JP2013097723A - Text summarization apparatus, method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To learn parameters so as to directly maximize a score in an evaluation scale such as ROUGE, which evaluates quality of a summary.SOLUTION: The text summarization apparatus comprises: a first storage section 14 which stores a training example database 12 storing training examples in which texts xare associated with summaries ypreliminary generated from the texts x; and a parameter estimation section 16 which generates a summary y' from the text xof the training example, obtains an error of the generated summary y' to the summary yof the training example by using an evaluation scale for evaluating quality of a summary, updates parameters used for generating a summary y so that the greater the obtained error is the greater a width of the update is, and stores the updated parameters in a parameter database 18. The parameters stored in the parameter database 18 are used for generating the summary y from the text x to be summarized by a text summarization section 24.

Description

  The present invention relates to a text summarization device, a text summarization method, and a text summarization program.

  In recent years, a large amount of digitized text has been distributed, and in order to quickly grasp the information described in the text, the text is summarized by a machine such as a computer (summary The need for technology to generate

  In the technology of summarizing text by machine, in general, one or more sentences (important sentences) that are considered to represent the contents of the text to be summarized (important sentences) or one or more such sentences (important sentence set) are represented. A summary is generated by picking and concatenating them. When evaluating whether a sentence or a set of sentences is representative of the content of the text to be summarized, each sentence or set of sentences is expressed as a feature vector, An inner product of a weight vector estimated by some method (hereinafter also referred to as a parameter) and a score is calculated (for example, see also Non-Patent Document 1).

  A parameter for calculating the inner product of a sentence or a feature vector representing a sentence set is a set of pairs consisting of a text to be summarized or a set of text to be summarized and a corresponding summary (hereinafter, Based on training examples). In learning the parameters, the parameters are learned to minimize some error function.

Wen-tau Yih, Joshua Goodman, Lucy Vanderwende and Hisami Suzuki, “Multi-Document Summarization by Maximizing Informative Content-Words.”, In Proceedings of International Joint Conference on Artificial Intelligence (IJCAI), 2007. Koby Crammer, Ofel Dekel, Joseph Keshet, Shai Shalev-Shwartz and Yoram Singer, “Online Passive-Aggressive Algorithms.”, Journal of Machine Learning Research, Vol. 7, 2006.

  However, when learning parameters used in a technology for summarizing text, for example, the technology described in Non-Patent Document 1 does not learn parameters that directly maximize the quality of the summary, and is easily included in the summary. The parameters are learned so that the sentence has a high score. As a scale for evaluating the quality of a summary generated by a machine, for example, an evaluation scale called ROUGE (see Non-Patent Document 2) is frequently used, but sentences that are easily included in the summary have a high score. There is no direct relationship between learning the parameters so that the evaluation scale such as ROUGE of the summary generated using the learned parameters has a high value, and there is room for improvement in parameter learning accuracy.

  The present invention has been made in consideration of the above facts, and is a text summarization apparatus, a text summarization method, and a text summarization method capable of learning parameters so that the value of an evaluation measure for evaluating the quality of summaries such as ROUGE is directly maximized. The purpose is to obtain a text summarization program.

  A text summarization apparatus according to the present invention includes a first storage unit that stores a training case database storing training cases in which texts and summaries generated in advance from the texts are associated with each other, and is included in the training cases. An error in the quality of the summary generated from the text contained in the training case, using the pre-update values of the parameters used to generate the summary from the text to be summarized relative to the quality of the summary associated with the text Is estimated using an evaluation scale for evaluating the quality of the summary, and the parameter is estimated by updating the value before updating the parameter so that the range of update increases as the calculated error increases. A parameter estimation unit.

Further, in the present invention, the parameter estimation unit is configured such that the _old value of the weight vector as the parameter is w _old , the text included in the training example is x _i , and the summary included in the training example is y _i. the summary generated from the text x _i included in the training set using a value before the update of the weight vector as the parameter y ', the feature vectors of text x and abstract y f (x, y), When the evaluation scale is ROUGE and the error is loss (y ′; y _i ),

And
loss (y ′; y _i ) = 1−ROUGE (y ′; y _i ) (3)

The parameter can be estimated by calculating the updated value w _new of the weight vector as the parameter according to the above equations (1) to (3).

  The present invention is also received by the second storage unit that stores the parameter database that stores the parameters estimated by the parameter estimation unit, the text input unit that receives the text to be summarized, and the text input unit. Based on the text to be summarized and the parameters stored in the parameter database, a text summarizing unit that generates a summary of the text to be summarized, and the summary generated by the text summarizing unit as text A text output unit for outputting.

  In the text summarizing method according to the present invention, a training example database storing training examples in which texts and summaries generated in advance from the texts are associated is stored in the first storage unit. A summary of the summary generated from the text included in the training case using the pre-update values of the parameters used to generate the summary from the summary text for the quality of the summary contained and associated with the text. An error in quality is obtained using an evaluation scale for evaluating the quality of the summary, and the parameter is updated by updating the value before update so that the range of update increases as the obtained error increases. A parameter estimation step of estimating.

  The text summarization program according to the present invention causes a computer to function as each means constituting the text summarization device according to any one of claims 1 to 3.

  The present invention has the effect that the parameter can be learned so that the value of the evaluation measure for evaluating the quality of the summary such as ROUGE is directly maximized.

It is a block diagram which shows the text summarizing apparatus demonstrated by embodiment. It is a schematic block diagram of the computer which functions as a text summarization apparatus. It is a flowchart which shows the outline of a text summary process. It is a chart which shows an example of the training example stored in the training example database. (A) is a chart showing an example of a text and a summary, and (B) is an explanatory diagram showing an example of a feature vector corresponding to the text and the summary shown in (A). It is a flowchart which shows an example of the algorithm of a parameter learning process. It is a chart which shows an example of a parameter database. It is a graph which shows an example of the input format of the text of the summary object. It is a flowchart which shows an example of the algorithm of a summary production | generation process.

  Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 shows a text summarizing apparatus 10 according to the present embodiment. The text summarizing apparatus 10 is stored in the training case database 12 and a first storage unit 14 that stores a training case database 12 that stores training cases necessary for estimating parameters (weight vector w in the present embodiment). A parameter estimation unit 16 that accepts training examples and estimates parameters; a second storage unit 20 that stores a parameter database 18 that stores parameters estimated by the parameter estimation unit 16; and accepts text to be summarized A text input unit 22, a text summary unit 24 that generates a summary by using the text to be summarized received by the text input unit 22 and parameters stored in the parameter database 18, and a summary generated by the text summary unit 24 And a text output unit 26 for outputting.

  The text summarizing apparatus 10 can be realized by, for example, a computer 30 shown in FIG. The computer 30 includes a CPU 32, a memory 34, a nonvolatile storage unit 36, a keyboard 38, a mouse 40, and a display 42, which are connected to each other via a bus 44. The storage unit 36 can be realized by an HDD (Hard Disk Drive), a flash memory, or the like. In the storage unit 36 as a recording medium, a text summarizing program 46 for causing the computer 30 to function as the text summarizing device 10, the training example database 12, and the parameter database 18 are stored. The CPU 32 reads the text summary program 46 from the storage unit 36 and expands it in the memory 34, and sequentially executes processes included in the text summary program 46.

  The text summarization program 46 includes a text input process 50, a parameter estimation process 52, a text summarization process 54, and a text output process 56. The CPU 32 operates as the text input unit 22 shown in FIG. 1 by executing the text input process 50. The CPU 32 operates as the parameter estimation unit 16 illustrated in FIG. 1 by executing the parameter estimation process 52. The CPU 32 operates as the text output unit 26 shown in FIG. 1 by executing the text output process 56. The text summary program 46 is an example of a text summary program according to the present invention.

  When the text summarizing apparatus 10 is realized by the computer 30, the storage unit 36 that stores the training case database 12 is used as the first storage unit 14, and a partial area of the memory 34 is used as the training case database 12. The storage unit 36 that stores the parameter database 18 is used as the second storage unit 20, and a partial area of the memory 34 is used as the parameter database 18. As a result, the computer 30 that has executed the text summarization program 46 functions as the text summarization apparatus 10.

  Next, as an operation of the present embodiment, an outline of text summarization processing realized by executing the text summarization program 46 by the CPU 32 of the computer 30 will be described first with reference to FIG. In step 60 of the text summarization process, the parameter estimation unit 16 receives a training case from the training case database 12, and learns parameters based on the received training case (details of parameter learning will be described later). In the next step 62, the parameter estimation unit 16 stores the learned parameters in the parameter database 18.

  In the next step 64, the text input unit 22 receives the input text to be summarized. The text to be summarized may be, for example, text input via the keyboard 38, may be text stored in advance in the storage unit 36, or may be another text connected to the computer 30 via, for example, a communication line. It may be text received from the device.

  In step 66, the text summary unit 24 generates a summary from the summary target text received by the text input unit 22 and the parameters stored in the parameter database 18 (details of the summary generation will be described later). .

  In step 68, the text output unit 26 outputs the summary generated by the text summary unit 24 to the outside. The output of the summary may be displayed as text on the display 42, for example, may be output as a voice that reads the summary, or may be recorded as text data on a recording medium or transmitted to another device. .

  Further, in FIG. 3, parameter learning and storage in the parameter database 18 (steps 60 and 62) and summary generation / output (steps 64 to 68) are shown as a series of processing. After learning and storing in the parameter database 18, a summary may be generated and output after a time interval.

  Next, in describing the details of the text summarization process, first, training examples stored in the training example database 12 will be described with reference to FIG. FIG. 4 shows an example of training cases stored in the training case database 12. In FIG. 4, the training case database 12 stores one training case in each record. Each training case stored in each record of the training case database 12 includes a case number for identifying an individual training case, a text to be summarized, and summary information generated from the text to be summarized. Is included. Note that the summary included in the training example is a correct summary that is created manually, for example, and can be used as a reference in learning a parameter (weight vector w) described later.

  Subsequently, generation of a summary using a parameter (weight vector w) will be described. Assume that the text to be summarized is x, and the summary generated from the text x to be summarized is y. Given the text x to be summarized, a feature vector f (x, y) is obtained from the text x to be summarized and the summary y generated from the text x to be summarized. The inner product w · f (x, y) of the feature vector f (x, y) and the parameter (weight vector w) is the score of the summary y. When the summary is generated, the inner product w · f (x, y) of the parameter (weight vector w) and the feature vector f (x, y) among the summaries y generated from the text x to be summarized is the maximum. Is generated as a summary y ′. The above is expressed by the following equation (4).

y ′ = arg _y max w · f (x, y) (4)
An example of a summary generation method will be described later, but any summary generation method can be applied.

  The feature vector f (x, y) obtained from the text x to be summarized and the summary y generated from the text x corresponds to a word whose elements are different from each other. For example, the summary y includes a certain word. A vector in which the value of the corresponding element is 1 if it is present and the value of the corresponding element is 0 if it is not included can be applied. An example of the feature vector f (x, y) obtained from the text x and the summary y shown in FIG. 5 (A) is shown in FIG. 5 (B). Since the summary y shown in FIG. 5A includes words such as “on the road” and “robbery”, the feature vector f (x, y) shown in FIG. It is 1. On the other hand, since the summary y shown in FIG. 5A does not include words such as “road” and “thief”, the feature vector f (x, y) shown in FIG. 5B corresponds to these words. The element value is 0.

Next, the parameter learning performed by the parameter estimation unit 16 will be described. Assuming that n training cases are given, let the text of the i-th training case be x _i and the summary be y _i . For example, the text of case number 1 is x1, and the summary is y1.

When learning the parameters, n training examples are picked up one by one. Under the current parameter w, when the summary y ′ generated from the text x _i and the correct summary y _i created manually are obtained, the goodness (quality) of the summary y ′ as a summary is , summary y value of ROUGE of 'rating scale for evaluating, using ROUGE as an example, reference to digest y summary _{y i' (ROUGE (y '} ; y i)) can be calculated as. Since the value of ROUGE is in the range of 0 to 1, the summary y ′ when the summary y _i is used as the reference (correct summary) by subtracting the value of ROUGE from 1 as shown in the above equation (3). , I.e., the error of summary y ′ with respect to summary y _i is obtained. In the above equation (3), the error function loss (y ′; y _i ) is used as an error.

  If the parameter can be updated so as to reduce the above error, a parameter capable of generating a better summary can be obtained. Therefore, the parameter is updated as in the following equation (5).

In equation (5), w _old is a current parameter (before update), and w _new is a parameter after update. The above equation (5) indicates that the summary y ′ in which the score of the correct summary y _i is generated using the current (before update) parameter w _old in the updated parameter w _new while minimizing the change of the parameter w. The parameter w is updated so that it always becomes larger than the score of. At this time, the residual obtained by subtracting the score of the summary y ′ from the score of the summary y _i is set to be equal to or larger than the error function loss (y ′; y _i ). When the above equation (5) is solved, the above-described parameter update equations (equations (1) and (2)) are obtained.

In the above update formula (formulas (1) and (2)), the error of the summary y ′ with respect to the summary y _i obtained by using ROUGE as the error function for the current (pre-update) parameter w _new increases. Since the update is performed so that the range of the update is increased, the parameter w can be updated (learned) so that the value of ROUGE directly increases as the parameter w is updated.

Next, an example of a specific algorithm of the parameter learning process (details of steps 60 and 62 of the text summarization process shown in FIG. 3) by the parameter estimation unit 16 will be described with reference to FIG. In step 70 the parameter learning processing, parameter estimation unit 16, a text x _i and abstract y _i included in the training set of case number i is assigned from the training case database 12 of n (case number i = 1 to n Acquire training examples). In step 72, the parameter estimation unit 16 acquires a preset repetition count T. In the next step 74, the parameter estimation unit 16 initializes the parameter (weight vector w) to a vector in which all element values are zero. In step 76, the parameter estimation unit 16 sets 1 to the variable t, and in the next step 78, the parameter estimation unit 16 sets 1 to the variable i.

In the next step 80, the parameter estimation unit 16 calculates the inner product w · f (x _i , y) between the current parameter (weight vector w) and the feature vector f (x _i , y) for the training example with the case number i. Is generated as a summary y ′ (see also the following equation (6)).

y ′ = arg _y max w · f (x _i , y) (6)

In the next step 82, the parameter estimation unit 16 first calculates the error loss (y ′; y _i ) of the summary y ′ with respect to the summary y _{i according} to the above equation (3), and then calculates the parameter update equation ( The parameter (weight vector w) is updated by (1) and (2). As a result, the parameter (weight vector w) is updated so that the update range increases as the error of the summary y ′ with respect to the summary y _i increases.

  In step 84, the parameter estimation unit 16 determines whether or not the variable i is equal to or greater than the total number n of training examples acquired in the previous step 70. If the determination in step 84 is negative, the process proceeds to step 86. In step 86, the parameter estimation unit 16 increments the variable i by 1, and then returns to step 80. Thus, steps 80 to 86 are repeated until the determination in step 84 is affirmed, and updating (learning) of the parameter (weight vector w) is repeated n times using n training cases one by one.

  If the determination at step 84 is affirmative, the routine proceeds to step 88. In step 88, the parameter estimation unit 16 determines whether or not the variable t has reached the number of repetitions T acquired in the previous step 72. If the determination in step 88 is negative, the process proceeds to step 90. In step 90, the parameter estimation unit 16 increments the variable t by 1, and then returns to step 78. Thereby, step 78 to step 90 are repeated until the determination in step 88 is affirmed, and each of the n training cases is used T times and the update (learning) of the parameter (weight vector w) is repeated. Thereby, the parameter (weight vector w) is updated (learned) so that the value of ROUGE, which is an evaluation measure for evaluating the quality of the summary, is directly maximized.

  If the determination in step 88 is affirmative, the updating of the parameter (weight vector w) is terminated, and the process proceeds to step 92. In step 92, the parameter estimation unit 16 updates the parameter (weight) updated (learned) by the above processing. The vector w) is stored in the parameter database 18, and the parameter learning process is terminated. An example of the parameter database 18 is shown in FIG. In the parameter database 18, information on “feature” such as whether or not a certain word is included in each record and “weight” of the feature is set. The first record in the parameter database 18 shown in FIG. 7 indicates that the weight of the feature that includes the word “road” is −0.03.

  After the parameter learning process described above is performed, in step 64 of the text summarization process shown in FIG. 3, the text input unit 22 receives the input text x to be summarized. An example of the input format of the text x acquired by the text input unit 22 is shown in FIG. In the input format of the text x shown in FIG. 8, the text x is divided for each sentence, and the result of morphological analysis is input. In FIG. 8, each row of the table corresponds to a single morpheme, and information of “notation”, “part of speech”, “reading”, and “standard form” is added to each morpheme.

  Next, an example of a specific algorithm of the summary generation process (details of step 66 of the text summary process shown in FIG. 3) by the text summarizing unit 24 will be described with reference to FIG. In step 100 of the summary generation process, the text summarization unit 24 acquires the text to be summarized x and the summary length K from the text input unit 22. In step 102, the text summarizing unit 24 acquires a parameter (weight vector) w from the parameter database 18. In step 104, the text summarizing unit 24 stores the text x acquired in step 100 in the sentence set U. In the next step 106, the text summarizing unit 24 sets the set S representing the summary to the empty set φ. initialize.

In step 108, the text summarizing unit 24 determines whether the sentence set U is larger than the empty set φ (whether any sentence is stored in the sentence set U). If the determination at step 108 is affirmative, the routine proceeds to step 110. In step 110, the text summarizing unit 24 calculates the score (= inner product of the feature vector f and the parameter w) when the sentence s _i is put into the summary set S from the sentences stored in the sentence set U. selecting the difference in scores when not put a sentence s _i, sentence s _i sentence s _i divided by the length of the maximum of the _(next (7) sentence s _i satisfying _equation) to a set of abstract S To do.

s _i = arg max _siεU ((w · f (x, {S, s _i } −w · f (x, S)) / length (s _i )) (7)
In equation (7), length (s _i ) is the length of sentence s _i .

In the next step 112, the text summarizing unit 24 determines whether or not the length obtained by adding the length of the sentence s _i selected in step 110 to the length of the sentence already in the summary set S is equal to or shorter than the summary length K. (See also the following equation (8)).
length ({S, s _i }) ≦ K (8)

If the determination at step 112 is affirmative, the routine proceeds to step 114. In step 114, the text summarizing section 24 adds the sentence s _i selected in step 110 to the summary set S (see also the following equation (9)), and then proceeds to step 116.
S = {S, s _i } (9)

If the determination in step 112 is negative, step 114 is skipped (without adding the sentence s _i to the summary set S) and the process proceeds to step 116.

In step 116, the text summarizing unit 24 removes the sentence s _i selected in step 110 from the sentence set U (U = U−s _i ). If the process of step 116 is performed, it will return to step 108 and will repeat step 108-step 116 until determination of step 108 is affirmed. When the sentence set U becomes the empty set φ, the determination in step 108 is negative and the process proceeds to step 118.

In step 118, the text summarizing unit 24 selects, as a sentence v, a sentence s _i having a length within the summary length K and having the highest score as a single sentence among sentences of the text x (next (10 (See also).)
v = arg max _{siεx: length (si) ≦ K} w · f (x, s _i ) (10)

  In the next step 120, the text summarizing unit 24 determines whether the score of the sentence included in the summary set S is equal to or higher than the score of the sentence v selected in the previous step 118. If the determination at step 120 is affirmative, the routine proceeds to step 122. In step 122, the text summarizing unit 24 outputs the sentences contained in the summary set S as a summary of the text x, and ends the summary generation process. If the determination at step 120 is negative, the process proceeds to step 124. In step 124, the text summarizing unit 24 outputs the sentence v selected in the previous step 118 as a summary of the text x, and ends the summary generation process.

  The summary of the text x output by the text summarization unit 24 in step 122 or step 124 of the summary generation process described above is output to the outside by the text output unit 26 in step 68 of the text summarization process shown in FIG.

  Note that the summary generation method shown as the summary generation process in FIG. 9 is an example, and the parameter update method in the present invention can be combined with any summary generation method.

  Moreover, although the aspect using ROUGE as an evaluation scale for evaluating the quality of the summary has been described above, the present invention is not limited to this, and an evaluation scale other than ROUGE can be applied.

In the above description, the training example database 12 is stored in the storage unit 36 provided in the computer 30. However, the present invention is not limited to this. You may memorize | store in the memory | storage part provided in another connected computer. In this case, when learning the parameters, n training examples each including the text x _i and the summary y _i can be received from the other computer via the communication line. Further, the text x to be summarized may be received from the other computer via a communication line.

  Furthermore, in the above description, the mode in which the computer 30 functions as the text summarizing device 10 by the computer 30 executing the text summarizing program 46 has been described. However, the present invention is not limited to this and is shown in FIG. Each functional block (parameter estimation unit 16, text input unit 22, text summarization unit 24, and text output unit 26) can also be configured by hardware.

  In the above description, the text summary program 46 is stored (installed) in the storage unit 36 in advance. However, the text summary program according to the present invention is recorded on a recording medium such as a CD-ROM or a DVD-ROM. It is also possible to provide it in the form.

  All documents, patent applications and technical standards mentioned in this specification are to the same extent as if each individual document, patent application and technical standard were specifically and individually stated to be incorporated by reference. Incorporated by reference in the book.

DESCRIPTION OF SYMBOLS 10 Text summary apparatus 12 Training example database 14 1st memory | storage part 16 Parameter estimation part 18 Parameter database 20 2nd memory | storage part 22 Text input part 24 Text summary part 26 Text output part 30 Computer 34 Memory 36 Memory | storage part 46 Text summary program

Claims (5)

A first storage unit that stores a training case database storing training cases in which texts and summaries generated in advance from the texts are associated with each other;
Generated from the text included in the training case using the pre-update values of the parameters used to generate the summary from the text to be summarized for the quality of the summary included in the training case and associated with the text An error in the quality of the summary obtained is obtained using an evaluation scale for evaluating the quality of the summary, and the value before the update of the parameter is updated so that the range of update becomes larger as the obtained error becomes larger. A parameter estimation unit for estimating the parameter;
Text summarization device. The parameter estimation unit, w _old value before updating the weight vector as the _parameter, the text x _i included in the training _set, the summary of the y _i included in the training _set, the weights of the said parameter The summary generated from the text x _i included in the training example using the values before the update of the vector is y ′, the feature vector of the text x and the summary y is f (x, y), the evaluation measure is ROUGE, When the error is loss (y ′; y _i ),

And
loss (y ′; y _i ) = 1−ROUGE (y ′; y _i ) (3)
(1) to (3) text summarizing apparatus according to claim 1, wherein for estimating the parameter by calculating a value w _{new new} updated weight vector as the parameter according to equation. A second storage unit for storing a parameter database for storing the parameters estimated by the parameter estimation unit;
A text input section that accepts text to be summarized;
A text summarization unit that generates a summary of the text to be summarized based on the text to be summarized received by the text input unit and the parameter stored in the parameter database;
A text output unit for outputting the summary generated by the text summary unit as text;
The text summarization apparatus according to claim 1, further comprising: In a state where a training case database storing a training case in which a text is associated with a summary generated in advance from the text is stored in the first storage unit,
Generated from the text included in the training case using the pre-update values of the parameters used to generate the summary from the text to be summarized for the quality of the summary included in the training case and associated with the text An error in the quality of the summary obtained is obtained using an evaluation scale for evaluating the quality of the summary, and the value before the update of the parameter is updated so that the range of update becomes larger as the obtained error becomes larger. A text summarization method comprising: a parameter estimation step for estimating the parameter.   A text summarization program for causing a computer to function as each means constituting the text summarization device according to any one of claims 1 to 3.