U.S. patent application number 16/005800 was filed with the patent office on 2019-01-31 for extractive query-focused multi-document summarization.
The applicant listed for this patent is INTERNATIONAL BUSINESS MACHINES CORPORATION. Invention is credited to Odellia Boni, Guy Feigenblat, David Konopnicki, Haggai Roitman.
Application Number | 20190034537 16/005800 |
Document ID | / |
Family ID | 62749547 |
Filed Date | 2019-01-31 |
![](/patent/app/20190034537/US20190034537A1-20190131-D00000.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00001.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00002.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00003.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00004.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00005.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00006.png)
![](/patent/app/20190034537/US20190034537A1-20190131-D00007.png)
United States Patent
Application |
20190034537 |
Kind Code |
A1 |
Boni; Odellia ; et
al. |
January 31, 2019 |
EXTRACTIVE QUERY-FOCUSED MULTI-DOCUMENT SUMMARIZATION
Abstract
A method, computer system, and computer program product for
generating a multi-document summary is provided. The embodiment may
include receiving a query statement, one or more documents, one or
more summary constraints, and quality goals. The embodiment may
include identifying one or more keywords within the query
statement. The embodiment may include performing a sentence
selection from the one or more documents based on the one or more
identified keywords. The embodiment may include generating a
plurality of candidate summaries of the one or more documents based
on the performed sentence selection, the goals, and a cross entropy
method. The embodiment may include calculating a quality score for
each of the plurality of generated candidate summaries using a
plurality of quality features. The embodiment may include selecting
a candidate summary from the plurality of generated candidate
summaries with the highest calculated quality score that also
satisfies a quality score threshold.
Inventors: |
Boni; Odellia; (Giva'at Ela,
IL) ; Feigenblat; Guy; (Givataym, IL) ;
Konopnicki; David; (Haifa, IL) ; Roitman; Haggai;
(Yoknea'm Elit, IL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
INTERNATIONAL BUSINESS MACHINES CORPORATION |
ARMONK |
NY |
US |
|
|
Family ID: |
62749547 |
Appl. No.: |
16/005800 |
Filed: |
June 12, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15843993 |
Dec 15, 2017 |
10019525 |
|
|
16005800 |
|
|
|
|
15660034 |
Jul 26, 2017 |
10127323 |
|
|
15843993 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 16/345 20190101;
G06F 16/9038 20190101; G06F 16/93 20190101; G06F 16/3334 20190101;
G06F 16/9535 20190101; G06F 16/90332 20190101 |
International
Class: |
G06F 17/30 20060101
G06F017/30 |
Claims
1. A computer system for generating a multi-document summary, the
computer system comprising: one or more processors, one or more
computer-readable memories, one or more computer-readable tangible
storage media, and program instructions stored on at least one of
the one or more tangible storage media for execution by at least
one of the one or more processors via at least one of the one or
more memories, wherein the computer system is capable of performing
a method comprising: receiving, by a processor, a query statement,
one or more documents, one or more summary constraints, and one or
more goals, wherein the query statement is a user-entered series of
words relating to a particular topic or asks a question towards
which a user desires more information, and wherein the one or more
documents are selected from a group consisting of a plurality of
articles, a plurality of stories, a plurality of papers, and a
plurality of repositories, and wherein the one or more documents
are capable of being search and summarized, and wherein the one or
more summary constraints are a plurality of factors that impact a
generated summary, and wherein the one or more goals are one or
more criteria to be achieved when generating a summary of the one
or more documents, and wherein the one or more summary constraints
are fixed and must be adhered to, and wherein the one or more goals
are fluid to allow for various generated summaries to each be
evaluated for quality; identifying one or more keywords within the
query statement, wherein each keyword relates to an important
aspect of the query statement; performing a sentence selection from
the one or more documents based on the one or more identified
keywords, wherein performing the sentence selection comprises
parsing each received document to identify one or more sentences
relevant to the query statement, and wherein the performed sentence
selection is a first step of a constrained global optimization
problem to identify a subset of one or more sentences in each
document that maximize a given quality target function, and wherein
the performed sentence selection utilizes a sentence filtering
technique, and wherein the sentence filtering technique is selected
from a group consisting of keyword identification and similarity
score rating; generating a plurality of candidate summaries of the
one or more documents based on the performed sentence selection,
the one or more goals, and a fully-polynomial randomized
approximation scheme (FPRAS) cross entropy method; calculating a
quality score for each of the plurality of generated candidate
summaries using a plurality of quality features, wherein the
calculated quality score is an integer between zero and one hundred
or a letter grade; and selecting a candidate summary from the
plurality of generated candidate summaries with the highest
calculated quality score that also satisfies a quality score
threshold.
Description
BACKGROUND
[0001] The present invention relates, generally, to the field of
computing, and more particularly to natural language
processing.
[0002] Natural language processing may relate to a field within
computer science dealing with the analysis of natural language text
by a computer. Natural language processing may have applications in
various categories, such as syntax, semantics, discourse, and
speech. Various applications within the mentioned categories
include text-to-speech, discourse analysis, relationship
extraction, natural language generation, sentiment analysis,
parsing, word segmentation, terminology extraction, machine
translation, and summarization.
SUMMARY
[0003] According to one embodiment, a method, computer system, and
computer program product for generating a multi-document summary is
provided. The embodiment may include receiving a query statement,
one or more documents, one or more summary constraints, and one or
more goals. The embodiment may also include identifying one or more
keywords within the query statement. The embodiment may further
include performing a sentence selection from the one or more
documents based on the one or more identified keywords. The
embodiment may also include generating a plurality of candidate
summaries of the one or more documents based on the performed
sentence selection, the one or more goals, and a cross entropy
method. The embodiment may further include calculating a quality
score for each of the plurality of generated candidate summaries
using a plurality of quality features. The embodiment may also
include selecting a candidate summary from the plurality of
generated candidate summaries with the highest calculated quality
score that also satisfies a quality score threshold.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0004] These and other objects, features and advantages of the
present invention will become apparent from the following detailed
description of illustrative embodiments thereof, which is to be
read in connection with the accompanying drawings. The various
features of the drawings are not to scale as the illustrations are
for clarity in facilitating one skilled in the art in understanding
the invention in conjunction with the detailed description. In the
drawings:
[0005] FIG. 1 illustrates an exemplary networked computer
environment according to at least one embodiment;
[0006] FIGS. 2A-2C are an operational flowchart illustrating a
multi-document summarization process according to at least one
embodiment;
[0007] FIG. 3 is a block diagram of internal and external
components of computers and servers depicted in FIG. 1 according to
at least one embodiment;
[0008] FIG. 4 depicts a cloud computing environment according to an
embodiment of the present invention; and
[0009] FIG. 5 depicts abstraction model layers according to an
embodiment of the present invention.
DETAILED DESCRIPTION
[0010] Detailed embodiments of the claimed structures and methods
are disclosed herein; however, it can be understood that the
disclosed embodiments are merely illustrative of the claimed
structures and methods that may be embodied in various forms. This
invention may, however, be embodied in many different forms and
should not be construed as limited to the exemplary embodiments set
forth herein. In the description, details of well-known features
and techniques may be omitted to avoid unnecessarily obscuring the
presented embodiments.
[0011] Embodiments of the present invention relate to the field of
computing, and more particularly to natural language processing.
The following described exemplary embodiments provide a system,
method, and program product to, among other things, extract
sentences from a corpus of documents to use in generating a summary
of the documents that satisfies a user-defined goal and a provided
summary constraint. Therefore, the present embodiment has the
capacity to improve the technical field of natural language
processing by utilizing centroid-level analysis to provide a high
quality extractive-based, unsupervised text summary without
requiring domain knowledge.
[0012] As previously described, natural language processing may
relate to a field within computer science dealing with the analysis
of natural language text by a computer. Natural language processing
may have applications in various categories, such as syntax,
semantics, discourse, and speech. Various applications within the
mentioned categories include text-to-speech, discourse analysis,
relationship extraction, natural language generation, sentiment
analysis, parsing, word segmentation, terminology extraction,
machine translation, and summarization.
[0013] Within the field of natural language processing,
summarization, or automatic text summarization, may relate to
extracting fragments of text from a corpus that condenses the
corpus to a shorter length that retains the crux of the corpus.
Various methods of summarization currently exist that can be
categorized into two main dimensions: extractive vs. abstractive
and supervised vs. unsupervised. Extractive-based summarization
relates to extracting objects from a corpus without modification.
For example, whole sentences may be removed from a document without
modification to create a short summary of the document.
Abstraction-based summarization relates to paraphrasing sections of
a document to produce a document summary. Typically,
abstraction-based summarization programs are more difficult to
develop than extractive-based summarization programs.
[0014] Supervised summarization relates to using a model and
training data to select the most appropriate text fragments to
include within a summary. Conversely, unsupervised summarization
relies on an algorithm to identify and select the text fragments to
include in a summary rather than relying on, usually a large amount
of, training data. While supervised methods have typically provided
better quality, such methods require more domain knowledge than
unsupervised summarization methods. As such, it may be advantageous
to, among other things, utilize an extractive-based, unsupervised
summarization program to generate a text summarization that
produces a quality summary without necessitating the presence of
training data.
[0015] According to one embodiment, a set of sentences may be
extracted from various documents within a corpus to be compiled
into a summary that satisfies a user-preconfigured summary length
constraint. A set of summaries may be iteratively generated so that
each may be evaluated and assigned a quality score based on how
accurately each summary satisfies one or more quality goals such as
how the summary adequately describes the corpus, the user-entered
query statement etc. An efficient global optimization
fully-polynomial randomized approximation scheme (FPRAS) based on
the cross-entropy method may be utilized to generate each summary
that maximizes the quality goal while satisfying the summary length
constraint.
[0016] The present invention may be a system, a method, and/or a
computer program product at any possible technical detail level of
integration. The computer program product may include a computer
readable storage medium (or media) having computer readable program
instructions thereon for causing a processor to carry out aspects
of the present invention.
[0017] The computer readable storage medium can be a tangible
device that can retain and store instructions for use by an
instruction execution device. The computer readable storage medium
may be, for example, but is not limited to, an electronic storage
device, a magnetic storage device, an optical storage device, an
electromagnetic storage device, a semiconductor storage device, or
any suitable combination of the foregoing. A non-exhaustive list of
more specific examples of the computer readable storage medium
includes the following: a portable computer diskette, a hard disk,
a random access memory (RAM), a read-only memory (ROM), an erasable
programmable read-only memory (EPROM or Flash memory), a static
random access memory (SRAM), a portable compact disc read-only
memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a
floppy disk, a mechanically encoded device such as punch-cards or
raised structures in a groove having instructions recorded thereon,
and any suitable combination of the foregoing. A computer readable
storage medium, as used herein, is not to be construed as being
transitory signals per se, such as radio waves or other freely
propagating electromagnetic waves, electromagnetic waves
propagating through a waveguide or other transmission media (e.g.,
light pulses passing through a fiber-optic cable), or electrical
signals transmitted through a wire.
[0018] Computer readable program instructions described herein can
be downloaded to respective computing/processing devices from a
computer readable storage medium or to an external computer or
external storage device via a network, for example, the Internet, a
local area network, a wide area network and/or a wireless network.
The network may comprise copper transmission cables, optical
transmission fibers, wireless transmission, routers, firewalls,
switches, gateway computers and/or edge servers. A network adapter
card or network interface in each computing/processing device
receives computer readable program instructions from the network
and forwards the computer readable program instructions for storage
in a computer readable storage medium within the respective
computing/processing device.
[0019] Computer readable program instructions for carrying out
operations of the present invention may be assembler instructions,
instruction-set-architecture (ISA) instructions, machine
instructions, machine dependent instructions, microcode, firmware
instructions, state-setting data, configuration data for integrated
circuitry, or either source code or object code written in any
combination of one or more programming languages, including an
object oriented programming language such as Smalltalk, C++, or the
like, and procedural programming languages, such as the "C"
programming language or similar programming languages. The computer
readable program instructions may execute entirely on the user's
computer, partly on the user's computer, as a stand-alone software
package, partly on the user's computer and partly on a remote
computer or entirely on the remote computer or server. In the
latter scenario, the remote computer may be connected to the user's
computer through any type of network, including a local area
network (LAN) or a wide area network (WAN), or the connection may
be made to an external computer (for example, through the Internet
using an Internet Service Provider). In some embodiments,
electronic circuitry including, for example, programmable logic
circuitry, field-programmable gate arrays (FPGA), or programmable
logic arrays (PLA) may execute the computer readable program
instructions by utilizing state information of the computer
readable program instructions to personalize the electronic
circuitry, in order to perform aspects of the present
invention.
[0020] Aspects of the present invention are described herein with
reference to flowchart illustrations and/or block diagrams of
methods, apparatus (systems), and computer program products
according to embodiments of the invention. It will be understood
that each block of the flowchart illustrations and/or block
diagrams, and combinations of blocks in the flowchart illustrations
and/or block diagrams, can be implemented by computer readable
program instructions.
[0021] These computer readable program instructions may be provided
to a processor of a general purpose computer, special purpose
computer, or other programmable data processing apparatus to
produce a machine, such that the instructions, which execute via
the processor of the computer or other programmable data processing
apparatus, create means for implementing the functions/acts
specified in the flowchart and/or block diagram block or blocks.
These computer readable program instructions may also be stored in
a computer readable storage medium that can direct a computer, a
programmable data processing apparatus, and/or other devices to
function in a particular manner, such that the computer readable
storage medium having instructions stored therein comprises an
article of manufacture including instructions which implement
aspects of the function/act specified in the flowchart and/or block
diagram block or blocks.
[0022] The computer readable program instructions may also be
loaded onto a computer, other programmable data processing
apparatus, or other device to cause a series of operational steps
to be performed on the computer, other programmable apparatus or
other device to produce a computer implemented process, such that
the instructions which execute on the computer, other programmable
apparatus, or other device implement the functions/acts specified
in the flowchart and/or block diagram block or blocks.
[0023] The flowchart and block diagrams in the Figures illustrate
the architecture, functionality, and operation of possible
implementations of systems, methods, and computer program products
according to various embodiments of the present invention. In this
regard, each block in the flowchart or block diagrams may represent
a module, segment, or portion of instructions, which comprises one
or more executable instructions for implementing the specified
logical function(s). In some alternative implementations, the
functions noted in the blocks may occur out of the order noted in
the Figures. For example, two blocks shown in succession may, in
fact, be executed substantially concurrently, or the blocks may
sometimes be executed in the reverse order, depending upon the
functionality involved. It will also be noted that each block of
the block diagrams and/or flowchart illustration, and combinations
of blocks in the block diagrams and/or flowchart illustration, can
be implemented by special purpose hardware-based systems that
perform the specified functions or acts or carry out combinations
of special purpose hardware and computer instructions.
[0024] The following described exemplary embodiments provide a
system, method, and program product to generate a query-focused
multi-document summarization using an extractive-based,
unsupervised summarization program.
[0025] Referring to FIG. 1, an exemplary networked computer
environment 100 is depicted, according to at least one embodiment.
The networked computer environment 100 may include client computing
device 102 and a server 112 interconnected via a communication
network 114. According to at least one implementation, the
networked computer environment 100 may include a plurality of
client computing devices 102 and servers 112, of which only one of
each is shown for illustrative brevity.
[0026] The communication network 114 may include various types of
communication networks, such as a wide area network (WAN), local
area network (LAN), a telecommunication network, a wireless
network, a public switched network and/or a satellite network. The
communication network 114 may include connections, such as wire,
wireless communication links, or fiber optic cables. It may be
appreciated that FIG. 1 provides only an illustration of one
implementation and does not imply any limitations with regard to
the environments in which different embodiments may be implemented.
Many modifications to the depicted environments may be made based
on design and implementation requirements.
[0027] Client computing device 102 may include a processor 104 and
a data storage device 106 that is enabled to host and run a
software program 108 and a multi-document summarization program
110A and communicate with the server 112 via the communication
network 114, in accordance with one embodiment of the invention.
Client computing device 102 may be, for example, a mobile device, a
telephone, a personal digital assistant, a netbook, a laptop
computer, a tablet computer, a desktop computer, or any type of
computing device capable of running a program and accessing a
network. As will be discussed with reference to FIG. 3, the client
computing device 102 may include internal components 302a and
external components 304a, respectively.
[0028] The server computer 112 may be a laptop computer, netbook
computer, personal computer (PC), a desktop computer, or any
programmable electronic device or any network of programmable
electronic devices capable of hosting and running a multi-document
summarization program 110B and a database 116 and communicating
with the client computing device 102 via the communication network
114, in accordance with embodiments of the invention. As will be
discussed with reference to FIG. 3, the server computer 112 may
include internal components 302b and external components 304b,
respectively. The server 112 may also operate in a cloud computing
service model, such as Software as a Service (SaaS), Platform as a
Service (PaaS), or Infrastructure as a Service (IaaS). The server
112 may also be located in a cloud computing deployment model, such
as a private cloud, community cloud, public cloud, or hybrid
cloud.
[0029] According to the present embodiment, the multi-document
summarization program 110A, 110B may be a program capable of
analyzing a corpus to extract sentences that satisfy a quality
goal, or threshold, as determined from a query statement. The
extracted sentences may be used to generate multiple summaries that
are each evaluated for quality and filtered out if an assigned
quality score does not satisfy a quality threshold. Once a stopping
criterion is met, the multi-document summarization program 110A,
110B may present the highest quality score summary to a user. The
collaborative moment method is explained in further detail below
with respect to FIG. 2.
[0030] Referring now to FIGS. 2A-2C, an operational flowchart
illustrating a multi-document summarization process 200 is depicted
according to at least one embodiment. In FIG. 2A, at 202, the
multi-document summarization program 110A, 110B receives a query
statement, one or more documents, one or more summary constraints,
and one or more goals. The query statement may be a user-entered
series of words that relates to a particular topic or asks a
question towards which the user desires more information. For
example, a user may enter the query statement "What is the state of
the United States economy?" and expect results relating to the
current state of the United States economy. The one or more
documents may be a corpus of information in which the
multi-document summarization program 110A, 110B may search within
based on the query statement. The one or more documents may be
articles, stories, papers, repositories, or any other media with
which the multi-document summarization program 110A, 110B may be
capable of searching and summarizing. The one or more summary
constraints may be factors that impact the generated summary. For
example, a length constraint may limit each generated summary to a
particular word count. Similarly, each generated summary may be
limited to a preconfigured number of sentences. The one or more
goals may be criteria the multi-document summarization program
110A, 110B may aim to achieve when drafting the summary of the one
or more documents. For example, the user may establish a goal for
the multi-document summarization program 110A, 110B to generate a
summary that is comprised of sentences as short as possible but
covers as many subjects as possible from the one or more documents.
Whereas the summary constraints are fixed and must be adhered to by
the multi-document summarization program 110A, 110B when generating
the summary, the one or more goals may be more fluid to allow
various generated summaries that each may be evaluated for
quality.
[0031] Then, at 204, the multi-document summarization program 110A,
110B identifies keywords in the received query statement. In order
to effectively generate a summary that answers the query statement,
the multi-document summarization program 110A, 110B may identify
keywords within the query statement using known text analysis
techniques. For example, in the previously discussed query
statement, "What is the state of the United States economy?", the
words "state" and "economy" and the phrase "United States" may be
identified as keywords since each relates to an important aspect of
the query statement.
[0032] Next, at 206, the multi-document summarization program 110A,
110B generates an expanded query statement using known query
expansion techniques on the identified keywords. Query expansion
may relate to the process of finding other words or phrases that
are relevant to the query statement or identified keywords in an
attempt to improve retrieved results. For example, for the
identified keywords "state", "economy", and "United States", the
multi-document summarization program 110A, 110B may use known query
expansion techniques, such as RM3, to identify the related keywords
"dollar" and "unemployment." Therefore, sentences containing the
keywords "dollar" and "unemployment" may contain information that
is relevant to the identified keywords "state", "economy", and
"United States".
[0033] Then, at 208, the multi-document summarization program 110A,
110B performs sentence selection on the one or more documents. The
multi-document summarization program 110A, 110B may perform the
sentence selection process by parsing each received document to
identify one or more sentences relevant to the expanded query
statement. The multi-document summarization program 110A, 110B may
perform sentence selection as the first step of a constrained
global optimization problem where the goal may be to find a subset
of sentences within the one or more documents that maximize a given
quality target function. In order to identify the sentences during
the first step of the constrained global optimization problem, the
multi-document summarization program 110A, 110B may parse through
each received document to identify and extract sentences according
to known sentence filtering techniques, such as keyword
identification or similarity score rating. For example, using a
keyword identification approach, the multi-document summarization
program 110A, 110B may analyze the received documents to identify
each sentence that contains or is related to the keywords "state",
"economy", "United States", "dollar", and "unemployment." Each
sentence identified during sentence selection may be a candidate to
be included within a generated summary.
[0034] Then, at 210, the multi-document summarization program 110A,
110B generates candidate summaries of the documents using the
sentences identified during sentence selection. The multi-document
summarization program 110A, 110B may utilize an unsupervised
approach based on the Cross Entropy Method to direct the generation
towards better quality summaries. The Cross Entropy Method is a
Monte-Carlo framework utilizing a Stochastic algorithm that is
capable of solving hard combinatorial optimization problems based
on rare event estimation. A Stochastic algorithm is an iterative
optimization method that utilizes random variables. For example,
the multi-document summarization program 110A, 110B may generate
100 summaries by selecting individual sentences from the identified
sentences using a random variable for each sentence representing
the sentence's chance of being selected for inclusion in a
candidate summary. Furthermore, the total number of summaries
generated by the multi-document summarization program 110A, 110B
may be preconfigured by a user or based on system limitations, such
as computing power.
[0035] In at least one embodiment, the multi-document summarization
program 110A, 110B may generate the candidate summaries without
consideration for the preconfigured summary constraints, and filter
out the candidate summaries that do not satisfy the summary
constraints once generation is complete. For example, if a summary
constraint establishes all generated summaries are to be 10
sentences or less, the multi-document summarization program 110A,
110B may generate a candidate summary that is 11 sentences in
length and delete the 11-sentence candidate summary through a
filtering process once all candidate summaries in the iteration are
generated.
[0036] The generation of candidate summaries may be an iterative
process performed by the multi-document summarization program 110A,
110B. When proceeding through a first iteration of generating
candidate summaries, the multi-document summarization program 110A,
110B may apply equal probability to all sentences selected in step
208. Therefore, each sentence may have an equal chance of being
included in a candidate summary. However, in subsequent iterations,
the multi-document summarization program 110A, 110B may apply an
updated algorithm from step 224 that increases the selection
probability for sentences that appear in summaries receiving high
quality scores.
[0037] Referring now to FIG. 2B, at 212, the multi-document
summarization program 110A, 110B calculates a quality score for
each generated summary based on quality features formulated in the
one or more received goals. After generating each of the summaries,
the multi-document summarization program 110A, 110B may evaluate
each candidate summary by assigning a score to each generated
summary that evaluates each summary's quality as represented
through a quality goals. The score may be any rating system of
evaluating quality, such as a 0-100 point system or a letter grade
system. The quality features utilized by the multi-document
summarization program 110A, 110B may include measuring the
Bhattacharyya similarity between the unigram language model (LM) of
the query and the unigram LM of the summary, measuring the relative
mass that the summary devotes to the query, measuring to what
extent the summary generally covers the one or more documents,
measuring each summary's sentence diversity by calculating the
bigram LM entropy for each summary, biasing sentence selection
towards sentences that appear earlier in the containing document,
and biasing sentence selection towards longer summaries that still
satisfy the length constraint that contain few long sentences
rather than summaries that contain many short sentences.
[0038] Then, at 214, multi-document summarization program 110A,
110B determines whether a preconfigured number of iterations have
been performed. According to one implementation, the multi-document
summarization process 200 may continue along the operational
flowchart if a preconfigured number of iterations have been
performed. The multi-document summarization program 110A, 110B may
iterate through summary generation a number of times so that a
sufficient number of summaries may be generated to provide an
accurate, quality summary. If the multi-document summarization
program 110A, 110B determines a preconfigured number of iterations
have been performed (step 214, "Yes" branch), the multi-document
summarization program 110A, 110B may identify the candidate summary
with the highest calculated quality score from step 212 and the
multi-document summarization process 200 may continue to step 216
to restructure the identified summary. If the multi-document
summarization program 110A, 110B determines a preconfigured number
of iterations have not been performed (step 214, "No" branch), the
multi-document summarization process 200 may continue to step 220
to filter out each evaluated candidate summary that does not
satisfy a filter threshold.
[0039] In another embodiment, the multi-document summarization
program 110A, 110B may determine whether an improvement to the
target function is made. According to one implementation, the
multi-document summarization process 200 may continue along the
operational flowchart if no improvements are made to the target
function thereby improving the quality of the generated summaries.
If the multi-document summarization program 110A, 110B determines
no improvement to the target function is made, the multi-document
summarization program 110A, 110B may identify the candidate summary
with the highest calculated quality score and the multi-document
summarization process 200 may continue to step 216 to restructure
the chosen summary. If the multi-document summarization program
110A, 110B determines an improvement to the target function is
made, the multi-document summarization program 110A, 110B may
determine higher quality summaries may be generated using the
updated target function and the multi-document summarization
process 200 may continue to step 220 to filter out each evaluated
candidate summary that does not satisfy a filter threshold.
[0040] In yet another embodiment, the multi-document summarization
program 110A, 110B may determine whether a quality score for a
candidate summary satisfies a quality score threshold. According to
one implementation, the multi-document summarization process 200
may continue along the operational flowchart if the quality score
associated with a candidate summary satisfies a quality score
threshold. The multi-document summarization program 110A, 110B may
compare the calculated quality score for each candidate summary
against a user preconfigured quality score threshold to determine
if the quality score associated with any of the evaluated candidate
summaries satisfy the threshold. If the multi-document
summarization program 110A, 110B determines a quality score does
satisfy the quality score threshold, the multi-document
summarization program 110A, 110B may determine that the candidate
summary associated with the quality score should be selected as a
quality summary of the one or more documents and the multi-document
summarization process 200 may continue to step 216 to restructure
the chosen summary. If the multi-document summarization program
110A, 110B determines a quality score does not satisfy the quality
score threshold, the multi-document summarization program 110A,
110B may determine that none of the candidate summaries provide a
quality summary of the one or more documents and the multi-document
summarization process 200 may continue to step 220 to filter out
each evaluated candidate summary that does not satisfy a filter
threshold.
[0041] Next, at 216, the multi-document summarization program 110A,
110B restructures the selected summary. Once the multi-document
summarization program 110A, 110B determines that a summary should
be selected based on the associated quality score satisfying the
quality score threshold, the multi-document summarization program
110A, 110B may analyze the sentence organization of the selected
summary and accordingly restructure the sentence organization using
known natural language processing techniques based on various
preconfigured characteristics, such as sentence length, summary
readability, sentence position in the document, which document the
sentence belongs to, etc.
[0042] Then, at 218, the multi-document summarization program 110A,
110B presents the restructured candidate summary to a user. Once
the multi-document summarization program 110A, 110B restructures
the selected summary, the multi-document summarization program
110A, 110B may present the selected summary to a user for review.
The multi-document summarization program 110A, 110B may present the
selected summary to the user by displaying the selected summary on
a graphical user interface of a user device, such as a client
computing device 102.
[0043] Referring now to FIG. 2C, at 220, if the multi-document
summarization program 110A, 110B determines a quality score does
not satisfy the quality score threshold, the multi-document
summarization program 110A, 110B extracts each evaluated candidate
summary that satisfies a filter threshold. The filter threshold may
be a benchmark which the assigned quality score must meet or exceed
in order to remain in the group of candidate summaries being
analyzed by the document summarization program 110A, 110B. For
example, if a numerical 0-100 grading system was used to evaluate
the quality of each summary and the quality threshold is set at 85,
the multi-document summarization program 110A, 110B may extract the
candidate summaries with corresponding quality scores at or above
85 for further analysis as to what information within the extracted
candidate summaries allowed each summary to receive a high quality
score. Similarly, if the quality threshold is set to 10%, the
multi-document summarization program 110A, 110B may extract all
evaluated summaries with quality scores that are within the top 10%
of all quality scores within the group of candidate summaries.
[0044] Next, at 222, the multi-document summarization program 110A,
110B identifies frequently appearing sentences in the candidate
summaries remaining in the group. Of the extracted candidate
summaries that do satisfy the filter threshold, the multi-document
summarization program 110A, 110B may analyze each remaining
candidate summary to identify sentences appearing in multiple
candidate summaries with high quality scores. If a sentence appears
in multiple summaries with high quality scores, the sentence may be
associated with a quality summary for the query statement and
should be included in a quality summary. Therefore, the
multi-document summarization program 110A, 110B may determine the
repeated use of a sentence in multiple high quality score summaries
as an indication that the sentence should be included in the
generation of subsequent candidate summaries to achieve a high
quality summary.
[0045] Then, at 224, the multi-document summarization program 110A,
110B updates the algorithm used to generate candidate summaries
based on the identified frequently appearing sentences. Once the
multi-document summarization program 110A, 110B has evaluated the
candidate summaries and identified the frequently appearing
sentences, the multi-document summarization program 110A, 110B may
update the probabilities of the random variables used to generate
candidate summaries in step 210 so that the frequently appearing
sentences have higher probabilities and, therefore, more likely to
be selected for inclusion in a candidate summary. Therefore, unlike
the first iteration of candidate summary generation where each
identified sentence had an equal chance of being selected for a
candidate summary, the updated algorithm may increase the
probability of the identified frequently appearing sentences
thereby increasing the chance that the identified sentences are
chosen for a candidate summary and improving the quality of each
generated candidate summary. Once the candidate summary generation
algorithm is updated, the multi-document summarization process 200
may return to step 210 to generate candidate summaries of the one
or more documents using the extracted sentences.
[0046] It may be appreciated that FIG. 2 provides only an
illustration of one implementation and does not imply any
limitations with regard to how different embodiments may be
implemented. Many modifications to the depicted environments may be
made based on design and implementation requirements.
[0047] FIG. 3 is a block diagram 300 of internal and external
components of the client computing device 102 and the server 112
depicted in FIG. 1 in accordance with an embodiment of the present
invention. It should be appreciated that FIG. 3 provides only an
illustration of one implementation and does not imply any
limitations with regard to the environments in which different
embodiments may be implemented. Many modifications to the depicted
environments may be made based on design and implementation
requirements.
[0048] The data processing system 302, 304 is representative of any
electronic device capable of executing machine-readable program
instructions. The data processing system 302, 304 may be
representative of a smart phone, a computer system, PDA, or other
electronic devices. Examples of computing systems, environments,
and/or configurations that may represented by the data processing
system 302, 304 include, but are not limited to, personal computer
systems, server computer systems, thin clients, thick clients,
hand-held or laptop devices, multiprocessor systems,
microprocessor-based systems, network PCs, minicomputer systems,
and distributed cloud computing environments that include any of
the above systems or devices.
[0049] The client computing device 102 and the server 112 may
include respective sets of internal components 302 a,b and external
components 304 a,b illustrated in FIG. 3. Each of the sets of
internal components 302 include one or more processors 320, one or
more computer-readable RAMs 322, and one or more computer-readable
ROMs 324 on one or more buses 326, and one or more operating
systems 328 and one or more computer-readable tangible storage
devices 330. The one or more operating systems 328, the software
program 108 and the multi-document summarization program 110A in
the client computing device 102 and the multi-document
summarization program 110B in the server 112 are stored on one or
more of the respective computer-readable tangible storage devices
330 for execution by one or more of the respective processors 320
via one or more of the respective RAMs 322 (which typically include
cache memory). In the embodiment illustrated in FIG. 3, each of the
computer-readable tangible storage devices 330 is a magnetic disk
storage device of an internal hard drive. Alternatively, each of
the computer-readable tangible storage devices 330 is a
semiconductor storage device such as ROM 324, EPROM, flash memory
or any other computer-readable tangible storage device that can
store a computer program and digital information.
[0050] Each set of internal components 302 a,b also includes a R/W
drive or interface 332 to read from and write to one or more
portable computer-readable tangible storage devices 338 such as a
CD-ROM, DVD, memory stick, magnetic tape, magnetic disk, optical
disk or semiconductor storage device. A software program, such as
the multi-document summarization program 110A, 110B, can be stored
on one or more of the respective portable computer-readable
tangible storage devices 338, read via the respective R/W drive or
interface 332, and loaded into the respective hard drive 330.
[0051] Each set of internal components 302 a,b also includes
network adapters or interfaces 336 such as a TCP/IP adapter cards,
wireless Wi-Fi interface cards, or 3G or 4G wireless interface
cards or other wired or wireless communication links. The software
program 108 and the multi-document summarization program 110A in
the client computing device 102 and the multi-document
summarization program 110B in the server 112 can be downloaded to
the client computing device 102 and the server 112 from an external
computer via a network (for example, the Internet, a local area
network or other, wide area network) and respective network
adapters or interfaces 336. From the network adapters or interfaces
336, the software program 108 and the multi-document summarization
program 110A in the client computing device 102 and the
multi-document summarization program 110B in the server 112 are
loaded into the respective hard drive 330. The network may comprise
copper wires, optical fibers, wireless transmission, routers,
firewalls, switches, gateway computers and/or edge servers.
[0052] Each of the sets of external components 304 a,b can include
a computer display monitor 344, a keyboard 342, and a computer
mouse 334. External components 304 a,b can also include touch
screens, virtual keyboards, touch pads, pointing devices, and other
human interface devices. Each of the sets of internal components
302 a,b also includes device drivers 340 to interface to computer
display monitor 344, keyboard 342, and computer mouse 334. The
device drivers 340, R/W drive or interface 332, and network adapter
or interface 336 comprise hardware and software (stored in storage
device 330 and/or ROM 324).
[0053] It is understood in advance that although this disclosure
includes a detailed description on cloud computing, implementation
of the teachings recited herein are not limited to a cloud
computing environment. Rather, embodiments of the present invention
are capable of being implemented in conjunction with any other type
of computing environment now known or later developed.
[0054] Cloud computing is a model of service delivery for enabling
convenient, on-demand network access to a shared pool of
configurable computing resources (e.g. networks, network bandwidth,
servers, processing, memory, storage, applications, virtual
machines, and services) that can be rapidly provisioned and
released with minimal management effort or interaction with a
provider of the service. This cloud model may include at least five
characteristics, at least three service models, and at least four
deployment models.
[0055] Characteristics are as follows:
[0056] On-demand self-service: a cloud consumer can unilaterally
provision computing capabilities, such as server time and network
storage, as needed automatically without requiring human
interaction with the service's provider.
[0057] Broad network access: capabilities are available over a
network and accessed through standard mechanisms that promote use
by heterogeneous thin or thick client platforms (e.g., mobile
phones, laptops, and PDAs).
[0058] Resource pooling: the provider's computing resources are
pooled to serve multiple consumers using a multi-tenant model, with
different physical and virtual resources dynamically assigned and
reassigned according to demand. There is a sense of location
independence in that the consumer generally has no control or
knowledge over the exact location of the provided resources but may
be able to specify location at a higher level of abstraction (e.g.,
country, state, or datacenter).
[0059] Rapid elasticity: capabilities can be rapidly and
elastically provisioned, in some cases automatically, to quickly
scale out and rapidly released to quickly scale in. To the
consumer, the capabilities available for provisioning often appear
to be unlimited and can be purchased in any quantity at any
time.
[0060] Measured service: cloud systems automatically control and
optimize resource use by leveraging a metering capability at some
level of abstraction appropriate to the type of service (e.g.,
storage, processing, bandwidth, and active user accounts). Resource
usage can be monitored, controlled, and reported providing
transparency for both the provider and consumer of the utilized
service.
[0061] Service Models are as follows:
[0062] Software as a Service (SaaS): the capability provided to the
consumer is to use the provider's applications running on a cloud
infrastructure. The applications are accessible from various client
devices through a thin client interface such as a web browser
(e.g., web-based e-mail). The consumer does not manage or control
the underlying cloud infrastructure including network, servers,
operating systems, storage, or even individual application
capabilities, with the possible exception of limited user-specific
application configuration settings.
[0063] Platform as a Service (PaaS): the capability provided to the
consumer is to deploy onto the cloud infrastructure
consumer-created or acquired applications created using programming
languages and tools supported by the provider. The consumer does
not manage or control the underlying cloud infrastructure including
networks, servers, operating systems, or storage, but has control
over the deployed applications and possibly application hosting
environment configurations.
[0064] Infrastructure as a Service (IaaS): the capability provided
to the consumer is to provision processing, storage, networks, and
other fundamental computing resources where the consumer is able to
deploy and run arbitrary software, which can include operating
systems and applications. The consumer does not manage or control
the underlying cloud infrastructure but has control over operating
systems, storage, deployed applications, and possibly limited
control of select networking components (e.g., host firewalls).
[0065] Deployment Models are as follows:
[0066] Private cloud: the cloud infrastructure is operated solely
for an organization. It may be managed by the organization or a
third party and may exist on-premises or off-premises.
[0067] Community cloud: the cloud infrastructure is shared by
several organizations and supports a specific community that has
shared concerns (e.g., mission, security requirements, policy, and
compliance considerations). It may be managed by the organizations
or a third party and may exist on-premises or off-premises.
[0068] Public cloud: the cloud infrastructure is made available to
the general public or a large industry group and is owned by an
organization selling cloud services.
[0069] Hybrid cloud: the cloud infrastructure is a composition of
two or more clouds (private, community, or public) that remain
unique entities but are bound together by standardized or
proprietary technology that enables data and application
portability (e.g., cloud bursting for load-balancing between
clouds).
[0070] A cloud computing environment is service oriented with a
focus on statelessness, low coupling, modularity, and semantic
interoperability. At the heart of cloud computing is an
infrastructure comprising a network of interconnected nodes.
[0071] Referring now to FIG. 4, illustrative cloud computing
environment 50 is depicted. As shown, cloud computing environment
50 comprises one or more cloud computing nodes 100 with which local
computing devices used by cloud consumers, such as, for example,
personal digital assistant (PDA) or cellular telephone 54A, desktop
computer 54B, laptop computer 54C, and/or automobile computer
system 54N may communicate. Nodes 100 may communicate with one
another. They may be grouped (not shown) physically or virtually,
in one or more networks, such as Private, Community, Public, or
Hybrid clouds as described hereinabove, or a combination thereof.
This allows cloud computing environment 50 to offer infrastructure,
platforms and/or software as services for which a cloud consumer
does not need to maintain resources on a local computing device. It
is understood that the types of computing devices 54A-N shown in
FIG. 4 are intended to be illustrative only and that computing
nodes 100 and cloud computing environment 50 can communicate with
any type of computerized device over any type of network and/or
network addressable connection (e.g., using a web browser).
[0072] Referring now to FIG. 5, a set of functional abstraction
layers 500 provided by cloud computing environment 50 is shown. It
should be understood in advance that the components, layers, and
functions shown in FIG. 5 are intended to be illustrative only and
embodiments of the invention are not limited thereto. As depicted,
the following layers and corresponding functions are provided:
[0073] Hardware and software layer 60 includes hardware and
software components. Examples of hardware components include:
mainframes 61; RISC (Reduced Instruction Set Computer) architecture
based servers 62; servers 63; blade servers 64; storage devices 65;
and networks and networking components 66. In some embodiments,
software components include network application server software 67
and database software 68.
[0074] Virtualization layer 70 provides an abstraction layer from
which the following examples of virtual entities may be provided:
virtual servers 71; virtual storage 72; virtual networks 73,
including virtual private networks; virtual applications and
operating systems 74; and virtual clients 75.
[0075] In one example, management layer 80 may provide the
functions described below. Resource provisioning 81 provides
dynamic procurement of computing resources and other resources that
are utilized to perform tasks within the cloud computing
environment. Metering and Pricing 82 provide cost tracking as
resources are utilized within the cloud computing environment, and
billing or invoicing for consumption of these resources. In one
example, these resources may comprise application software
licenses. Security provides identity verification for cloud
consumers and tasks, as well as protection for data and other
resources. User portal 83 provides access to the cloud computing
environment for consumers and system administrators. Service level
management 84 provides cloud computing resource allocation and
management such that required service levels are met. Service Level
Agreement (SLA) planning and fulfillment 85 provide pre-arrangement
for, and procurement of, cloud computing resources for which a
future requirement is anticipated in accordance with an SLA.
[0076] Workloads layer 90 provides examples of functionality for
which the cloud computing environment may be utilized. Examples of
workloads and functions which may be provided from this layer
include: mapping and navigation 91; software development and
lifecycle management 92; virtual classroom education delivery 93;
data analytics processing 94; transaction processing 95; and
multi-document summarization 96. Multi-document summarization may
relate to selecting sentences from within a corpus, generating
various candidate summaries of the corpus using a cross entropy
method, calculating a quality score for each of the various
candidate summaries, presenting, to a user, the summary with the
highest calculated quality score that also satisfies a quality
score threshold.
[0077] The descriptions of the various embodiments of the present
invention have been presented for purposes of illustration, but are
not intended to be exhaustive or limited to the embodiments
disclosed. Many modifications and variations will be apparent to
those of ordinary skill in the art without departing from the scope
of the described embodiments. The terminology used herein was
chosen to best explain the principles of the embodiments, the
practical application or technical improvement over technologies
found in the marketplace, or to enable others of ordinary skill in
the art to understand the embodiments disclosed herein.
* * * * *