WO2020161506A1 - Procédé et système de capture de métadonnées dans un objet de document ou un format de fichier - Google Patents
Procédé et système de capture de métadonnées dans un objet de document ou un format de fichier Download PDFInfo
- Publication number
- WO2020161506A1 WO2020161506A1 PCT/GB2020/050297 GB2020050297W WO2020161506A1 WO 2020161506 A1 WO2020161506 A1 WO 2020161506A1 GB 2020050297 W GB2020050297 W GB 2020050297W WO 2020161506 A1 WO2020161506 A1 WO 2020161506A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- text
- user
- document
- metadata
- input
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/194—Calculation of difference between files
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/16—File or folder operations, e.g. details of user interfaces specifically adapted to file systems
- G06F16/164—File meta data generation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/93—Document management systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/166—Editing, e.g. inserting or deleting
- G06F40/169—Annotation, e.g. comment data or footnotes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/197—Version control
Definitions
- the application relates to a system and method for capturing metadata in a document object.
- the system and method also provide a corresponding file format in which the metadata and document data can be stored, and a user interface and system for end-to- end editing and publishing of text documents with the attached metadata.
- the present invention described herein relates to a system and method for capturing metadata in a document object or file format.
- the document object or file format may be used to track and record inputs to a document as metadata during the process of writing the document.
- the document object or file format forms an audit trail of metadata detailing the steps taken by an author or authors of the document from conception and research related to the preliminary writing of the document, through to reviewing, editing, and the eventual publication of the document.
- the document object or file format may be used in an end-to-end authoring system for creating, editing and publishing documents.
- the authoring system is designed to analyse the sources used throughout the writing process and record them as metadata in the document object or file format. Once the document is written by the author or authors it may be published in the system.
- the system may be proprietary and secure, such that the editing of published documents is prohibited or tracked further using the document object or file format.
- Each published document is linked or published with the corresponding document object or file format for that particular document, providing the reader with an improved reading experience which is both more informative and more secure.
- the file format of the present invention comprises a logical organisation of data in memory.
- a portion of the memory receives and stores data input by a user including one or more of text, numerical data and graphics such as images, tables and graphs.
- Another portion of the memory receives and stores data indicative of one or more characteristics of the user input data, the characteristics including at least one of: designation of the user who inputted the user input data such as a user identification; a date and/or time when the user input data was inputted; a source location such as a uniform resource locator (URL) address, indicating where the user obtained the user input data and/or what processes or information the user accessed before and at the time of inputting the user input data; an indication of the method of input the user performed to input the user input data; and any edits or modifications to the user input data, performed by a user, since the original user input the user input data.
- URL uniform resource locator
- the file format may be in the form of a table, directed acyclic graph, or any other structure in which user input data is associated, either by position or by an explicitly defined link, to the characteristics of the user input data.
- the file format is updated to create new entries each time new data is input or when existing data is modified by a user.
- the file format is therefore configured to contain user input data as well as information concerning the input or modification of the data. This means that the file format of the present invention is very useful in tracking, as metadata, how a user inputs and modifies the content of a document during the writing process of the document.
- the file format is configured to be accessible and readable when opened in a reader application configured to read the file format. This may be a proprietary system, hence keeping the contents of the file format secure outside of the reader application.
- Figure 1 is a diagram of the system according to the present invention.
- Figure 2 is a diagram of an example linkage structure according to the present invention.
- Figures 3 to 7 are diagrams illustrating the structure of a logical file format for use with the present invention.
- Figure 8 is an illustration of a published document according to the present invention.
- Figure 9 is a flow diagram showing the method of the present invention.
- FIG. 10 is a detailed flow diagram showing the method of the present invention.
- the present invention relates to a file format or document object for storing data in a memory.
- the data comprises data input by a user, and metadata that describes characteristics relating to, amongst other things, how and when the user input data is inputted.
- the file format or document object is used to record user input data on a computer device and corresponding metadata for a document during the writing process of the document.
- the file format describes the structure and the way in which data is stored in a computer file.
- the computer file structured according to the file format of the present invention is saved to the memory of a computer readable medium.
- the memory may be implemented using any type of volatile or non-volatile memory devices, or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic or optical disk.
- SRAM static random access memory
- EEPROM electrically erasable programmable read-only memory
- EPROM erasable programmable read-only memory
- PROM programmable read-only memory
- ROM read-only memory
- magnetic memory a magnetic memory
- flash memory a flash memory
- magnetic or optical disk a magnetic or optical disk.
- the computer device used for inputting user input data may comprise a processor, and various peripheral interface modules, the peripheral interface modules being, for example, a keyboard, a click wheel, buttons, and the like.
- the computer device may also include a power supply component configured to implement power management on the device, a wired or wireless network interface configured to connect the device to a network, and an input/output interface.
- the device may operate based on an operating system such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, or the like stored in the memory.
- the computer device itself may be a personal computer, tablet computer, mobile device such as an e-reader or a mobile telephone, wearable technology or the like.
- the methods of the present invention relating to the file format, described herein, may be provided on a non-transitory computer readable storage medium including instructions, such as a memory including instructions, executable by the processor of the computer device.
- the non-transitory computer-readable storage medium may be a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disc, an optical data storage device, and the so on.
- Figure 1 illustrates a system 100 comprising a user interface module 101 and a data processing module 102.
- the user interface 101 comprises a text editing module 103 and a text publishing module 104.
- the data processing module 102 may be formed as part of a computer, server, or distributed computer network for example.
- the data processing 102 module includes or is communicatively coupled to a memory module 107. When data is recorded or saved to the data processing module 102, it is therefore saved to the memory module 107.
- the system 100 is communicatively coupled to a database 105.
- the user interface 101 and the data processing module 102 are configured to communicate with each other.
- the database 105 may be the internet as shown in Figure 1 , and accessed through a web browser.
- Content from the database 105 may include html data 105a.
- the web browser may include a database application 106, which is configured to communicate with the data processing module 102.
- the database may be a proprietary database containing data available for use by the analyst or researcher, or a combination of a proprietary database and the Internet accessed through the web browser including the database application 106.
- the system 100 provides an end-to-end platform for capturing metadata in a document object or file format, for a document in the process of being written.
- the system 100 tracks and records information that is used and accessed in the process of writing, and stores this information in a linkage structure.
- the linkage structure forms part of the file format or document object. This will be described later with respect to Figure 2 and a possible implementation of the linkage structure will be explained with reference to Figures 3 to 7.
- the linkage structure includes linkage metadata that describes not only where the written text of the document originates, but also how and if it is modified, and what actions are taken by an author during the process of writing.
- a document is published in the system 100, and the linkage structure or at least some of the linkage metadata is made available with the published document.
- the system 100 therefore provides an information-rich document which not only includes text, but also the linkage metadata that defines how the text of the published document was found, inputted and modified.
- the combination of the text with the linkage structure means that users of the system 100 who read the published document have access to much more information than regular text documents.
- the document object or file format provided by the invention may be made available to the public, so that it is possible to determine for any document how it was constructed. In alternative embodiments, the document object or file format may be kept confidential within a proprietary system.
- the document object or file format therefore provides attributions to the author or authors that are involved in the process of writing a document, starting from conception and research of the document through to the publication of the document.
- the document object or file format is configured to preserve, as metadata, information regarding the author interactions with the system 100 as well as information regarding sources that are deemed to be used by the author or are otherwise relevant to the document being written.
- the system 100 and the process of capturing metadata will now be described.
- the process starts by inputting data into the system 100.
- the text editing module 103 is configured to receive input data that is to be included in the published text document. Data is inputted into the text editing module 103 of the user interface 101 by a user 110 who performs one or more of three methods.
- the user 1 10 who inputs the data is known as the author. From here forth the input data is referred to as input text.
- the data inputted by the author may include text data, numerical data, graphical data, or any combination of text, numerical and graphical data.
- the first method of inputting text comprises directly inputting text into the text editing module 103.
- the author may therefore type, dictate, or otherwise input text into the text editor 103 of the user interface 101.
- the second method of inputting text comprises pasting text into the text editing module 103.
- the pasted text is copied from a piece of text hereby referred to as a source.
- the third method of inputting text comprises uploading all or a part of a predefined text document to the text editing module 103.
- the predefined text document may also be referred to as a source. This may be a PDF file or other graphical or image based file containing one or more of text or graphical information.
- the text-editing module may include a predetermined template in which the author inputs text into different sections.
- the text editing module 103 communicates with the data processing module 102 to determine whether any linkage metadata exists relating to the input text.
- the linkage metadata contains the source or sources used by the author as well as other related tracking data concerning the author of the input text.
- the author performs a search of the database 105 for sources relevant to the intended topic of the text document before inputting text.
- the database 105 may be any structure capable of storing information that is retrievable by the system 100, such as the internet accessed through a web browser as seen in Figure 1 , or a propriety database as discussed previously.
- the data processing module 102 tracks the actions of the author, the actions including any interactions with the database such as clicks, website visits and downloads for example.
- the actions of the author may be recorded in the linkage structure by the data processing module 102.
- the database application 106 When the author uses the internet as the database 105 to search for information from sources, the recording of author actions and the related data may be facilitated by the database application 106, which provides the communicative coupling between the database 105 and the data processing module 102.
- the database application is a web-browser extension, plugin or the like.
- the sources identified in the search are directly used for inputting text into the text editing module 103, by copying and pasting from a source or by uploading a source document respectively.
- a source may still be indirectly in the first method for inputting text, in that the author may obtain information from a source found in the search, and use said information for informing the text which they input into the text editing module 103. Therefore, a source may be used for each of the three methods for inputting text into the text editing module 103.
- the linkage structure contains linkage metadata, which describes how the source is used to inform the text inputted into the text editing module 103 of the user interface 101.
- the linkage structure maps each source used by the author from the database 105 to the corresponding inputted text.
- the linkage structure maps each separate piece of inputted text to its respective sources.
- the visit of the website will be captured by in the linkage structure as an author action.
- Data related to the website itself may also be recorded as a source, such as the html and text content of the visited website. This may then be stored in the linkage structure as the source corresponding to the user action of visiting the website.
- URL Uniform Resource Locator
- Sources are recorded by the linkage structure either at the time they are identified in the search or when the author inputs text, or an image for example, that relates to a particular source.
- the html data and other text content from the webpage may therefore be extracted and recorded in the linkage structure at the data processing module 102 at the time it is identified in the search.
- the data processing module 102 may communicate with the database 105 to search for a particular source relating to the inputted text, according to the linkage metadata of the linkage structure that corresponds to the captured author actions.
- the search of the database 105 may be done only according to linkage metadata that is first determined to be relevant to the inputted text.
- the text editing module 103 may communicate the topic of ‘the economy’ with the data processing module 102, such that the data processing module can identify any linkage metadata relating to the topic.
- the data processing module 102 may then communicate with the database, through the database application 106 or otherwise, to identify any sources relating to the linkage metadata concerning the topic of‘the economy’. If sources, such as a website previously visited by the author, are found in this way they are then recorded to the data processing module 102.
- This alternative method may be considered more efficient because source information is only recorded to the data processing module 102 once it is determined that it has been used to inform the inputted text.
- the inputted text may be compared, semantically or otherwise, to source information recorded in the data processing module 102. If the comparison determines that there is a match between a portion of the inputted text and the source information, then the linkage structure is updated with linkage metadata to reflect that the portion of inputted text was informed from the specific source information. It is to be understood that a‘match’ does not necessarily mean that text from the source has been directly copied in the inputted text. Alternatively, a match may occur if there are semantically similar words, sentences or statistics in the inputted text as in the text from the source.
- Determining that there is a match between the inputted text and the text from the source may be subject to a similarity threshold score, above which the inputted text is identified as a match of the text from the source.
- the linkage metadata may also reflect who the author was that inputted the text, what changes, if any, they made to the original source information, and when the text was inputted compared to when the source information was accessed.
- comparison between the inputted text and the source information may be performed according to any text or language based recognition algorithm, examples of which will not be discussed here.
- the database application 106 tracks and captures the copy action of the author, and sends the copied text to the data processing module 102.
- the copied text is sent to the data processing module 102 with linkage metadata that includes the URL related to the copied text, for example.
- the linkage metadata is used to update or build the linkage structure that exists in the data processing module 102.
- the text editor module 103 When the author pastes the copied text into the text editor module 103, to effectively input the copied text, the text editor module 103 then communicates with the data processing module 102 to indicate that text has been pasted. The data processing module 102 then identifies that the text inputted in the text editor module 103 is the copied text, and is thus able to attribute the inputted text to the source information and associated linkage metadata received from the database application 106 when the original source information was copied.
- the second method of inputting text therefore involves recording the copied text and associated linkage metadata from the database 105 once it is copied, such that if the text is pasted into the text editing module 103, the data processing module 102 already contains the recorded source information and linkage metadata necessary for linking the source information to the text inputted by the author.
- the author action of copying the source information may be sent to the data processing module 102 from the database application 106 without the source information itself.
- the source information such as html data, is only retrieved from the database 105 once the copied text is pasted into the text editing module 103.
- This alternative method may therefore be considered to be more efficient as source information is only recorded in the data processing module 102 once input text is pasted.
- the text editing module 103 then communicates with the data processing module 102 and the data processing module 102 communicates with the database 105 via the database application 106 to obtain the specific source information.
- the uploaded document itself may be recorded as an online source by the database application 106 and sent to the data processing module 102.
- the third method of inputting text may however occur offline.
- the database 105 may be contained within the system 100.
- the author may have already saved the predefined document, or the database 105 itself may be proprietary as discussed previously.
- the database 105 when a predefined document is downloaded, the database 105, or the database application 106, communicate with the data processing module 102 such that linkage metadata of the downloaded document is captured and used to update the linkage structure. Furthermore, the document may be stored as a source at the data processing module 102.
- the author may also input text including numerical data into the system 100 as discussed previously, according to any of the first, second or third methods for inputting text.
- the text including the numeric data is considered as text for the purposes of tracking and recording metadata in the linkage structure of the document object or file format as discussed above.
- the system 100 may also be configured to provide input text to the text editing module 103 according to a fourth method, which focuses on the retrieval and manipulation of numerical data in a source.
- This fourth method for including input text comprises, firstly, identifying, by the data processing module 102, numerical data in a source within the database 105.
- the source may be a source identified in the initial search performed by the author, or otherwise identified by metadata in the linkage structure of the document object or file format.
- the system 100 may alternatively search the database 105 in the background for numerical data relating to the subject matter of input text already received by the text editing module 103. In order to achieve this, the system 100 may perform text and language recognition on the input text acquired from the author.
- the numerical data may be part of a graph or table contained in a source.
- the graph or table is a structured data set.
- Graphs or tables that have structured data sets comprise formatting, such that values within the graphs or tables are fixed in length, font, or relative position for example.
- the graphs and tables may have titles, subheadings or axis titles that can be identified by the system 100 to determine the subject of certain numerical values. For instance, if a source comprises a table with the column title‘quarterly revenue’, the system 100 identifies that the numerical contents of that column are associated with the quarterly revenue.
- the linkage structure may be updated with linkage metadata relating to the numerical data from the structured data set and the corresponding relationship to the source from where it was found.
- the system 100 may also identify unstructured numerical data, such as numerical values found within extracts of prose.
- the system 100 may identify the numerical value using a text-based recognition algorithm.
- the system 100 may then identify a relationship for the numerical value with a specific subject, by performing language recognition on the sentence from where the numerical value was identified. For instance, for the sentence: ‘The quarterly revenue for the third quarter was £2,345,800’ the system 100 may first identify the numerical value‘£2,345,800’ using text recognition, before using language recognition to determine that the numerical value is associated with the third quarter revenue.
- the linkage structure may then be updated with linkage metadata relating to the numerical data and the corresponding relationship to the source from where it was found.
- the fourth method for inputting numerical data into the text editing module 103 may then include aggregating and combining numerical values that are related by a certain subject. This may involve the system 100 identifying common subjects in the linkage structure according to the linkage metadata of the numerical data. For instance, numerical values that relate to‘quarterly revenue’ are combined by the system 100. The system 100 may then perform an analysis and processing on the combination of numerical values, such as applying mathematical formulae, or segregating related values dependent on sub-divisions of the subject such as the year or date, for example, in order to produce a meaningful result. The result may then be used to produce a graphic such as table or graph. The graphic is then provided to the text editing module as a graph or table to be published.
- the graph or table to be published contains numerical data that may be accumulated from several sources, but that is interrelated according to the specific subject.
- the table may for instance, show the quarterly revenue for a company searched or written about by the author elsewhere in the writing process.
- the system 100 may perform a search of the linkage structure and/or the database 105 for numerical data according to sources author by the user or input text inputted by the author.
- the manipulation and analysis may occur in the background and be presented to the author in the text editing module 103 as a finished graph or table to be published.
- the author may then be able to accept, reject or edit the table or graph for publishing.
- the linkage structure of the document object or file format is updated with metadata according to the table or graph for publishing and the author’s interactions with it, in terms of accept, reject or edits.
- the document object or file format preserves the source information and manipulation of the numerical data, such that when a document published in the system 100 contains graphs or tables, the document object or file format provides the reader with information regarding the source, references, and any mathematical manipulation of the original numerical data.
- the method of inputting numerical data according to the fourth method is different from the first, second and third method for inputting text in that the system 100 automatically performs the input in the fourth method, by using text and language recognition, followed by data analysis and processing.
- the publisher module 104 is configured to publish the text document that contains the inputted text.
- the publisher module 104 publishes the text document in the user interface module 101.
- the published text document includes the at least some of the linkage metadata of the linkage structure used to map the use of sources from the database 105 to the inputted text.
- a user of the user interface module 101 may access the published text document through interaction with the user interface module 101. It is to be understood that the user of the interface module 101 who accesses the published text document may or may not be the author of the published text document. Therefore, the user who accesses the published text document is referred to as a‘reader’.
- the published text document contains the text input of the author as well as at least some of the linkage metadata that describes how a source or sources were used to inform the input text, the reader is provided with an information-rich publication, from which sources of further information can be easily identified.
- Figure 2 shows an example of the linkage structure 110.
- the linkage structure 110 maps how a source 1 11 and an author input 112 into the text editing module 103 are used by the system 100 to arrive at the published document 1 13.
- the linkage structure 110 also includes linkage metadata 114 that provides information on author actions, such as a click, scroll, website visit, copy or the like, and how these actions link between sources 1 11 and author inputs 1 12.
- the linkage structure 1 10 forms at least part of the document object or file format of the present invention.
- the linkage structure 1 10 is arranged in a Directed Acyclic Graph format (DAG), wherein the nodes of the graph can only be traversed in one direction.
- DAG Directed Acyclic Graph format
- the linkage structure 1 10 contains multiple sources 11 1 and author inputs 1 12, and maps the entire creation of the published document. However, it is to be understood that the linkage structure 1 10 may overwrite itself if a second author input overwrites input text of a first author input. An overwrite of the linkage structure may be performed to save space on the data processing module 102.
- the data processing module 102 tracks the actions of the author as discussed previously, the actions including any interactions with the database such as clicks, website visits and downloads for example.
- the author actions 1 14 are recorded within the linkage structure 110.
- the author actions 114 stored in the linkage structure 1 10 may be directly related to sources 11 1.
- the author actions 114 may be a visit of a particular website, and the source 1 11 may be the text or html content of the particular website.
- the source 11 1 in this instance may be incorporated into the linkage structure 110 at the time of the search done by the author, or it may be included in the linkage structure 1 10 at a later time, once the author has inputted text into the text editing module 103 that is determined to be informed by a particular source as discussed previously. It is to be understood that incorporating the source 1 11 into the linkage structure 1 10 may comprise including the whole source, including extracted text and/or html data, or alternatively, it may comprise including a pointer or address to the source, so that the source can be accessed only when necessary and without putting undue burden on memory allocated for the linkage structure 110, for example.
- the linkage structure 1 10 contains metadata corresponding to the source 11 1 , the author actions 114 and the author input 112.
- the author input 1 12 is compared to the source 11 1 to determine in what way the author has used the source 11 1 to inform the text input into the text editing module 103.
- the method of text input from one or more of the first, second and third method of inputting text is determined and the associated author action, such as typing, copy and pasting, or uploading is recorded in the linkage structure 110.
- the first method of inputting text there may be the further step of directly comparing inputting text with the source text and scoring a similarity to determine whether the source text was used to inform the input text, as discussed previously.
- the linkage structure 1 10 thus contains information regarding what sources are used for input text, and where and how they are used in the input text. As can be seen in Figure 3, two or more sources may be used for one input text, such that words, sentences or paragraphs from different sources can be inputted into the same input text and still be tracked and linked back to their respective sources 11 1 by the linkage structure 110.
- the linkage structure 1 10 may keep track of inputs from separate authors, who collaborate on writing into the word editing module 103.
- the first author input ‘AUTHOR INPUT#T 112 and the second author input ‘AUTHOR INPUT#2’ 1 12 in Figure 3 may therefore represent the inputs of different authors.
- the example of Figure 3 shows that the second author input uses the first author input, but also adds from the source 11 1 used by the first author as well as from a new second source 1 11.
- the actions that the second author performs on the first source, the second source and the first author input are recorded as author action metadata 114.
- the linkage structure 1 10 is therefore capable of keeping track of revisions and edits to the inputted text by separate authors, and how each author input interacts with source information and other author inputs.
- the linkage structure overwrites linkage metadata such as source information 1 11 , author actions 114 and author inputs 112 if a previous author input 1 12 is overwritten by a later author input 112.
- the text editing module 103 communicates with the data processing module 102 to indicate that text inputting has ended.
- the linkage structure 110 updates accordingly to reflect that the inputted text is used in the published document 1 13.
- Figure 3 therefore shows an example of a complete linkage structure 1 10, from the sources 1 11 used to the finished document 113.
- Figure 3 shows a data structure for capturing metadata according to an embodiment of the invention.
- a tabular structure is used for the purposes of illustration and to show how the different aspects of the document object data may be associated with each other.
- the document object has three columns for storing respectively a data linkage item number 301 , a data linkage content item 302 and a metadata linkage object 303.
- the data linkage item number 301 provides a unique ID for each row in the table so that it can be referenced unambiguously.
- the data linkage content item 302 contains an element of text or a graphics element input by the user, using any of the techniques noted above, and metadata linkage object 303 contains metadata detailing when and how the content item 302 was created.
- Figure 3 shows the case where the user has begun a new document or a new section of a document and entered text into the text editor.
- the user is creating a document about “Artificial Intelligence in the healthcare industry” and has begun by typing the words“The opportunities for Artificial Intelligence in health care” into the text editing module 103 via the user interface 101.
- The“If symbol indicates a carriage return.
- the data processing module 102 detects the text entry and populates the data object model 300 with the text data that the user has entered, the carriage return information, as well as metadata, such as the identity of the user entering the text, the time the text was entered by the user (12.56pm), and the nature of the input - in this case“manual input”.
- the user is represented by an indicator I D#27 by way of example. More or less metadata may be captured in other implementations of the preferred embodiment.
- the user browses though the database 105 to discover documents that relate to this topic.
- the database 105 may be the Internet, or a proprietary database in which reports or other documents are made available.
- the user searches for relevant sources, the user finds and view a number of documents with respective individual links ⁇ link #1 >, ⁇ link #2>, and ⁇ link #3>.
- the user cuts and pastes the following paragraph of text from the document available at ⁇ link #1 > into the text editing module 103:
- the data processing module 102 detects that the user has both browsed to the document in the web browser or a proprietary search window provided in the database module 105, and copied and pasted text from the document into the text editing module 103.
- the text copy and pasted by the user appears in the text editing window underneath the text that the user typed and which is stored in record #1.
- the data processing module captures the data as new record #2.
- Metadata is stored with the input text to indicate that the cutting and pasting operation was carried out at 1.02pm, by user with ID#27.
- the entry type is therefore “copy from” and the link from which the text was sourced is also added to the metadata, ⁇ I i n k# 1 > .
- the user also copies and pastes the following paragraph of text from the document available at ⁇ link #2> into the text editing module 103:
- Machine Learning algorithms such as those provided in the field of Artificial Intelligence, can also be trained using real medical data, benefitting from the collective experience of the medical professionals involved in the training. Such algorithms can typically extrapolate that training better than humans, reducing the element of luck in the diagnosis. Machine learning algorithms provided by Medical Al Ltd have shown to be 20% more reliable than doctors in actual field trials.
- the data processing module 102 detects that the user has both browsed to the document in the web browser or a proprietary search window provided in the database module 105, and copied and pasted text from the document into the text editing module 103.
- the text copy and pasted by the user appears in the text editing window underneath the text that the user typed and which is stored in record #2.
- the data processing module captures the data as new record #3.
- Metadata is stored with the input text to indicate that the cutting and pasting operation was carried out at 1.05pm, by user with ID#27.
- the entry type is therefore“copy from” and the link from which the text was sourced is also added to the metadata, ⁇ link#2>.
- the user browses to and reads a document at ⁇ link#3>. However, the user does not copy and paste any text from this document into the text editing module 103.
- the data processing module 102 captures this fact by creating a record #4 that stores the actions of the user, ⁇ link#3>,“1.06pm browsed”. If the user subsequently cuts and pastes text from the third document, available at ⁇ link#3> then record #4 may be updated in the manner illustrated in records #2, and #3 with a new time and a“copy from” indication above the link.
- record #4 remains in the metadata as a record of the user’s actions browsing and inspecting the third document, that is available for later auditing.
- the position of the record #4 is not indicative of the layout or special arrangement of the text information in the document model, and the data processing module 102 attempts to keep the record #4 positioned in the document model as close to the correct time as possible according to its time stamp. This will be illustrated in Figure 7.
- the fact that the user does not copy and paste text from the third document can be detected if the user browses away from the third document to another document, or begins another action, such as manually typing text into the text editor.
- embodiments of the invention may detect whether the text being typed by the user matches the text present in the database 105 document window or web browser window, so as to capture the fact that the user is copying the text, but has not used the clipboard function.
- the data processing module captures the actions of the user and updates the document model accordingly. Sections of the original copied text that are unedited by the user are preserved as new separate records #2 and #4, and a new record #3 is created reflecting the edits made to the original text. The records with a higher number are renumbered accordingly.
- the metadata for each of the data linkage content items #2 and #4 is updated to “unedited”, while the metadata for record #3 is set as“edited”, and data showing the edits is preserved.
- the document model records both the fact that the user’s document contains text copied from another document, and the fact that this text has been changed. Using the record structure of the document model, it is possible to determine what aspects of the copied text have been edited and what aspects have not been edited.
- record #3 is created to reflect the editing of a portion of text beginning with the user’s first edit and ending with the last. It is possible to make the creation of records and the indication of what sections of the text are“edited” or“unedited” more or less fine grained. For example, records marked “edited” might be limited to show only adjacent deletions and insertions of text, such that in Figure 5, the section of text in record #3 that reads “staff typically diagnose a patient’s underlying health issues, based on a visual inspection of a patient medicaf’ would be included in a new record marked“unedited”. However, to avoid proliferation of records (and to make the description here easier to follow), it will be understood that record #3 strictly shows both edited and unedited text, and that record #3 shows a section of edited text having both edited and unedited sections.
- the data processing module 102 may make the decision of whether to create a more fine grained record, or a record showing a section of edited text, based on the numbers of words in the record to be marked as a section of “edited” text. For example, in Figure 5, if record #3 contained more than 250 characters for example, data processing module 102 might seek to create more than one record to capture the associated editing information.
- Figure 6 shows a further editing action performed by the user and captured by the data processing module 102.
- the user deletes the final sentence of text“Some conditions may be more difficult to detect than others", and the data processing system captures this deletion as a further edit of the text from link ⁇ link#1 > creating new record #5, and renumbering former records #5 and #6 to be #6 and #7.
- a further method of data or text entry might be to upload data such as an image of text data, a chart, or some other graphical image.
- Most computer systems offer functions for capturing as a graphic part of the screen, or saving a link which may then be pasted or uploaded into a document and reproduced.
- the user finds at 1.15pm a chart which illustrates the applications of image based processing techniques to different types of patient data, and has added this chart to the document being created.
- the data processing module 102 captures the link information (a shortcut or pathway to the file) or the graphical information (the image file itself) of the chart and includes this in the document object model as the data linkage content item. As before, metadata is added to the document model to indicate the user ID, the link where the chart was found, ⁇ link #4>, and the time of addition.
- the system tracks the user activities and captures in the document object model, the necessary data describing the content and provenance of the data.
- the data may refer to text data or numerical data, or both.
- the linkage structure captured in the document object model or at least some of the linkage metadata may be made available to the reader. This feature will now be discussed with reference to Figure 8.
- Figure 8 shows an example of a published document 120 according to the present invention.
- the published document 120 includes a title 121 and a body of text.
- a portion 122 such as a paragraph, may be selected by a reader by a click, hover or the like 125. Selecting the portion of text 122 opens a textual representation 123 of the linkage structure 110.
- the textual representation of the linkage structure 123 includes linkage metadata 124 related to the selected portion of text 122.
- the published document 120 is therefore information-rich, in that the reader of the published document can interact with the document 120 to obtain more information 123 regarding the document 120.
- the metadata 124 included in the textual representation 123 of the linkage structure 110 is determined based on the selected portion 122 of the published text document 120. As shown in Figure 8, the metadata available may include a summary of the selected portion of text 122, what parts of the text are inputted by the author, what parts are excerpts or quotes direct from sources, and information regarding the sources. Furthermore, as can be seen from Figure 8, the textual representation 123 may also include revision or tracked history data, showing version 1 v1 , version 2 v2, and version 3 v3 of the selected portion of text 122.
- the linkage metadata information for each portion of text 122 is retrieved from the linkage structure 110 from the data processing module 102 once the text inputting is complete.
- the publishing module 104 then publishes the inputted text with links to the relevant linkage metadata as shown in Figure 8.
- the textual representation 123 may include an address or link to an entire source document, or may alternatively highlight and include relevant parts of the source document.
- An advantage of the present invention is that the reader is able to quickly and efficiently have access to much more information regarding the topic of the published document 120 than is actually in the document itself, by simply having access to the system 100. Since the document is written and published in the system 100, and the system 100 maintains all linkage metadata information and the linkage structure 1 10, when the reader enters the system 100 they may have access to all recorded and saved information.
- the system 100 may include several published documents 120, each with their own respective linkage structures 1 10.
- the system 100 becomes a database in itself, in which authors can contribute to new or existing published documents, and readers can read the documents 120 as well as see exactly what authors have done and what sources they have used according to the linkage metadata.
- Each linkage textual representation 123 of a linkage structure 1 10 may further link to sources identified in similar linkage structures from different published documents 120.
- historic data relating to the different published documents 120 may be used and provided by the different respective linkage structures 110.
- the system 100 may first search for similar published documents 120 and their corresponding linkage structures 1 10 for historic data that is relevant to the topic of search, and suggest them to the author.
- the search for similar published documents may be done using language recognition algorithms for example.
- the linkage structure 1 10 of each published document 120 is maintained such that the reader and/or author can audit the writing process and determine the reliability of publications. Since the linkage structure 1 10 is maintained, this promotes security within the system 100 and clarity of information to the reader.
- the system 100 may have a feedback functionality.
- the feedback functionality is provided using the data collected by the database application 106 and stored in the data processing module 102.
- the database application 106 captures sources and related information as discussed previously, which is then recorded on the data processing module 102.
- the database application 106 may also track the navigation behaviour of the author on the database, such that website visits, search parameters, and number of clicks, for example, are recorded by the database application 106. This information may then be analysed to determine search practices of the author and how sources were used during the report writing process. This includes identifying what search terms resulted in finding a source in the search which was then used by the author, which sources were visited and used, and which sources were visited but not used by the author.
- This analysis of the search and the usefulness of sources may be used to inform the author or an administrator of the system 100, via a metric such as a score, search analysis document or graphic, of the productivity of the author and/or the quality of the search and output of the search.
- a metric such as a score, search analysis document or graphic
- the author or administrator of the system can determine author related statistics concerning the report writing process, such as how productive and efficient the author is in searching and identifying relevant source information, as well as where to find relevant source information.
- This feedback allows the administrator to see which authors create specific reports, and what behaviour in terms of the search led to good or bad quality output or to fast or slow delivery of the report.
- the system 100 automatically produces these author-related statistics, and report on the results.
- the report of the results may take the form of a score, search report document or graphic, such as authoring time vs search behaviour, or report feedback scores vs search behaviour.
- the author-related statistics may then be published to identify how an author is performing, and which areas or addresses of the database, and therefore which sources, are most used and most useful.
- the sources used in previous report-writing may be ranked from most used to least used, or according to most commonly found during a search. This ranking may be published within the system 100 for the administrator and one or more author to view.
- the database application 106 may prompt the author with author-related statistics regarding the source. For example, the database application may display information to the current author relating to the interactions of the previous author or authors with the source. This may include stating that the previous author or authors used the source or did not use the source, and statistics and/or rankings regarding how popular the source is with authors, or common search terms inputted by authors who used the source.
- This feedback functionality may also apply to specific paragraphs within sources, such that paragraphs within sources are ranked according to their selection or non-selection by previous authors. Manually selected paragraphs may therefore also be included in the author- related statistics.
- the aim of the feedback functionality is to drive ongoing improvements to the report writing process, so that the system administrator and/or future authors are informed with historic data concerning previous report-writing processes.
- the published document 120 is accessed in the system 100, it may be exported and read by the reader outside the system 100. However, in order to maintain security and to keep the linkage structure updated, it is preferable that the author or authors of a document perform the process of writing within the system 100.
- system 100 is referred to as a collection of modules and a database 105. It is however to be understood that the system 100 may be comprised by any computer terminal or a set of computer readable instructions recorded on a computer-readable medium. The system 100 may therefore exist over a network, server or the like and as such is not required to be accessed locally. The author and/or reader may access the system 100 over a connection to the internet or any other remote connection to the system 100.
- the method according to the present invention relates directly to the workings of the invention as discussed above with reference to Figures 1 to 4.
- the method according to the present invention may be performed on a set of computer readable instructions recorded on a computer readable medium, such as a computer program written on a computer, sever, or distributed computer network.
- an author of the document to be written may perform a first search according to the topic of the document to be written.
- the author searches for sources comprised within a database.
- the author may be prompted with sources or previously published documents relating to the topic of the document to be written, from searching historical data.
- the author may be provided with a predetermined template for inputting text according to the topic of the document to be written.
- the topic of the document to be written can be determined using language recognition algorithms or other text-based analysis on the search string of the first search.
- the activity of the author is tracked and captured and used to build a linkage structure including linkage metadata.
- the activity of the author that is tracked and captured includes any interaction from the author during the first search, and with sources of information identified in the search, or sources otherwise found by the author. These interactions may include clicks, website visits, downloads and copy and pastes for example. Throttling may be performed on the tracked actions of the author, such that only significant actions that are useful in determining sources are captured and tracked.
- sources may be identified according to the author actions and added to the linkage structure. Such sources may include websites, spreadsheets, or other such documents.
- the linkage structure connects the sources to the related author actions.
- the source connected to that user action may be the html content of the website.
- author input is recorded at a user interface, the author input being an input of text.
- the author may input text to the user interface according to any one or more of three methods.
- the first method of inputting text comprises directly inputting text into the user interface.
- the author may type, dictate, or otherwise input text in this way.
- the second method of inputting text comprises pasting text.
- the pasted text is copied from a piece of text originally found in a source.
- the third method of inputting text comprises uploading all or a part of a predefined text document.
- the linkage structure is updated according to the author input, and if the source used is not already recorded, it is then fetched and recorded in the linkage structure.
- the linkage structure is searched for sources that compare closely with the inputted text to determine if the author used these sources in informing the inputted text.
- the linkage structure is updated to reflect that the author has copied and pasted from a particular source.
- Linkage metadata from the source such as a URL, html or text data is included in the linkage structure.
- Step 203 repeats every time that new text is input into the user interface.
- a database application may be used to communicate between the database, the user interface and a data processor.
- the data processor processes and stores the linkage structure and any identified sources.
- the inputted text is published into a published document.
- the linkage structure is conserved and published with the published document as extra information to the inputted text, wherein the extra information corresponds to the recorded linkage metadata of the linkage structure.
- a reader of the published document may access the extra information by interacting with the published document.
- the extra information may include a summary of a selected portion of text, what parts of the text are inputted by the author, what parts are excerpts or quotes direct from sources, and information regarding the sources.
- the extra information may also include revision or tracked history data and suggestions for similar published documents and/or sources.
- the reader is therefore provided with an information-rich publication with links to linkage metadata that is relevant to the topic of the published document.
- the reader is also provided with information regarding the author or authors’ actions during the process of writing in step 203, and as such, can clearly audit the writing process of the author or authors.
- the method according to the present invention provides users with a more informative and transparent method of writing and publishing documents, with conserved linkage between author actions and information sources, which can be accessed again once the document has been published.
- the method includes presenting via a first user interface controls for a user to interact with a document database.
- the method includes presenting, via a second user interface, controls for a user to enter and record text based input into a document object or file format.
- the method includes capturing as metadata, the user’s interactions with the document database and adding it to the document object or file format.
- the method includes capturing input text entered by the user and adding it to the document object or file format.
- the method includes determining an association between the text entered by the user and the captured metadata, and updating the metadata to reflect the association.
- the method includes capturing an editing action from the user modifying text already entered by the user.
- the method includes adding metadata to the document object or file format to identify the modifications made to the text by the editing action.
- the method includes saving in memory the document object or file format, including the text entered by the user, the modifications made to the text by the editing action, and the metadata capturing the user interactions and the determined association.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Human Computer Interaction (AREA)
- Business, Economics & Management (AREA)
- Document Processing Apparatus (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
L'invention concerne un procédé et un système de capture de métadonnées dans un objet de document ou un format de fichier. Le procédé comprend les étapes consistant à fournir une interface utilisateur à un utilisateur ; recevoir une entrée d'utilisateur telle qu'une entrée de texte (201) au niveau de l'interface utilisateur ; suivre et capturer simultanément l'activité de l'utilisateur en tant que métadonnées dans un objet de document ou un format de fichier (202) lorsque l'utilisateur navigue dans une base de données de documents ; déterminer une association entre l'entrée de texte et les métadonnées dans l'objet de document ou le format de fichier ; mettre à jour les métadonnées dans l'objet de document ou le format de fichier pour refléter l'association ; et publier l'objet de document ou le format de fichier avec l'entrée de texte (204) lorsque l'utilisateur a fini d'éditer ou d'utiliser l'interface utilisateur. Le procédé préserve ainsi des informations concernant le processus d'édition d'une entrée de texte et publie ces informations avec l'entrée de texte.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
GB1901830.8 | 2019-02-08 | ||
GB1901830.8A GB2582536A (en) | 2019-02-08 | 2019-02-08 | Method and system for capturing metadata in a document object or file format |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2020161506A1 true WO2020161506A1 (fr) | 2020-08-13 |
Family
ID=65998593
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/GB2020/050297 WO2020161506A1 (fr) | 2019-02-08 | 2020-02-10 | Procédé et système de capture de métadonnées dans un objet de document ou un format de fichier |
Country Status (2)
Country | Link |
---|---|
GB (1) | GB2582536A (fr) |
WO (1) | WO2020161506A1 (fr) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2022151903A1 (fr) * | 2021-01-18 | 2022-07-21 | 华为技术有限公司 | Procédé d'accès à un fichier distribué et dispositif associé |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPWO2021039129A1 (fr) * | 2019-08-29 | 2021-03-04 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5873107A (en) * | 1996-03-29 | 1999-02-16 | Apple Computer, Inc. | System for automatically retrieving information relevant to text being authored |
EP2584509A1 (fr) * | 2011-10-17 | 2013-04-24 | Research In Motion Limited | Interface de compilateur de notes |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7340685B2 (en) * | 2004-01-12 | 2008-03-04 | International Business Machines Corporation | Automatic reference note generator |
US20060218034A1 (en) * | 2005-03-23 | 2006-09-28 | Kelly Laird R | System and method for monitoring and recording research activity |
US9361375B2 (en) * | 2008-07-29 | 2016-06-07 | Excalibur Ip, Llc | Building a research document based on implicit/explicit actions |
US20100030765A1 (en) * | 2008-07-30 | 2010-02-04 | Yahoo! Inc. | Automatic generation of attribution information for research documents |
-
2019
- 2019-02-08 GB GB1901830.8A patent/GB2582536A/en not_active Withdrawn
-
2020
- 2020-02-10 WO PCT/GB2020/050297 patent/WO2020161506A1/fr active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5873107A (en) * | 1996-03-29 | 1999-02-16 | Apple Computer, Inc. | System for automatically retrieving information relevant to text being authored |
EP2584509A1 (fr) * | 2011-10-17 | 2013-04-24 | Research In Motion Limited | Interface de compilateur de notes |
Non-Patent Citations (3)
Title |
---|
LASSE LEHMANN ET AL: "Capture of lifecycle information in office applications", INTERNATIONAL JOURNAL OF TECHNOLOGY ENHANCED LEARNING, vol. 2, no. 1/2, 2010, pages 41, XP055688812, ISSN: 1753-5255, DOI: 10.1504/IJTEL.2010.031259 * |
MARTIN WOLPERS ET AL: "Tracking Actual Usage: the Attention Metadata Approach", JOURNAL OF EDUCATIONAL TECHNOLOGY & SOCIETY, July 2007 (2007-07-01), Palmerston North, pages 106 - 121, XP055688472, Retrieved from the Internet <URL:https://www.fit.fraunhofer.de/content/dam/fit/de/documents/01_Tracking-Actual-Usage-the-Attention-Metadata-Approach.pdf> [retrieved on 20200424] * |
SRIRAM RAGHAVAN ET AL: "Eliciting file relationships using metadata based associations for digital forensics", CSI TRANSACTIONS ON ICT, vol. 2, no. 1, March 2014 (2014-03-01), pages 49 - 64, XP055688629, ISSN: 2277-9078, DOI: 10.1007/s40012-014-0046-4 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2022151903A1 (fr) * | 2021-01-18 | 2022-07-21 | 华为技术有限公司 | Procédé d'accès à un fichier distribué et dispositif associé |
Also Published As
Publication number | Publication date |
---|---|
GB201901830D0 (en) | 2019-04-03 |
GB2582536A (en) | 2020-09-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7140536B2 (en) | Method and system for highlighting modified content in a shared document | |
US7899829B1 (en) | Intelligent bookmarks and information management system based on same | |
Schönfelder | CAQDAS and qualitative syllogism logic—NVivo 8 and MAXQDA 10 compared | |
CN103098051B (zh) | 搜索引擎优化助理 | |
US8533199B2 (en) | Intelligent bookmarks and information management system based on the same | |
WO2009154153A1 (fr) | Système de recherche de documents | |
US20090322756A1 (en) | Using visual techniques to manipulate data | |
US9495424B1 (en) | Recognition of characters and their significance within written works | |
US20130036348A1 (en) | Systems and Methods for Identifying a Standard Document Component in a Community and Generating a Document Containing the Standard Document Component | |
US20050081146A1 (en) | Relation chart-creating program, relation chart-creating method, and relation chart-creating apparatus | |
JP7333832B2 (ja) | リビング文書の生成および対話型編集のためのシステムおよび方法 | |
KR102055407B1 (ko) | 정책정보 제공 방법, 정책정보 제공 시스템, 이를 위한 컴퓨터 프로그램 | |
WO2020161506A1 (fr) | Procédé et système de capture de métadonnées dans un objet de document ou un format de fichier | |
US20080208803A1 (en) | System and method for characterising a web page | |
US20100325528A1 (en) | Automated formatting based on a style guide | |
KR100522186B1 (ko) | 동적으로 홈페이지를 제작하는 방법 및 이 방법을 웹에서구현하는 장치 | |
Edhlund et al. | NVivo for Mac essentials | |
JP4288247B2 (ja) | 履歴作成装置、活動履歴作成方法、及び活動履歴作成プログラム | |
US20100211562A1 (en) | Multi-part record searches | |
CN112418875A (zh) | 跨平台税务智能客服语料迁移方法及装置 | |
US20120150899A1 (en) | System and method for selectively generating tabular data from semi-structured content | |
KR101835994B1 (ko) | 키워드 맵을 이용한 전자책 검색 서비스 제공 방법 및 장치 | |
US20220138407A1 (en) | Document Writing Assistant with Contextual Search Using Knowledge Graphs | |
JP2009223679A (ja) | 電子文書検索装置、及び電子文書検索プログラム | |
Edhlund et al. | NVivo 12 for Mac Essentials |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20705787 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08.11.2021) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20705787 Country of ref document: EP Kind code of ref document: A1 |