are input to AoA) is calculated, and the calculated output value is set as a weight to be given to the word whose weight is to be obtained. The weight assigning unit 22 performs a process of obtaining a weight using the predetermined function f ₁ (μ, AoA), with each of the input word and the word selected by the word selecting unit 21 as a word whose weight is to be obtained. Do.

f ₁ (μ, AoA) is a function that outputs a larger value as the absolute value of the difference between μ and AoA is smaller. An example of f ₁ (μ, AoA) is a function weight (μ, AoA) of the following equation (2). In equation (2), π is the pi and σ is a predetermined positive number. For example, σ = 3. As in this example, f ₁ (μ, AoA) may be defined using a Gaussian probability density function.

  For example, it is assumed that the input word is the word “grandma” and the word selection unit 21 selects the words “mama”, “dad”, “baoba”, and “older brother”. Also, assuming X = 50, the X% acquisition age of the word “Mama” is 9.87, the X% acquisition age of the word “papa” is 10.9, and the X% acquisition age of the word “Baba” is 13.1, and the word “grandma” has an X% acquisition age of 16.9, and the word “older brother” has an X% acquisition age of 21.9.

  In this case, the weight assigning unit 22 inputs the word “grandma” and “mama”, “dad”, “baoba”, “older brother” selected by the word selection unit 21 based on the above formula (2). The weight of the word is calculated. As a result, for example, the weight assigning unit 22 assigns a weight of 0.0085 to the word “Mama”, assigns a weight of 0.0180 to the word “daddy”, and assigns a weight of 0.0591 to the word “Baba”. A weight of 0.133 is given to the word “grandma”, and a weight of 0.033 is given to the word “older brother”.

  In this way, the weight assigning unit 22 assigns a greater weight to a word that is closer to the X% acquired age of the word for which the X% acquired age of the word is input, for example, it is already remembered or likely to be remembered. The weight of the word is low, the weight of the word that is likely to be remembered in the word that is not yet remembered or that is likely to be remembered is increased, and the weight of the word that is likely to take a while to be remembered can be adjusted to be slightly lower.

<< Example 2 of weighting >>
In the weighting example 2, the weighting unit 22 has a weight of a word whose X% acquisition month is earlier than the X% acquisition month of the input word, and is X% higher than the X% acquisition month of the input word. The weight given by the weight assigning unit 22 for all or some of the words is corrected so that the weight of the word having a late acquisition age becomes larger, and the weight after correction is substituted for the weight given by the weight assigning unit 22 Are given as final weights to all or some of the words.

  For example, first, the weight assigning unit 22 assigns a weight to each of the input word and the word selected by the word selecting unit 21 by, for example, the method of weight assignment example 1. Then, the weight assigning unit 22 applies a predetermined attenuation coefficient R (0 ≦ R) to the weight w assigned by the method of Example 1 only for a word whose X% acquisition age is earlier than the X% acquisition age of the input word. A value w × R multiplied by <1) is calculated, and the calculated value w × R is used as a weight of the word instead of w. For example, R = 0.5.

  For example, the input word is the word “grandma”, and the word selection unit 21 selects the words “mama”, “dad”, “baoba”, and “older brother”. For example, the weight assigning unit 22 assigns a weight of 0.0085 to the word “Mama”, assigns a weight of 0.0180 to the word “Daddy”, assigns a weight of 0.0591 to the word “Baba”, and reads “Grandma”. Assume that a weight of 0.133 is given to the word and a weight of 0.033 is given to the word “Older brother”. Similarly to weighting example 1, if X = 50, the word “Mama” has an X% acquisition age of 9.87, the word “papa” has an X% acquisition age of 10.9, The X% acquisition age of the word "13.1", the X% acquisition age of the word "Grandma" is 16.9, and the X% acquisition age of the word "Older" is 21.9.

  In this case, the words whose age of X% acquisition is earlier than that of the input word “grandma” is “Mama”, “Daddy”, and “Baba”. For this reason, the weight assigning unit 22 calculates a value 0.0085 × R obtained by multiplying the weight 0.0085 given by the weight assigning unit 22 for the word “Mama” by the attenuation coefficient R, and replaces 0.0085 with this calculated value 0.0085. Let xR be the weight of the word “Mama”.

  The weight assigning unit 22 calculates a value 0.0180 × R obtained by multiplying the weight 0.0180 given by the weight assigning unit 22 for the word “daddy” by the attenuation coefficient R, and replaces the 0.0180 with the calculated value 0.0180 × R. Let R be the weight of the word “daddy”.

  Further, the weight assigning unit 22 calculates a value 0.0591 × R obtained by multiplying the weight 0.0591 given by the weight assigning unit 22 of the word “Baba” by an attenuation coefficient R, and replaces 0.0591 with this calculated value. Let 0.0591 × R be the weight of the word “baoba”.

It is a function that outputs a value of 0 or more and less than 1 with X% acquisition age AoA _input of an _input word, X% acquisition age of a word as AoA, and f ₂ (AoA _input , AoA), When AoA _input > AoA, the function outputs a smaller value than when AoA _input ≤AoA. For example, let w be the weight calculated by the method of Example 1 above for a certain word. In this case, the weight assigning unit 22 may use w × f ₂ (AoA _input , AoA) as a weight for the certain word instead of w. The weight assigning unit 22 uses the predetermined function f ₂ (AoA _input , AoA) as a word for which the input word and the word selected by the word selection unit 21 are to be obtained. May be performed.

The above function f ₂ (AoA _input , AoA) is an example. Other functions may be used as long as the weight of a word whose X% acquisition month is earlier than the X% acquisition month of the input word can be corrected.

  According to the method of weighting example 2, a word whose X% acquisition age is earlier than the input word's X% acquisition age, in other words, the weight of the word that will already be remembered is relatively lowered, and the input word It is possible to relatively increase the weight of words whose X% acquisition age is later than the X% acquisition age, in other words, words that will be remembered in the future.

<< Example 3 of weighting >>
In weighting example 3, assuming that the information related to the word acquisition time is at least one of the information related to the word understanding curve and the information related to the utterance curve, the weighting unit 22 uses the information related to the input word acquisition time as information. A greater weight is given to a word having information on the acquisition time of a near word. In other words, the weight assigning unit 22 assigns a greater weight to a word having a closer feature than some AoA itself that can be extracted from the understanding (utterance) curve of the input word. Any characteristic that can be extracted from the understanding (utterance) curve is any characteristic that can be extracted from the curve, such as the inclination of the tangent having the maximum inclination among the tangents of the curve, and the number of inflection points, in addition to the specific examples described later. Good.

Specifically, the weight assigning unit 22 reads the parameters of the understanding curve and the utterance curve of the input word and the word selected by the word selection unit 21 from the storage unit 1. Then, AR _i and AS _i are the acquisition ratios of the input curve A and the curve I of the input word A identified from the read parameters, respectively, and the age of the curve B and the utterance curve of the word B whose weight is to be obtained. With the acquisition ratio of _i as BR _i and BS _i , respectively, the weight assigning unit 22 calculates the value of weight (A, B) defined by the following equation (3), and obtains the weight using the calculated value. The weight of the word B you want. The weight assigning unit 22 performs a process of obtaining a weight using this function weight (A, B), with each of the input word and the word selected by the word selecting unit 21 as a word B for which a weight is to be obtained. In Formula (3), I is the maximum age at which the understanding curve and the utterance curve are defined.

  The above equation (3) is based on the understanding curve and utterance of the word A as the histogram A for each age of the comprehension curve and utterance curve for the word A, and the histogram B for the understanding curve for the word B and the age of each utterance curve. Represents the curve normalized by adding the Bhattacharyya coefficient to the maximum value of 1.

  In this way, by assigning a greater weight to a word having a closer understanding curve and utterance curve, it is possible to assign a weight in consideration of the whole way of remembering the infant's words.

<< Example 4 of weighting >>
In the weighting example 4, when the information about the words acquired so far is obtained by the web diary described below, for example, the weighting unit 22 is information about the words acquired so far. Is used to correct the weight.

  First, the web diary will be explained. There is a service called a web diary that records the growth of the infant's language. This service is provided, for example, by a web diary service providing server indicated by a broken line in FIG.

  FIG. 4 shows an example of a web diary interface screen provided by the web diary service providing server 4 and displayed on the terminal device 5. Figure 4 shows the growth diary of XX, “One-line diary”, “Height / weight”, “☆ Words I learned for the first time today”, “ There is an entry field such as “Sentence spoken for the first time today”.

  When a caregiver uses the terminal device 5 to enter the word “One One” and its meaning “Inu” in “☆ Words Learned for the First Time”, for example, the word “One One” in 2017 The learned information is transmitted to the web diary service providing server 4 and stored in the word storage unit 41 of the web diary service providing server 4. In this way, the caregiver sequentially fills in the web diary with the words that the infant has learned, so that the words that the infant has learned so far and the information about the age at which the infant has learned the words are the words of the web diary service providing server 4. Stored in the storage unit 41.

  In this way, the word storage unit 41 of the web diary service providing server 4 stores information about the word that the infant has learned so far and the age at which the word has been learned, in other words, information about the word that the infant has acquired so far. Is stored.

  In addition, it is assumed that the word storage unit 41 of the web diary service providing server 4 stores information about the current age of the infant.

  In the following, an explanation will be given by taking an example of processing when the word “grandma” is entered in the “☆ words learned for the first time” column of the web diary.

  When a caregiver uses the terminal device 5 to input the word “grandma” in the “☆ words learned for the first time” field of the web diary, the information about the input word “grandma” is the terminal device. 5 to the web diary service providing server 4. The web diary service providing server 4 outputs information about the inputted word “grandma” to the word selection unit 21 of the word presentation unit 2.

  The word selection unit 21 of the word presentation unit 2 reads a word related to the word “grandma” from the storage unit 1 by the process of step S21. For example, when the words “mama”, “dad”, “baoba”, and “older brother” are selected by the word selection unit 21, information on these selected words is output to the weighting unit 22. .

  In this case, first, the weight assigning unit 22 inputs the word “grandma” and the “mama” “daddy” selected by the word selecting unit 21 by the method of weight grant example 1 or weight grant example 3, for example. A weight is assigned to each of the words “baoba” and “older brother”. For example, the weight assigning unit 22 assigns a weight of 0.0085 to the word “Mama”, assigns a weight of 0.0180 to the word “Daddy”, assigns a weight of 0.0591 to the word “Baba”, and reads “Grandma”. Assume that a weight of 0.133 is given to the word and a weight of 0.033 is given to the word “Older brother”.

  Here, the weight assigning unit 22 refers to the word storage unit 41 of the web diary service providing server 4 and obtains information about whether or not each selected word is a word acquired so far. For example, it is assumed that the word “daddy” among the selected words has not been acquired and is not stored in the word storage unit 41.

  In this case, the weighting unit 22 corrects the weights given by the weighting unit 22 for all or part of the words so that the weights of the words not stored in the word storage unit 41 become larger, and gives weights. Instead of the weight given by the unit 22, the corrected weight is given to all or some of the words.

  In this example, the weight assigning unit 22 assigns the weight given by the weight assigning unit 22 to the word “papa”, for example, so that the weight of the word “papa” not stored in the word storage unit 41 becomes larger. Correction is performed by multiplying 0.0180 by a predetermined value (for example, 2), and the corrected weight (0.0360 = 0.0180 × 2) is assigned to the word “daddy” instead of the weight 0.0180 assigned by the weight assigning unit 22. . The predetermined value is greater than 1 and may be a predetermined positive number other than 2.

  In this way, by emphasizing the weight of unacquired words, it is possible to boost the acquisition of vocabulary similar to that of children of the same generation, in other words, children of similar ages.

  In the above example, the weight assigning unit 22 corrects only the weights given by the weight assigning unit 22 for a part of the word “papa”, but the words not stored in the word storage unit 41 are corrected. If the weight can be corrected to be larger, the weights given by the weight assigning unit 22 for words other than the word “daddy” (for example, all words) are stored in the word storage unit 41. You may correct using the function which can correct | amend so that the weight of the word which does not become larger.

  Note that the weighting unit 22 is a word that is not stored in the word storage unit 41 and has an X% acquired age earlier than the current age, in other words, past the X% acquired age as viewed from the current age. The weights given by the weighting unit 22 for all or part of the words are corrected so that the weights of the unacquired words are larger, and the weights given by the weighting unit 22 are corrected. Later weights may be assigned to all or some of the words. In this case, the weight assigning unit 22 uses information about the current age of the infant stored in advance in the word storage unit 41 of the web diary service providing server 4 as the current age.

<< Example 5 of weighting >>
In the weighting example 5, when the information about the words acquired so far is obtained by, for example, the web diary described below, the weighting unit 22 provides information about the words acquired so far. Is used to correct the weight. In the weighting example 5 and the weighting example 4, the weight correction method is different.

  In the weighting example 5, the weighting unit 22 is a word stored in the word storage unit 41, and the difference between the X% acquisition age of the word and the input word X% acquisition age is the magnitude of the difference. The weights given by the weighting unit 22 for all or some of the words are corrected so that the smaller the word, the larger the weight, and the corrected weights are replaced with the weights given by the weighting unit 22 in whole or Append to some words.

  As in the case of weighting example 4, a process when the word “grandma” is input in the “☆ words learned for the first time” field of the web diary will be described below as an example.

  When a caregiver uses the terminal device 5 to input the word “grandma” in the “☆ words learned for the first time” field of the web diary, the information about the input word “grandma” is the terminal device. 5 to the web diary service providing server 4. The web diary service providing server 4 outputs information about the inputted word “grandma” to the word selection unit 21 of the word presentation unit 2.

  The word selection unit 21 of the word presentation unit 2 reads a word related to the word “grandma” from the storage unit 1 by the process of step S21. For example, when the words “mama”, “dad”, “baoba”, and “older brother” are selected by the word selection unit 21, information on these selected words is output to the weighting unit 22. .

  In this case, first, the weight assigning unit 22 inputs the word “grandma” and the “mama” “daddy” selected by the word selecting unit 21 by the method of weight grant example 1 or weight grant example 3, for example. A weight is assigned to each of the words “baoba” and “older brother”. For example, the weight assigning unit 22 assigns a weight of 0.0085 to the word “Mama”, assigns a weight of 0.0180 to the word “Daddy”, assigns a weight of 0.0591 to the word “Baba”, and reads “Grandma”. Assume that a weight of 0.133 is given to the word and a weight of 0.033 is given to the word “Older brother”.

  Here, the weight assigning unit 22 refers to the word storage unit 41 of the web diary service providing server 4 and information about whether or not each selected word is a word stored in the word storage unit 41. And when it is memorize | stored in the word memory | storage part 41, the information about the age at which the word was acquired (the age at which the infant remembered the word) is acquired.

  In weighting example 5, the word “Mama” and the word “Daddy” among the selected words are the words stored in the word storage unit 41, and the age at which the word “Mama” was acquired. Is 13 months, the age at which the word “daddy” was acquired is 14 months, and the infant's current age is 15 months.

  In this case, the weight assigning unit 22 is a word stored in the word storage unit 41, and the weight is larger as the difference between the X% acquired age of the word and the current age of the infant is smaller. As described above, the weights given by the weighting unit 22 for all or some of the words are corrected, and the corrected weights are given to all or some of the words instead of the weights given by the weighting unit 22.

For example, for the word stored in the word storage unit 41, the weight assigning unit 22 sets the difference between the X% acquired age of the word and the current age of the infant as d, and the word weight assigning unit 22 Correction is performed by applying the attenuation coefficient Q ^d to the assigned weight w, and the corrected weight w × Q ^d is ^assigned to the word instead of the weight w assigned by the weight assigning unit 22. Q is a predetermined number greater than 0 and less than 1. For example, Q = 0.9.

Specifically, the weight assigning unit 22 corrects the weight 0.0085 assigned by the weight assigning unit 22 for the word “Mama” by applying an attenuation coefficient Q ^(15-13), and is given by the weight assigning unit 22. Instead of the weight of 0.0085, the corrected weight of 0.0085 × Q ^(15-13) is ^assigned to the word “Mama”. Further, the weight assigning unit 22 corrects the weight 0.0180 given by the weight assigning unit 22 for the word “daddy” by applying the attenuation coefficient Q ^(15-14) , and the weight given by the weight assigning unit 22 is 0.0180. Instead, the corrected weight of 0.0180 × Q ^(15-14) is assigned to the word “papa”.

  In this way, by relatively increasing the weight of the recently learned word, the infant can review the recently learned word.

  In the above example, the weight assigning unit 22 corrects only the weights given by the weight assigning unit 22 for the word “mama” and the part of the word “papa”. If it can be corrected so that the weight of the memorized word and the difference between the X% acquired age of the word and the current age of the infant is smaller, the weight is increased. The weights given by the weight assigning unit 22 for words other than words (for example, all words) are words stored in the word storage unit 41, and the X% acquisition age of the word and the current age of the infant You may correct using the function which can correct | amend so that a weight may become so large that the magnitude | size of the difference of this is small.

<< Example 6 of weighting >>
In the weighting example 6, the weighting unit 22 assigns weights based on word categories.

  First, the weight assigning unit 22 obtains the input word and the category of the selected word by referring to the storage unit 1. Then, the weight assigning unit 22 assigns a greater weight to a word included in a category having a category distance closer to the input word based on the obtained category information.

  For example, D = 1, the information that the animal category includes the carnivore category and the herbivore category is stored in the storage unit 1, and the input word is “lion” It is assumed that information indicating that the category of “lion” is a carnivore is stored in the storage unit 1. In this case, for example, the weight assigning unit 22 assigns weight 1 to the words included in the carnivore category that is the category of the input word “lion”, and the carnivores other than the words included in the carnivore A weight of 0.8 is assigned to words included in the category of animals that are parent categories that are separated by a hierarchical distance 1 based on the category, in other words, words included in the category of herbivores belonging to the animal category and the animal category, and so on. 0.5 is added to the words included in the category.

<< Example 7 of weighting >>
The weighting unit 22 receives the weight of the input word and the selected word obtained by any one of the methods <<<< weighting example 1 >> to << weighting example 6 >>. The weight of the normalization is normalized by the weight of the input word to obtain the normalized weight, and the obtained normalized weight is assigned to the input word and the selected word as the final weight, respectively. May be.

<Document Search Unit 3>
The document search unit 3 receives information about the input word and the selected word output by the word presentation device and the weights assigned to these words.

  The document search unit 3 performs a document search using the input word and the selected word as a keyword, outputted from the word presentation device, with a large weight given to each of the input word and the selected word. The emphasis is placed on this (step S3).

  An example of the document search unit 3 is an information presentation device described in, for example, Japanese Patent Application Laid-Open No. 2014-235723 or Japanese Patent Application Laid-Open No. 2006-148927.

Assume that the document search unit 3 includes a search target media data DB in which feature vectors v _n (n = 1,..., N) of each media data are stored. Here, the media data is, for example, document data.

Then, the document search unit 3 sets the word corresponding to the input word and the corresponding element of the selected word with respect to the feature vector having the number of types of words appearing in at least one volume in the search target media data DB. weight granted is inputted to, as a feature vector v _query of the other elements query feature vector consisting of zero, the query feature vector v _query similarity (e.g., cosine similarity) is higher feature vector v the information about the media data having a _n, is acquired by referring to the search target media data DB. For example, the document search unit 3 acquires information about a predetermined number of media data such as 30 in descending order of similarity. Note that the document search unit 3 ignores the weights assigned to the input words and the words in the selected words that did not appear in the media data stored in the search target media data DB. Then, the similarity is calculated.

  Then, the document search unit 3 outputs information about the acquired media data as a document search result corresponding to the input word and the selected word.

  In this way, the word presentation device selects a word related to the input word, and outputs the input word, the selected word, the input word, and the weight assigned to each of the selected word. Then, the document search unit 3 performs a document search based on the information output from these word presentation devices, so that not only the word input to the word presentation device but also words related to the input word are taken into consideration. Document search can be performed.

  In addition, in << weighting example 4 >> << weighting example 5 >> of the weighting unit 22, the caregiver uses the terminal device 5 to read “☆ words learned for the first time today” in the web diary. When a word is input in the column, a document search result corresponding to the input word may be displayed in the web diary.

  For example, when the word “grandma” is entered in the “☆ words learned for the first time today” field of the web diary, information about the entered word “grandma” is documented via the web diary service providing server 4. Output to search device. It is assumed that the document search apparatus performs the processing from step S21 to step S3 using the word “grandma” as the input word, and obtains the document “Aneto Ototo” as the document search result. In this case, information about the document search result is output to the web diary service providing server 4. The web diary service providing server 4 displays a document “Aneto Ototo” in, for example, the “Recommended Picture Book” field in the upper right of FIG.

  In this way, by displaying not only the word entered in the “☆ words learned for the first time today” column of the web diary but also the document search result considering the words related to the entered word in the web diary, A person can obtain information about documents suitable for the daily language development of infants.

[Modifications, etc.]
In the above embodiment, the example of acquiring the vocabulary of an infant has been described. However, the subject may not be an infant, in other words, a preschooler who is learning a native language. For example, when the subject is a person who is studying a second foreign language, the above “acquired age” corresponds to the first appearance page of a word in a textbook. When the subject is a person who is studying a new field, “acquired age” corresponds to the number of units that appear in the textbook for that field. In addition, the above “category” corresponds to a word that appears for the first time in the same unit or a word whose frequency of co-occurring in the same unit is higher than a predetermined reference.

  Further, the word presentation device and method may output only the word input to the word presentation device and the word selected by the word selection unit 21, and may not output the weights assigned to these words. In this case, the word presentation device may not have the weight assigning unit 22, and the word presentation method may not perform the process of step S22. In this case, the document search unit 3 performs a document search using the input word and the selected word output by the word presentation device as keywords without considering the weight.

  In the above description, the understanding curve, the utterance curve, and the X% acquired age are defined using the age of the moon. You may define using the parameter | index showing time.

  The processes described in each of the above apparatuses and methods are not only executed in time series in the order described, but may be executed in parallel or individually as required by the processing capability of the apparatus that executes the processes.

  Various processing functions in each of the above devices may be realized by a computer. In that case, the processing contents of the functions that each device should have are described by a program. Then, by executing this program on a computer, various processing functions in each of the above devices are realized on the computer.

  The program describing the processing contents can be recorded on a computer-readable recording medium. As the computer-readable recording medium, for example, any recording medium such as a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory may be used.

  The program is distributed by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Further, the program may be distributed by storing the program in a storage device of the server computer and transferring the program from the server computer to another computer via a network.

  A computer that executes such a program first stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its storage unit. When executing the process, this computer reads the program stored in its own storage unit and executes the process according to the read program. As another embodiment of this program, a computer may read a program directly from a portable recording medium and execute processing according to the program. Further, each time a program is transferred from the server computer to the computer, processing according to the received program may be executed sequentially. Also, the program is not transferred from the server computer to the computer, and the above-described processing is executed by a so-called ASP (Application Service Provider) type service that realizes the processing function only by the execution instruction and result acquisition. It is good. Note that the program includes information provided for processing by the electronic computer and equivalent to the program (data that is not a direct command to the computer but has a property that defines the processing of the computer).

  In addition, each apparatus is configured by executing a predetermined program on a computer, but at least a part of these processing contents may be realized in hardware or manually.

DESCRIPTION OF SYMBOLS 1 Memory | storage part 2 Word presentation part 21 Word selection part 22 Weight provision part 3 Document search part 4 Web diary service provision server 41 Word storage part 5 Terminal device

Claims (16)

As information about the acquisition time when the infant acquires a word, as information about the acquisition time of the word, a storage unit storing information about each word and the acquisition time of each word, and information about the acquisition time of the input word A word presentation device including a word selection unit that selects from the storage unit a word having information on the acquisition time of a word close to
A document search unit that performs a document search using the input word output by the word presentation device and the selected word as a keyword;
Document retrieval device including The document search device according to claim 1,
The word presenting unit uses the information about the input time of the input word and the information about the time of acquisition of the selected word to determine the importance of the word for each of the input word and the selected word. And further includes a weight assigning unit for assigning a weight of
The word presentation unit outputs the input word and the selected word, and the weight given to each of the input word and the selected word,
The document search unit is provided with each of the input word and the selected word for a document search using the input word and the selected word output by the word presentation device as keywords. Focus on words with greater weight,
Document retrieval device. Information about the acquisition time when the infant acquires a word is information about the acquisition time of the word, and the word belongs to a predetermined category, and each word, information about the acquisition time of each word, and the category of each word , A word selection unit that selects words in the same category as the input word from the storage unit, and information on the acquisition time of the input word and acquisition of the selected word A weight assigning unit that assigns a weight that is the importance of the word to each of the input word and the selected word, using the information about the time, and the input word and the selected word A word presenting device that outputs the input word and the weight assigned to each of the selected words, and
Document search using the input word and the selected word output by the word presenting device as keywords is given more importance to a word having a higher weight given to the input word and the selected word. A document search unit
Document retrieval device including The document search device according to claim 2 or 3,
The weight assigning unit assigns a greater weight to a word having information related to a word acquisition time that is closer to the information related to the input word acquisition time.
Document retrieval device. The document search device according to any one of claims 2 to 4,
X is a predetermined number of 0 ≦ X ≦ 100, and the information regarding the acquisition time of the word is X% acquisition age at which X% of the infant acquires the word,
Document retrieval device. The document search device according to claim 5,
The weight assigning unit weights a word whose X% acquisition age is earlier than the X% acquisition age of the input word, and is slower than the X% acquisition age of the input word. The weights given by the weighting unit for all or part of the words are corrected so that the weights of the words are larger, and the corrected weights are replaced with the weights given by the weighting unit. To give to a word
Document retrieval device. The document search device according to any one of claims 2 to 4,
Let the understanding curve of a word be the curve that represents the percentage of infants that understand the word relative to the infant's age, and the utterance curve of the word represents the percentage of infants who can speak that word relative to the infant's age As described above, the information about the acquisition time of the word is at least one of information about an understanding curve of the word and information about an utterance curve.
Document retrieval device. The document search device according to any one of claims 2 to 5 and 7,
A word storage unit that stores information about words acquired so far;
The weighting unit corrects the weights given by the weighting unit for all or some of the words so that the weights of words not stored in the word storage unit are larger, and the weighting unit In place of the assigned weight, the corrected weight is given to all or some of the words.
Document retrieval device. The document search device according to any one of claims 2 to 5 and 7,
A word storage unit that stores information about words acquired so far;
X is a predetermined number of 0 ≦ X ≦ 100, and X% acquisition age of the word is the age at which X% of the infant acquires the word, and the weight assigning unit stores the word stored in the word storage unit The weights given by the weighting unit for all or some of the words are corrected so that the smaller the difference between the X% acquired age of the word and the current age of the infant is, the higher the weight is. Then, instead of the weight given by the weight giving unit, the corrected weight is given to all or some of the words.
Document retrieval device. A storage unit in which information about an acquisition time when an infant acquires a word is stored as information about an acquisition time of the word, and information about an acquisition time of each word and each word,
A word presenting unit including a word selection unit that selects from the storage unit a word having information about the acquisition time of a word close to the information about the input time of the input word;
Word presentation device including Information about the acquisition time when the infant acquires a word is information about the acquisition time of the word, and the word belongs to a predetermined category, and each word, information about the acquisition time of each word, and the category of each word And a storage unit storing
Using the word selection unit that selects a word in the same category as the input word, the information about the acquisition time of the input word and the information about the acquisition time of the selected word, the input word and the Each of the selected words includes a weight assigning unit that assigns a weight that is the importance of the word to each of the input word, the selected word, the input word, and the selected word. A word presentation unit for outputting the assigned weight;
Word presentation device including In the storage unit, information on the acquisition time when the infant acquires a word is stored as information on the acquisition time of the word, and information on the acquisition time of each word and each word is stored.
A word selection step in which the word selection unit selects from the storage unit a word having information on the acquisition time of the word close to the information on the input time of the word;
A document search step in which a document search unit performs a document search using the input word and the selected word as a keyword;
Search method including documents. In the storage unit, information about the acquisition time when the infant acquires a word is information about the acquisition time of the word, and the word belongs to a predetermined category, and information about the acquisition time of each word and each word, Assuming that each word category is stored,
The word presenting unit selects a word in the same category as the input word from the storage unit, and uses the information regarding the acquisition time of the input word and the information about the acquisition time of the selected word. A weight that is the importance of the word is assigned to each of the selected word and the selected word, and is given to each of the input word, the selected word, the input word, and the selected word. A word presentation step for outputting the weights obtained,
The document search unit performs a document search using the input word and the selected word as a keyword, focusing on a word having a larger weight assigned to each of the input word and the selected word. A document search step;
Search method including documents. In the storage unit, information on the acquisition time when the infant acquires a word is stored as information on the acquisition time of the word, and information on the acquisition time of each word and each word is stored.
A word selection step in which the word selection unit selects, from the storage unit, a word having information on the acquisition time of the word that is close to the information on the acquisition time of the input word;
Word presentation method including In the storage unit, information on the acquisition time when the infant acquires a word, information on the acquisition time of the word, information on the acquisition time of each word and each word, and the word belong to a predetermined category, Assuming that each word category is stored,
The word presenting unit selects a word in the same category as the input word from the storage unit, and uses the information regarding the acquisition time of the input word and the information about the acquisition time of the selected word. A weight that is the importance of the word is assigned to each of the selected word and the selected word, and is given to each of the input word, the selected word, the input word, and the selected word. A word presentation step for outputting the weights obtained,
Word presentation method including   The program for functioning a computer as each part of the document search device in any one of Claim 1 to 9, or the word presentation apparatus of Claim 10 or 11.

Images