US 20070083552 A1
A natural language-based information organization and collaboration tool for a computer system is disclosed. The present invention includes an apparatus and method for processing text expressions in a computer system, the apparatus including: 1) an object database defining an information object with an associated keyword; 2) a user input device for receiving an input text expression; 3) a parsing device for identifying the keyword in the input text expression, the parsing device including functions for linking the input text expression to the information object based on the keyword identified in the input text expression; and 4) a user output device for displaying to the user the identity of the information object to which the input text expression was linked. The apparatus of the present invention further includes supplemental information in the object database which is related to the information object, the user output device further including functions for displaying the supplemental information when a corresponding keyword is identified in the input text expression. The apparatus of the present invention further includes a method and apparatus for collaboration between users of a time and project management system.
1. An apparatus for processing text expressions in a computer system, the apparatus comprising:
an object database defining an information object with an associated keyword;
a user input device for receiving an input text expression;
a parsing device for identifying the keyword in the input text expression, the parsing device including functions for linking the input text expression to the information object based on the keyword identified in the input text expression; and
a user output device for displaying to the user the identity of the information object to which the input text expression was linked.
2. The apparatus as claimed in
The present invention relates to the organization and access to information stored in a computer system. More specifically, the present invention relates to the analysis of natural language input to produce structured information output and the processing of notes in a computer system. The present invention also relates to time and action/project management using a computer system. More specifically, the present invention relates to a method and apparatus for collaboration between two or more persons for time and project management.
Many application programs exist in the prior art for organizing information in particular ways or for manipulating specific types of information. For example, word processing applications are specifically designed for manipulating text documents in a computer system. Similarly databases in the prior art provide means for structuring data in well defined ways. Further, calendaring systems provide a structured way for tracking events or actions required at specified dates and times. Although these prior art applications provide ways for organizing information in particular ways, it is often inconvenient to require a user to switch back and forth between application programs when updates of information are required. Moreover, these types of systems require a user to organize the information prior to entering data into the computer system. For example, the user must know to activate a calendaring program if an appointment or action date is to be entered. In separate actions, the user may also need to update lists or databases associated with the appointment or action for which a calendar entry was made. In many prior art systems, the user is required to spend time navigating around a user interface to link information to the desired lists or categories to which it pertains.
U.S. Pat. No. 5,115,504 entitled “Information Management System” describes a system for linking elements representing stored information in a database. The system comprises a link structure formed in a section of the database independent of the elements, a pointer in the link structure indicating the location of a first element, and a second pointer in the link structure indicating the location of a second element. The database contains items comprising textual data and a plurality of categories into which the items may be categorized such that each item may be linked to more than one category. The system automatically assigns an element in a database to a parent category if it has been assigned to a child category of the parent. The system also generally features a means for assigning an element in a database to one or more of a plurality of categories, the categories being hierarchically arranged. The system constructs views as screens of information organized into sections having categories and section heads and items presented one after another beneath a given section head of a category to which the item has been assigned. After entering an item, the user can make further assignments directly by moving to the columns of the view and entering an existing name of a sub-category under the column head. In this manner, a link structure is created.
Unfortunately, the system disclosed in U.S. Pat. No. 5,115,504 still requires a user to directly manipulate information categories on a display screen. Using this approach, a user is still required to organize the information in some fashion on entry of the data into the desired category. In many situations, it is inefficient and inconvenient for a user to pre-organize and explicitly store information in this fashion. Moreover, further efficiencies could be obtained if a user could provide input in a convenient free form or natural language representation. It would also improve prior art systems if a user could update an information item or action item easily, quickly and without losing the context in which he/she is currently engaged.
Another disadvantage of the prior art systems described above, is that they do not provide a mechanism for collaboration between users or between applications. Organizing one's own information is important, but actions/projects are often shared between two or more people. In other words, it is not sufficient to organize one's own To Do lists and calendars. The user should also be able to collaborate with other users to assign projects, accept project assignments, and inform others. Furthermore, an ability to notify others of the status of projects is advantageous.
One prior art method of allowing multiple users to work together includes using e-mail to send messages to others. Although this prior art application allows users to communicate regarding any topic, it is not linked to calendars, lists, or external databases maintained by the users. Thus, as an agreement evolves with each e-mail exchange, the user is often involved in updating everything associated with the project. It would further improve prior art systems if a user could update an action item easily, quickly and without losing the context in which he/she is currently engaged. Further, it would be an improvement over the prior art to provide a system whereby multiple users could collaborate and track action items between many participants and across many computer systems.
Another prior art method utilizes top down delegation. This allows a supervisor to delegate projects to subordinates. However, it does not allow negotiation or collaboration between users. Rather, it is rigidly hierarchical. Thus, this method is not useful for collaboration and negotiation between peers. Another disadvantage of such prior art systems is that they use a client-server system. That is, a server has to be provided as a repository of information regarding the collaboration process. Thus, only clients that are connected to the server can participate in the collaboration. This means that persons who are not linked to the same server can not participate in the collaborative environment.
It would be an improvement over the prior art to provide a system that allows collaboration between two or more users. Further, it would be an improvement over the prior art to provide a system whereby multiple users could collaborate and track action items between many participants and across many computer systems.
Thus, a better natural language information organization and collaboration tool is needed.
The present invention is a natural language based information organization and collaboration tool for a computer system. The present invention includes an apparatus and method for processing text expressions in a computer system, the apparatus including: 1) relational object database defining an information object with an associated keyword, project, list, contact, date/time event or enclosure; 2) a user input device for receiving an input text expression; 3) a parsing device for identifying the keyword in the input text expression, the parsing device including functions for linking the input text expression to the information object based on the keyword identified in the input text expression; and 4) a user output device for displaying to the user the identity of the information object to which the input text expression was linked. The apparatus of the present invention further includes supplemental information in the object database which is related to the information object, and the user output device further includes functions for displaying the supplemental information when a corresponding keyword is identified in the input text expression. The apparatus of the present invention further includes a method and apparatus for collaboration between users of a time and project management system.
The present invention is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
The present invention is a natural language based information organization and collaboration tool for a computer system. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one of ordinary skill in the art that these specific details need not be used to practice the present invention. In other circumstances, well-known structures, circuits, and interfaces have not been shown in detail in order to not obscure unnecessarily the present invention.
The control logic or software implementing the present invention can be stored in main memory 104, mass storage device 107, or other storage medium locally accessible to processor 102. Other storage media may include floppy disks, memory cards, flash memory, or CD-ROM drives. It will be apparent to those of ordinary skill in the art that the methods and processes described herein can be implemented as software stored in main memory 104 or read only memory 106 and executed by processor 102. This control logic or software may also be resident on an article of manufacture comprising a computer readable medium 108 having computer readable program code embodied therein and being readable by the mass storage device 107 and for causing the processor 102 to operate in accordance with the methods and teachings herein.
The software of the present invention may also be embodied in a handheld or portable device containing a subset of the computer hardware components described above. For example, the handheld device may be configured to contain only the bus 101, the processor 102, and memory 104 and/or 106. The handheld device may also be configured to include a set of buttons or input signalling components with which a user may select from a set of available options. The handheld device may also be configured to include an output apparatus such as a liquid crystal display (LCD) or display element matrix for displaying information to a user of the handheld device. Conventional methods may be used to implement such a handheld device. The implementation of the present invention for such a device would be apparent to one of ordinary skill in the art given the disclosure of the present invention as provided herein.
The present invention is a natural language based, parsable, always available, intelligent note editor that captures user thoughts, action requests, and information in a computer system. The present invention uses natural language parsing to identify keywords and date information amongst a free form text input expression (denoted keynote herein) entered by a user and establishes links to other information objects based on the identified words. These linked other objects include projects, contacts, date/time events, lists, and document identifier objects. Keywords are pre-defined one word or multiple word text strings with or without punctuation that are associated or linked to one or more related information objects. Lists are user-established collections of related keynotes. Lists can be action-related or merely archived memos. The present invention further classifies the text input (i.e., keynote) as a particular type of keynote, such as an action, a memo, a personal keynote, a shared keynote, an action request, an FYI (for your information) message, or one of several other different types of keynotes. A personal keynote is one not intended to be sent to anyone else. A shared keynote is sent to others. Once the keynote is classified, the present invention takes action upon the keynote by sending an action request to a linked contact, updating a linked list, contact, date/time expression, or project data, sharing the keynote to others in a collaboration group, or storing information related to the keynote in an organized and efficient manner. The present invention includes a real-time and interactive user interface for receiving input text expressions from a user and for providing selectable supplemental information to the user regarding the classification of the keynote. In addition, the present invention includes a parser for processing natural language in the manner described above.
Referring now to
Referring now to
Referring again to
The second function served by the icons 240 and a related dropdown list control is a means for a user to select the display of the linked object type corresponding to a particular selected icon. The output produced by parser 300 is displayed in region 250 of shadow 230. Although shadow 230 may initially be displayed beneath keynote 220, the user may bring the shadow region 230 in front of the keynote region 220 by clicking on the shadow region 230 with the cursor control device 123 or by typing a pre-specified key entry on the keyboard 122. Alternatively, the shadow region 230 may also be selected for display using a menu command.
Referring now to
Referring now to
Referring now to
Drop down list indicators 260 are provided to cause a list to expand downward so a multiple line list of objects or information is displayed in a drop down portion of region 250. Conventional methods exist for providing drop down list indicators on a computer display device.
Thus, user interface 200 and its corresponding keynote and shadow regions 210 provide a means and method for receiving a natural language text expression from a user and for concisely and efficiently displaying the parsed and linked structured output of the text expression in an area on display device 121. In the following sections, the detailed description of the processing performed by parser 300 and lexical analysis tool 400 is provided.
As can be seen from
The parser 300 of the preferred embodiment receives natural language text expressions from user interface 200 and produces structured information including links to information objects, such as projects, contacts, lists, date/time calendar items, and enclosed documents corresponding to those identified to keywords in the input text expression. Although many parsing algorithms exist in the prior art, the parser 300 of the present invention is unique in its ability to effectively identify and suggest keywords and/or date/time calendar events in an input text string and respond with interactive user real-time performance. Parser 300 of the preferred embodiment accomplishes these objectives with a novel internal architecture and set of methods for processing a natural language text expression. The architecture and methods used by the parser 300 of the present invention will be described in the following sections.
The present invention solves the problem of interpreting structure and meaning from natural language text. This meaning is a set of structured information related to or linked to other pertinent information known to and pre-defined by the user. The following example illustrates the operation of the present invention.
Suppose a user enters the following sample keynote to the user interface 200 of the present invention:
“call Scott tomorrow to arrange the next Engineering meeting.”
The parser 300 of the present invention is used to analyze this keynote in real-time as the user enters the keynote character by character. Note that the entire keynote is parsed after the entry of each new character. After the entire keynote is entered by the user and analyzed by parser 300, the following structured information output is produced by parser 300:
In this example, parser 300 of the present invention recognized the keyword “call” in the input keynote and determined that this text input keyword should be linked to or related to the “Calls” list. The parser 300 of the present invention also recognized the keyword “Scott” and determined that this contact name should be linked to the contact object “Scott Jones”. The word “tomorrow” was also recognized by parser 300, which calculated tomorrow's date (i.e., current date+1 day) and linked this date object to the input keynote. The keywords “Engineering meeting” were recognized by parser 300 as a link to the previously defined “arrange Engineering meetings until Dennis gets back” project object. Note that the present invention displays these links between the input keynote and corresponding linked object types in the data areas for the list, project, contact, or date/time calendar event objects in display region 250.
One important goal of the present invention is to ease the computer user's workload by anticipating his/her intentions based on the natural language text expression that has-been entered. By anticipating the user's intentions with a reasonable degree of accuracy, the present invention allows the user to capture information in a much more efficient and comfortable manner. For example, after simply typing the input keynote in the example presented above: “call Scott tomorrow to arrange the next Engineering meeting”, the user is not required to spend valuable time navigating around the display screen or an application user interface to link the input keynote to the desired list, project, contact, and date/time calendar event objects. The present invention automatically handles the linkage of the unstructured information in the input keynote to corresponding structured information objects. The user is thus able to save the keynote, send the keynote, or initiate action upon the keynote very quickly with very little user intervention.
Referring now to
The date/time parser of keyword and date/time parser 810 is used to scan the input keynote for the presence of information corresponding to a date or time event. The operation of the date/time parser is described in more detail in a later section of this document.
Keyword and date/time parser 810 interfaces with a lexical analysis tool 400. It will be apparent to one of ordinary skill in art that the functions performed by lexical analysis tool 400 may equivalently be implemented as an integrated part of parser 300 or keyword and date/time parser 810. However, in the preferred embodiment of the present invention, lexical analysis tool 400 is provided as a software module independent from keyword and date/time parser 810. Upon receiving an input natural language keynote from user interface 200, keyword and date/time parser 810 passes this keynote to lexical analysis tool 400 on line 410. Lexical analysis tool 400 is responsible for producing a set of tokens from the input keynote.
Referring now to
The new token, the token type, and token value information is then added to an output token list, which represents the token list that will be returned to parser 300 at the completion of processing performed by lexical analysis tool 400. Also in processing block 926, the token buffer is initialized to set up for collection of the next token and processing is transferred to processing block 930 where the next character in the keynote is processed. Using this basic processing flow performed by lexical analysis tool as shown in
The interface between parser 300 and lexical analysis tool 400 may also be used to exchange tokens between parser 300 and lexical analysis tool 400 for the purpose of obtaining a multiple word token that may be used to represent a particular date/time event. The parser 300 and lexical analysis tool 400 exchange date/time tokens until the parser 300 finds a date/time processing path that succeeds for the date/time token.
Referring now to
Object dictionary 851 includes a trigger table 856, a keyword dictionary 852, and keyword definitions table 854. Trigger table 856 includes entries called triggers for each of the tokens from which keywords are formulated. Associated with each trigger is a reference count identifying the number of keywords of which the corresponding trigger is a member. Keyword dictionary 852 includes the identity of each of the keywords pre-defined using methods or calls provided by the keyword parser of keyword and date/time parser 810. Blocks 1210, 1212, and 1214 shown in
Referring now to
Referring again to decision block 1014, if the retrieved token is not a trigger as determined by access to trigger table 856, processing path 1016 is taken to decision block 1020 where the contents of the current keyword buffer are checked. If the current keyword buffer is empty, processing continues at decision block 1028 through the bubble labeled A. In this case, the current token is simply thrown away. If the current keyword buffer is not empty however, processing continues at the bubble labeled B illustrated in
Referring now to
Referring now to
Referring now to
In the second example shown in
Referring now to the third example illustrated in
Referring to the fourth example illustrated in
The present invention includes methods and objects for suggesting new keywords to a user given a predefined object dictionary 851 and an input user keynote.
Referring now to
If the token is not already in keyword dictionary 852 and the token is not on the list of non-suggestible words, the token may be suggested as a keyword in processing block 1416. If the token is already present in keyword dictionary 852 or the token is on the list of non-suggestible words, the token cannot be suggested. In this case, the token can be augmented in a variety of ways to render the token distinctly different from other keywords currently residing in keyword dictionary 852. As an example of such an augmentation of a token, the first character of the next sequential token in the input keynote may be used and concatenated with the token currently being processed.
This augmented token may then be compared with the contents of keyword dictionary 852 to determine if the augmented token is not currently present in the keyword dictionary and not on the list of non-suggestible words. If the augmented token is found in keyword dictionary 852 or the augmented token is on the list of non-suggestible words, the augmented token may be further augmented using additional characters of the next token or the previous token in the input keynote. This process continues until a unique and suggestible augmented token is generated. This unique and suggestible augmented token may then be suggested as a keyword to the user in processing block 1418. Finally, all tokens in the input keynote may be appended together as a combined token. This combined token is compared with the contents of keyword dictionary 852 and the list of non-suggestible words to determine if the combined tokens are currently defined as a keyword in the keyword dictionary or as non-suggestible. If not, the combined tokens may be suggested as an additional keyword in processing block 1420.
In an alternative embodiment, the logic for suggesting keywords may employ different heuristics based on the type of object for which keywords are being suggested. Thus, for example as described above for contact objects, it may be desirable to augment a keyword to include characters from subsequent tokens or to create initials from multiple tokens as one may do for the name of a contact. However, this process may not be appropriate for creating augmented keywords for list or project objects. In these cases, a somewhat different augmentation process may be used. The use of different heuristics based on the type of object is easily implemented with the present invention because the type of object will be known at the time the suggesting process is employed. By knowing the type of object for which a keyword suggestion is being generated, the appropriate heuristic may be selected.
Thus, the processing performed by the present invention for parsing keywords is described.
The implementation of the date/time parser of keyword and date/time parser 810 of the present invention is described in the following sections. The date/time parser of keyword and date/time parser 810 uses lexical analysis tool 400 to break an input keynote into date relevant tokens and to identify tokens in the input expression that may be relevant to date parsing. The lexical analysis tool 400 uses a method for breaking the input keynote into date relevant tokens similar to the method described above in connection with
The date/time parser of keyword and date/time parser 810 is responsible for parsing the input keynote to interpret dates and times from the natural language input keynotes as entered by a user. The goal of the day/time parser 814 of the present invention is to parse a set of date and time tokens from an input keynote and determine with perfect accuracy the intended date or time constraint applied by a user. The format used by the present invention is simple, short, intuitive, and powerful enough to allow a user to express almost any date/time value by typing a short and simple expression directly as natural language text rather than being forced to navigate through a maze of dialogs or to manipulate numerous command buttons, check boxes, or other graphical user interface components.
The date/time parsing performed by the date/time parser of keyword and date/time parser 810 uses a novel form of recursive descent with infinite look ahead technique. This technique provides an O(N) complexity methodology. Similarly, the keyword parser of keyword and date/time parser 810 achieves an O(N) complexity. This is the best complexity theoretically possible. Thus, the performance of the present invention is well suited for real-time applications requiring quick response.
The following is a list of date expressions recognized and a list of rules employed by the keyword and date/time parser 810 of the preferred embodiment:
Note: Assume today is Monday, Jun. 10, 1996 while reading this table.
Recurring events can occur on a daily, weekly, monthyly, annual basis. The keyword and data/time parser 810 also handles the entry of recurring events. The keyword and date/time parser 810 recognizes the following types of recurring event specifications and associated date-related keywords or key expressions. Note that other forms may similarly be handled.
Note that recurring dates are assumed to occur forever in the preferred embodiment. The keyword and date/time parser 810 always picks an intelligent starting date, but does not report an ending date. For example, assuming that today is Monday, Jun. 11, 1996, “every Tuesday” will be interpreted by the parser 810 as “every Tuesday” starting Tuesday, Jun. 12, 1996 (the nearest weekday starting from today).
As with dates, abbreviations and numerical/ordinal substitutions are correctly interpreted.
DLL Interface of the Preferred Embodiment of the Present Invention
The following section describes the interface to the parser 300 dynamic link library (DLL) component of the preferred embodiment of the present invention. The generic use of DLL's is well known to those of ordinary skill in the art.
The parser 300 of the preferred embodiment is composed of a single DLL. In this embodiment, the DLL is written in the C++ programming language. It will be apparent to those of ordinary skill in the art that other programming languages, such as C, Basic, etc., may alternatively be used.
The parser 300 capabilities of the preferred embodiment include the following:
From the point of view of parser 300, date and time parsing requires no special knowledge of data. On the other hand, parsing, completing, and suggesting keywords, lists, projects, and contacts requires explicit knowledge of the existing or predefined lists, projects, contacts, and keywords. The parser 300 of the preferred embodiment is data-independent. In other words, it is not aware of any files or databases. Therefore, the parser 300 must be initialized with lists, projects, contacts, and keywords. The initialization process usually occurs during the boot time of the application that uses the parser 300, or when the application switches to another set of data. The following pseudo code illustrates a typical initialization of the parser 300 of the preferred embodiment:
After this initialization, the parser 300 knows about all the lists, projects, contacts, and keywords. It is now able to correctly auto-fill, parse, and suggest keywords upon request from the client. Of course, the parser 300 must be kept in synchronization with the data in the object database 850; changes in the object database 850 should be reflected in the parser 300. Updates are accomplished using Add, Delete, and Rename function calls. As an example, consider the following situation: a user deletes an existing project named “Paint Fence”. The application removes the project from the object database 850 and removes (or updates) its associated keywords. This change must be reflected in the parser 300 and can be done with a single function call as follows:
This single function call will remove the project and any references to it from the parser 300. The project name will no longer auto-complete and all of the keywords that are associated with the “Paint Fence” project will be automatically removed or updated. Note that DeleteProject( ), AddProject( ), and RenameProject( ) return values indicating success or failure of the function. For the sake of simplicity, the previous examples ignore the return values.
The Rename functions support renaming of lists, projects, contacts, and keywords. Renaming a list (“list” is used as an example—it can be replaced by “project” or “contact”) is easily done in the object database 850. It is effectively a simple database update operation. Because the list has a primary key in the object database 850, and the keywords that are associated with the list are related to the list via this key (rather than the list's name), modifying the name of the list will not affect its associated keywords. In other words, the keywords will remain linked to the list after the name change.
The Rename functions of the preferred embodiment are as follows:
A single function call to any of the above functions will handle the entire renaming process, and will simplify the client's task.
Retrieving Results from the Parser
In order to retrieve information from the parser 300, a client must allocate buffers and pass them into the parser 300 DLL via function calls. Parsing information is retrieved using the ParseOutput Data Structure 830, which is defined as follows:
The client of the parser 300 DLL allocates a ParseOutput data structure (“struct” or “type” in Visual Basic), and passes it along with the input expression to be parsed using any of the following function calls:
Keyword suggestions are retrieved from the parser 300 using the KeywordSuggestion structure, defined as follows:
The client of the parse DLL allocates a KeywordSuggestion data structure, and passes it along with an input expression using any of the following function calls:
Auto-completion (auto-fill) requires only a character buffer which can be declared by the client as:
Handling Recurring Dates
The present invention supports recurring date parsing by use of the following fields of the Parse Output Data Structure. These fields are:
Finally, a boolean value is provided to signal whether we are dealing with a simple date or a recurring date:
The parser 300 doesn't require end dates. That is, recurring events are assumed to go on “forever” (the user must use the user interface 200 to specify a full range). The “sEndDate” field in the ParseOutput data structure, is nonetheless provided to support specified end dates. A starting date, however, is always provided when a recurring date is parsed. Although the parser 300 doesn't require starting dates (i.e., every Friday starting on Aug. 8, 1997), it always tries to guess (intelligently) a starting date for the recurring event. This starting date will be passed via the ParseOutput struct in the sDate field.
Collaboration support is provided in the preferred embodiment of the present invention. The parser 300 looks for two possibilities at the beginning of each keynote. Note that other similar keywords triggering collaboration could also be provided.
The parser 300 also accepts a single or multiple punctuation (comma, period, colon, and semicolon) between the contact keyword and the word “please” (e.g., “Brian. Please . . . ” or “Please, Brian . . . ”).
When the parser 300 recognizes this keynote sequence, parser 300 fills the sDelegate field of the ParseOutput data structure with the contact name (pointed to by the contact keyword). The next contact keyword (if any) will appear in the sContact field. For example, consider the two contacts, “Brian Smith” (keyword=“brian”) and “Danny Jones” (keyword=danny”). The input expression:
Further details on collaboration support of the present invention are provided in a subsequent section this document.
The object database 850 of the present invention supports arbitrary association of one type of object in the database with one or more other objects of any type. The object database 850 also supports collaboration (including negotiation and tracking action requests to completion) between two users of the present invention who may or may not share a common server.
Referring now to
Types of Objects
Association of Object Types to Tables Where They Reside
Every type of object listed in
Object Links Table
The links table of the preferred embodiment is a special table in the database of the present invention that allows free association of one object of any type to another object of any type. For example, the links table allows the present invention to associate a Person object type to an Email Address object type. Note that the same Person could also be associated with additional email addresses, each represented in the Links table as separate entries.
Every object in the object database 850 of the present invention has a unique identifier, or key, associated with the object. These keys are stored as part of the record, or entry, describing an object in a particular table. For example, Brian Smiga is an instance of a Person object type with a key of 101; Brian's first name and last name, as well as his object instance key, will be stored directly in the People table as part of a single record.
As indicated previously, every object in the database of the present invention also has a type associated with it. Given the object key and object type of one object and the key and type of another object, the two objects may be “linked via a single entry (record) in the links table, a sample of which is shown in
For example, if “Brian Smiga” represents an instance of a Person object type with a key 101 and “email@example.com” represents an instance of a corresponding Email Address with a key 102, the “Brian Smiga” object instance may be linked to his corresponding “firstname.lastname@example.org” email address instance in the link table as follows:
Referring now to
Given the initial database table content shown in
Further suppose the user attaches the Person “Jim Salter” and the Project “Patents”, also sending an FYI keynote to Tom Hagan as described in detail above. In this example, the parser 300 of the present invention will automatically determine many of the links that need to be established in the tables shown in
Once the new keynote is filed by user Dennis, the following significant data will be recorded in the object database 850 by the present invention:
An envelope (Type=0) will be added as a container for the keynote being sent to a delegate person. The Envelope includes such data as a subject for the note, when the note was used, or when it is received in a standard email client. Assume that this Envelope was assigned a database key of 212 by the object database 850.
A List Item (Type=1) will be added for the note and would read “Brian, please call Jim tomorrow re patent status”. Assume that this List Item was assigned a key of 213 by the object database 850.
Several link table entries would be created for the new keynote in this example. The sample link table resulting from this sample input keynote is shown in
It will be apparent to those of ordinary skill in the art that the object database 850 implementation of the present invention as described herein is of broader applicability than strictly for use with the natural language parser 300 as described herein. In an alternative embodiment, the input text data may be provided as a structured record or buffer from which the object database 850 extracts the information necessary to create the link table shown by example in
Collaboration Between Two or More Users of the Present System
The present system allows a user to manage his or her own actions/projects and time more effectively. In many cases actions/projects have to be handled by more than one person. The collaboration cycle described below allows interaction between users to further completion of actions/projects and to allow information to be efficiently exchanged between users of the present system.
In addition to parsing input text, the parser 300, described above, further utilizes the keywords to “classify” the text entered. In one embodiment, the message types are shared and personal. Shared messages include: FYI and action requests. Personal messages include personal action and personal memo. Of course, other message types may be utilized.
An action request is input text which is sent out to at least one other person, and requires a response. It generally asks another user to do something. In one embodiment, the keyword “please” may initiate an action request. As discussed above, other keywords may be added to the list, at the user's discretion. For example, a user may add the keywords “I need you to”, “pls”, or similar words to indicate that the text entered is an action request. Thus, the sentence “Joe, I need you to call Bill about the Project X deadline” is classified as an action request, targeted at Joe. The contact is Bill, and the project is Project X. There is no date, since the requester did not include a date. Such analysis may be done using the parsing methods described above. However, the present system is not limited to the parsing method described above. Other methods of identifying the target, contact, project and date may be utilized.
An FYI, or “for your information,” is input text that is sent out to at least one other person, and requires no response. It is used to inform others about facts. For example, and FYI might be used to inform others that a new manager has been brought in. Keywords for an FYI type may be “FYI”, “For your information”, “Please note”, and any other keywords which the user included in the keyword list. Generally, the targets of the FYI are deduced from the proximity to the keyword. For example, an FYI which read “John, FYI, Tom is in charge of Project X now.” would send a copy of this FYI to John.
A personal action/memo is text that is not sent to anyone. The absence of the other keywords would indicate that an entry is classified as a personal action/memo. For example, the input text “Call Jim about Project X” is a personal action. It is filed in the user's own system, as described above, but is not forwarded to anyone else.
Framework for Collaboration
The present system may be implemented on a number of different frameworks.
The distributed peer-to-peer system enables communication with users who do not utilize the present system. Thus, the present system allows seamless integration of all action requests, and to-do-lists, regardless of whether the recipient is a subscriber to the present system or not. In one embodiment, the peer-to-peer distributed system model is used for the present invention.
Keeping Track of Action Requests
Utilizing a distributed system necessitates an alternative means of keeping track of action requests. Because there is no server which tracks each action request, a method of identifying each action request and its associated objects, such as project, sender, etc., is needed. This method enables the system to match replies to the proper action request.
In one embodiment, a foreign key table is utilized to keep track of action requests. To clarify, the example described above with respect to
The format of the action request does not reflect the actual format of the action request displayed on the present system. The appearance of such an action request is described below. This action request was entered by Dennis, and sent to Brian. In addition, an FYI copy of the action request was sent to Tom. This is to alert Tom to the action request. When the action request is generated, an envelope is generated, to contain the action request.
The envelope includes a subject, and links to the list item, as well as the e-mail addresses of recipients. Below, only the actions occurring in Brian's system are described. However, similar activity occurs in Tom's system.
When Brian receives the action request, several Link table entries are created for the new action request. These Link table entries parallel the entries in Dennis' table, described above. Since the local key numbers are unique to the database of the individual, these key numbers may be different.
In addition, Brian's system generates a number of entries into a Foreign Key table. The Foreign Key table is utilized in mapping a collaborative action request in one database, part of the sender's system, to the same action request in another database, part of the recipient's system. This link is represented by associating the Envelopes containing the corresponding List Items.
The creator column is the creator of the entries in the Foreign Key table. The creator column displays the device which received the action request. In this example, it was the simple mail transfer protocol (SMTP) plug-in which received the action request from Dennis.
The Local Database column contains a unique identifier identifying the delegate's (Brian's) database. The Foreign Database column contains a unique identifier identifying the requester's (Dennis') database. In one embodiment, the identifier is a Globally Unique Identifier (GUID), which is a 128 bit value based on the Ethernet address of the machine on which the GUID is generated, as well as the time at which the GUID was created. Alternatively, other unique identifiers may be utilized.
The Local Key and Foreign Key are the key numbers from the Database tables of the individuals. The example keys described above with respect to Figure X are included in the Foreign Key table, i.e. Dennis' database entries.
For example, the local key for note creator (Dennis) is the key number associated with Dennis in Brian's database. The “me” key is a special purpose value utilized because the user's database may or may not contain information identifying the user himself or herself, and it is unreliable to match user names via text comparison. This method allows for a match to be indicated. The “my e-mail” key is a similar key for the e-mail address of the database owner. Additionally, a parallel table is generated in Dennis' system when Dennis receives a reply from Brian. Of course, in Dennis' system, Dennis' database ID is the Local Database, while Brian's is the Foreign Database.
Foreign Key entries are also created on both sides (requester and delegate) for the Person and Project linked to the action request, as well as the Creator (another Person) of the action request, the Sender (another Person) of the action request, the Delegate, and any Email Address associated with the action request. The Person, Project, etc. entries are created to ensure that the correct links are created/maintained on reply and that duplicate entries are not created in any user's database.
In an alternative embodiment, the e-mail addresses of the sender and the recipient, coupled with a unique identification attached to the action request itself identifies the action request. In another alternative embodiment, a unique local identification coupled with a public key/private key identification of the sender/recipient is utilized to identify each action request.
At block 2110, there is collaboration between the systems of the requester and delegate or delegates. This collaboration involves a complex series of negotiation steps that are designed to arrive at a final answer. This is described in more detail below.
At block 2120, the process queries whether the final answer is affirmative or negative, or done. An affirmative answer occurs when the parties agree to perform the task. A negative answer occurs when the parties decide to not perform the task.
If the final answer is affirmative, the process continues to block 2130. At block 2130, the final answer is filed in the appropriate calendars and lists, including lists associated with contact, project, if appropriate. These lists will be referred to hereinafter as project/contact list. Where it is filed depends on the interpretation of the original action request, and the collaboration process, as described below.
If the final answer is negative, the process continues to block 2140. At block 2140, the action request, and collaborative updates of the original action request are deleted from the calendar and lists. This process is further described below.
At one point, unless a negative reply was received, the delegate sends a Done reply to the requester. A Done reply may be sent by the delegate using the process described below. Alternatively, when the delegate checks the action/project off his or her calendar and/or project/contact lists, an automatic Done reply may be generated and sent to the requester.
When the requester receives the Done reply, the original action request is marked done in the requester's system. Additionally, in one embodiment, an automatic acknowledgment form is generated. In one embodiment, an acknowledgment form consists of a generic text, such as “Thank you for completing my action request regarding the ‘Project name’ project.” In one embodiment, different types of acknowledgments may be associated with different delegates. For example, the requester may identify certain delegates that should receive a thank you voice mail, or document. In those instances, such an acknowledgment may be automatically generated by the system. In one embodiment, the user may select the type of acknowledgment. In one embodiment, the requester only receives a notification that an acknowledgment should be sent.
At block 2205, the process identifies the delegate. An action request can be addressed to one or more parties. These parties are the delegate. As described above, in one embodiment, the names prior to, or following, the keyword “please” are generally considered the delegate. As described above, the user may have added additional keywords which indicate that the present entry is an action request.
The delegate may be an individual or a group. For example, an action request could be addressed to “managers”. In that instance, the keyword “manager” could include a plurality of managers. In one embodiment, for multiple delegates, separate action requests are spawned for each delegate, and each delegate is dealt with individually. For simplicity's sake, the remainder of this flowchart will assume that there is only a single delegate.
In one embodiment, a header is generated when the delegate is identified. In one embodiment, the header includes a number of fields. In one embodiment, these fields include: Delegate, FYI, and Enclosures. The Delegate field includes the delegates, which are determined as described above. The FYI field is determined in a similar way. In one embodiment, the Delegate field maps to the To field, the FYI maps to the CC field in other messaging applications.
The Enclosures field enables the requester to attach a variety of items to the action request. In one embodiment, the enclosures may include arbitrary files or information about the projects and the contacts related to the action request. This is especially useful when an action request is sent to a delegate who is not using the present system. In that instance, the message received by the delegate may not be linked to database(s) with information about contacts or projects. Thus, by attaching those items, the requester can simplify the recipients' work.
The requester's name is placed in the From: field. The header may also include a subject. The subject may be the first few words of the action request, the project to which the entry was parsed, or may be entered by the author of the action request. In one embodiment, the subject appears as “Action Request: <first few words of request>,” or FYI: <first few words of FYI>.” This makes apparent to the recipient the type of message received, in addition to giving some information about the subject matter of the message. At block 2210, the action request is sent to the delegate. The action request may be sent via electronic mail or any other means. In one embodiment, the action request is sent directly to the delegate's in-box in the system of the present invention. In one embodiment, if the delegate is not utilizing the present system, the action request is sent to the delegate's address. In one embodiment, this may include the delegate's e-mail address, fax number, voice email number, or pager.
Much of the filing and similar actions described in the present application require the use of the system of the present invention. However, action requests may be sent to any individual who has a receiver object, which can receive text or voice in some format. In one embodiment, if the delegate does not have an e-mail connection, the action request can be faxed to the delegate. In one embodiment, the present system may format the action request in a rich text format (RTF) and fax it to the delegate. In one embodiment, the action request may be turned into a voice mail message and sent by the present system. In another embodiment, the action request may be sent as a pager message to a pager.
In one embodiment, the action request is formatted into an ASCII format, which is readable by a user. In one embodiment, the action request is reformatted to read as follows:
It relates to:
The text in quotation marks is inserted based on the information from the action request. This text format is sent, and is readable by delegates who are not subscribers to the present system.
At block 2215, the action request is filed, and penciled into the appropriate calendars and/or project/contact lists. In one embodiment, the action request is filed in the “Waiting For” list. This is a list which contains action requests which have not been resolved. In one embodiment, a copy of the action request is also filed in the project/contact list to which it was parsed. Additionally, if appropriate, the action request is penciled into any lists, projects, contacts, and calendars that it was parsed to. “Penciling” indicates entry into a calendar or project/contact list in a different color. This allows the user to easily identify items which are not yet agreed upon. In one embodiment, penciled items appear in gray, compared to normal entries in black or blue.
At block 2220, the process waits for a reply from the delegate. The process does not remain in a wait state, but rather continues to execute other processes. However, this action request cycle waits for completion. When the reply from the delegate is received, the process continues to block 2225. The reply from the delegate appears in the in-box of the requester. In one embodiment, if there is no reply a reminder is sent before the due date.
At block 2225, the process queries whether the reply is affirmative, negative, or an other category. These categorizations are described below with respect to
If the reply is affirmative, the process continues to block 2230. At block 2230, the original action request is updated and refiled. As described above, the action request is originally in the Waiting For list. Now, the action request is recategorized to the appropriate list, as determined from the parsing. The action request is also placed into the calendars and project/contact lists in ink. In one embodiment, writing in ink comprises entering the request in a different color from penciling. In one embodiment, ink is a black or blue color.
If the reply is negative, the process continues to block 2235. At block 2235, the action request is deleted from the calendar and from the Waiting For category. At this point, the process is closed, since the delegate has refused to complete the request. However, the negative reply remains in the in-box of the delegate. Thus, if the requester wishes to reassign the project, he or she can do so using the copy in the in-box.
If the reply is other, the process continues to block 2240. At block 2240, the requester and delegate negotiate. The negotiation process is described in more detail with respect to
At block 2250, the opened action request is displayed to the delegate. In one embodiment, if the delegate is using the present system, the action request is displayed in the format described above, with respect to
Once the action request has been opened, the delegate may further delegate the action request. Of course, if the delegate is not using the present system, he or she can not do this, except through standard e-mail communication. At this point, in one embodiment, the user can further delegate the action request by adding a “Please ‘new delegate’” to the action request. This, in reparsing, directs the action request to the new delegate. In an alternative embodiment, the user can manually select a new delegate in the header, and thereby forward the action request. The action request is readdressed to the new delegate, and sent on. The original delegate becomes a requester at this point. However, the original delegate is still has to respond to the original requester.
At block 2260, the user is prompted to enter a reply. In one embodiment, when the user opens the action request in the in-box, it appears with reply classification choices in a reply box on the displayed action request. In one embodiment, the reply box is a pull-down menu. In an alternate embodiment, the reply box includes radio buttons, or other means of indicating one choice from a number of listed items.
One of these choices can be selected by the user, as the reply to the action request. In one embodiment these choices include: Yes, Yes if, No, Comment, and Done. The Yes reply indicates that the delegate accepts the delegation, and will perform the action requested. A Yes if reply indicates that the delegate is willing to perform the action requested, but is making a counter-suggestion. For example, the counter suggestion may be to change the meeting date. A No reply indicates that the delegate is refusing the delegation. The Done reply indicates that the delegate accepts the delegation, and has completed the project assigned. And finally, the Comment reply indicates something outside of these categories. For example, if the delegate feels that the requester misunderstands the project, this reply may be utilized. Other reply choices may be incorporated without changing the fundamental purpose of the present invention. In one embodiment, these choices appear when the delegate opens the action request in his or her in-box. In one embodiment, the user must select one of these choices. In an alternative embodiment, no such reply options appear. In that embodiment, the user replies in a free-form text. In that embodiment, a parser is used to parse the user's reply, and fit it into one of the above categories.
When the user selects one of these choices, a reply form is created. If the reply was either a Yes, No, or Done, the reply form is complete. The user need not enter any further information. However, the user may enter further information. In one embodiment, if the reply is Yes, No or Done, a header is automatically added to the reply, and it is automatically sent. The user is not prompted for entry.
If, on the other hand, the choice selected is a Yes If or a Comment, a reply form is automatically generated, with the appropriate header information. In one embodiment, if the reply was Yes If, a phrase such as “Yes, I will do it, if” appears, followed by the cursor. The user can then complete this phrase. In an alternative embodiment, the user can delete the phrase and enter his or her own words. In one embodiment, if the Comment button is selected, the phrase such as “I have a comment,” followed by the cursor is displayed. Again, the user can either finish the phrase, or erase it and write it differently. In this way, the reply format is automatically filled in based on the reply choice selected.
At block 2265, the action request and reply are placed on the appropriate project/contact lists and calendars of the delegate. The information placed on the lists and calendars is based on the parsed action request and parsed reply. In one embodiment, only the date information is reparsed, and all other information remains. When the delegate opens the action request, preliminary project/contact list, contact, and other information is indicated in the shadow of the action request, in parentheses. However, the delegate can change this information, either directly in the shadow, or by changing the information in the reply.
At block 2270, the reply is sent to the requester. In one embodiment, this occurs when the user presses a button. In one embodiment, there is a send button.
At block 2275, the process queries whether the just sent reply was affirmative, a negative, or other. The affirmative reply includes Yes and Done. The negative reply includes No. The Yes if and Comment replies are classified as other. Similarly, if different categories are utilized, any category which refuses the delegation is negative. Any category which accepts the delegation, without attempting to change it in any way is affirmative. Any other categories are Other.
If the reply is affirmative, the process continues to block 2280. At block 2280, the original action request is updated and filed. As described above, the original action request is on the Out-box list of the delegate. At this point, the action request is refiled based on the results of the parsing. The action request is also placed into the appropriate calendars and lists in ink. In one embodiment, placing the request in the calendar in ink comprises entering the request in a different color from penciling. In one embodiment, ink is a black or blue color. In one embodiment, a notification of the reply remains in the requester's in-box. In another embodiment, the user may select whether or not to receive notification.
If the reply is a negative, the process continues to block 2285. At block 2285, the action request is deleted from the calendar and from the Waiting For category. At this point, the request is dosed, since the delegate has refused to complete the request. In one embodiment, a notification of the reply remains in the requester's in-box. In another embodiment, the user may select whether or not to receive notification.
If the reply is Other, the process continues to block 2290. At block 2290, the requester and delegate negotiate. The negotiation process is described in more detail with respect to
At block 2310, the negotiation process starts. At block 2320, the recipient of the last message is prompted for a reply. The negotiation process is entered when the delegate returns a reply which is either a “Yes, if” or a “Comment,” or any other reply which is not affirmative or negative.
Thus, in the first iteration, at block 2320, the requester is prompted for a reply, in response to the delegate's initial reply. The reply form that appears before the requester is very similar to the reply form for the delegate. In one embodiment, the categories that may be selected are: OK, OK If, No, Withdraw/Done, and Comment. These categories parallel the categories of Yes, Yes If, No, Done and Comment. The category names may be changed without affecting the present process. The OK reply means that the requester accepts the change or comment proposed by the delegate. An OK If reply is a counterproposal by the requester. A No is a rejection of the delegate's proposal. A Withdraw is notification of the delegate that the original action request is being canceled, and the delegate no longer has to do anything in connection with the action request. An OK is dassified as affirmative. A No or Withdraw is classified as a negative. And OK If and Comment are dassified as Other. In an alternative embodiment, the reply form does not contain any categories. In that embodiment, the user enters a free-form reply. A parser is used to parse the reply, and determine the reply choice into which it belongs.
As described above, with respect to blocks 2270 and 2320, based on the reply choice selected, a preformatted reply appears, along with the appropriate header information. This simplifies the negotiation process. At block 2330, the reply is sent and a copy of the reply is placed in the appropriate calendars and project/contact lists. The message is also appropriately updated in light of the reply just sent.
At block 2340, the process tests whether the reply just sent was an affirmative, negative or other. As described in the above classification, a delegate's Yes, Done, and a requester's OK are classified as an affirmative. The delegate's No, and the requester's No or Withdraw is classified as a negative. All other answers, i.e. Yes If, OK If, and Comment, are dassified as other. Either an affirmative or a negative answer is a final answer. That is, it is a conclusion to the negotiation. If the answer was either affirmative or negative, the process continues to block 2350. At block 2350, the negotiation process terminates.
If, at block 2340, the answer was found to be Other, the process continues to block 2360 At block 2360, the process tests whether there is an auto-terminate that is activated. In one embodiment, the auto-terminate is an option which a requester can select. The auto-terminate automatically ends the negotiation process after a preset number of exchanges. In one embodiment, the user enters the number of exchanges after which the negotiation ends. For example, if the parties can not agree after five e-mail exchanges, the process automatically terminates. This is a method to avoid endless cycles of negotiation when it is apparent that the parties can not agree.
If the auto-termination process is not activated, the process-returns to block 2320, and prompts the recipient of the last message for a reply. For example, if the last message was written by the requester to the delegate, the process prompts the delegate to respond to the message.
If, on the other hand, the auto-termination process is activated, at block 2360, the process continues to block 2370. At block 2370, the final answer is set to a No. This indicates that no agreement was reached between the requester and the delegate. The process then continues to block 2350, where the negotiation process terminates.
At block 2410, the process starts. This occurs when a user enters information started with a keyword which indicates that the data entered is an FYI. The FYI is parsed, as described above. At block 2420, the FYI is filed in the system of the originator.
At block 2430, the recipient or recipients are identified. In one embodiment, the name, names, or group names which appear next to the keyword indicating that this is an FYI are identified as the recipient(s). At block 2440, headers are added to the FYI. The header contains the To: field, From: field, cc: field, enclosures: field, and a subject field. These header entries are as described above with respect to the action request. At block 2450, the FYI is sent to the recipients. When the recipients receive the FYI, they can open it and file it. In one embodiment, no option to distributed reply is provided. In this way, no extraneous communication about information is encouraged.
At graphic 2500, Dennis is entering text. As described above, the text is being concurrently parsed. Thus, the project, contact, date and appropriate project/contact list come up in the shadow 2515, as Dennis types in the action request 2512. As discussed above, the parsing algorithm may be the algorithm described above, or any other algorithm. In one embodiment, as Dennis is entering the action request 2512, the parser is parsing the text. When the keyword indicating that this is an action request is found, a header is placed on the action request. The spaces in the header, such as delegate, FYI, enclosures etc. are added as they are determined by the parsing algorithm and Dennis' actions. In an alternative embodiment, the action request 2512 is parsed only when Dennis indicates that he has finished entering text. When Dennis finishes entering the action request, he sends it.
At graphic 2502, Tom has received the action request, and opened it. The header 2425 that was automatically generated is displayed. Underneath, the original action request 2530 is displayed. Above the header, a reply block 2520 is displayed. The reply block 2520 displays the possible reply options. In one embodiment, the reply block 2520 is a drop-down list from which one reply may be selected. In another embodiment, the reply block 2520 may be in any other format which allows the user to select a reply. In this instance, the Yes If reply has been highlighted. For contact and project information, suggestions derived from the sender are provided to the recipient initially. In one embodiment, the project name and contact name are in parentheses. This is to indicate that the project and contact names may not be the same for Tom as they were for Dennis. As described above, the keywords vary, because each user can enter his or her own keywords.
At graphic 2504, Tom has selected the Yes If reply button from the reply block 2520, and the reply 2545 is displayed. In one embodiment, reply automatically starts with a “Yes, I'll do it, if . . . ” The cursor is placed behind that phrase, enabling Tom to complete the phrase. In this instance, Tom has completed the phrase by typing “we can reschedule for Tuesday.” The text entered by Tom is differentiated by being placed in italics in this instance. It is understood that in the actual application, the text need not be differentiated in this way. In the shadow 2540, the project name has been altered. This may be done manually by the user. Alternatively, the process may parse the original note, using Tom's databases and keyword lists. In this instance, the contact name remained the same. However, the project name was changed. Each user is responsible for naming his or her own projects, since two users may refer to the same project by different names. At this point, Tom may send the reply. At that point, a header is placed on the reply, and it is sent back to the requester, i.e., Dennis.
At graphic 2506, Dennis has received Tom's reply and opened it. In the shadow 2560, the new date appears. Tom suggested a new date for the meeting. Because the reply was reparsed for this factor, the new date/time appears in the shadow 2560. Once again, a reply box 2550 appears. Because Dennis is the requester, a slightly different reply box 2550 appears. The entries are explained above, with respect to
At graphic 2508, Dennis' system files the finalized information. In one embodiment, the finalized information is displayed in a box showing the history of the communications between the parties. The original shadow 2515 is updated to an updated shadow 2570. The original penciled entries in Dennis' calendars and lists are also updated to be in ink. This indicates that an agreement was reached.
At graphic 2510, Tom's system files the finalized information. Additionally, the original shadow 2535 is updated to a new shadow 2585, containing the finally agreed upon information. The original penciled entries in Brian's calendars and lists are also updated to be in ink. This indicates that an agreement was reached.
Thus, through this process, one instance of the present invention has been illustrated in a graphical form. It is understood that the actual screen displays may not be identical to the displays illustrated in this Figure. In this way, the interaction between a requester and a delegate is simplified. This allows a delegate and a requester to arrive at a mutually satisfactory way to complete actions/projects. It provides sufficient flexibility for both parties, and works as an automated conversation type of automated negotiation.
The present invention can be used to receive and parse an input keynote in the manner described above. In addition, the present invention supports collaboration with other users, each of whom may or may not have the functionality of the present invention. Because a receiver of a keynote may not have the functionality of the present invention, the present invention automatically formats a conventional electronic mail message which can be sent to a receiver or delegate of a keynote. Using the parser 300 as described above, the present invention takes an input keynote and builds the conventional electronic mail message from information associated with keywords matched in the input keynote. In addition, the present invention also builds a scripted (i.e., encoded) version of the electronic mail message. The scripted version of the electronic mail message is formatted in a structured form to allow a receiver or delegate of the keynote to process the message automatically if the receiver or delegate has the functionality of the present invention. If the receiver or delegate has the functionality of the present invention, the scripted version of the keynote can be interpreted and automatically processed as a keynote on the receiving end. Thus, the present invention, by building an electronic message with both a conventional text format and a scripted format combined in the same automatically generated electronic message, allows keynote collaboration with anyone on the receiving end of such an electronic mail message.
The following sample automatically generated keynote electronic mail message illustrates the dual format electronic message structure automatically created by the present invention given the following input keynote:
A sample electronic mail message produced by the present invention from the above input keynote follows. The first portion of the message below represents the conventional electronic mail format (ASCII) readable by a receiver without the functionality of the present invention. The second portion of the message starting with the second occurrence of the text string, “-dreldbssbtdwrvkval” represents the scripted version of the message which can be processed by a receiver with the functionality of the present invention. Thus, this sample electronic mail message illustrates the dual format electronic mail message generation capability of the present invention. A further explanation of the collaboration capability of the present invention is provided in a later section of this patent application.
Thus, a natural language based information organization and collaboration tool for a computer system is described. Although the invention has been described with respect to specific examples herein, it will be apparent to those of ordinary skill in the art that the invention is not limited to the use of specific examples but may extend to other embodiments as well. The present invention is intended to include all of these other embodiments as defined in the following claims.