This disclosure concerns delivery of facsimile (fax) documents over a value added network, such as a store-and-forward network, and more particularly to an automated and integrated method and apparatus to facilitate delivery of a fax document after an initial delivery attempt is unsuccessful.
As a mechanism to carry information over long distances, store-and-forward (S&F) networks offer an efficient, low-cost alternative to the existing public switched telephone network (PSTN). In general, S&F networks operate parallel to, and are accessed by, the PSTN.
In summary, transport of information from the source fax machine to the destination fax machine using the S&F network requires three distinct steps:
Store-and-forward networks offer a number of significant advantages over standard telephone networks for transport of facsimile. For example, a fax document can be carried 16 times more efficiently using packet technology employed by S&F networks. A common annoyance in telephony is the inability to complete a call, usually because the destination device is busy or does not answer. Although sophisticated voice mail systems have been designed to overcome this problem in voice telephony, similar practical and cost effective solutions do not exist for fax. S&F networks offer a viable solution. A properly implemented S&F network will employ a sufficiently large number of telephone circuits such that a customer fax machine never encounters a busy signal. At the destination end, it is a common practice to design into S&F networks the ability to automatically redial those call attempts which encounter “busy” or “no-answer” signals. Normally, the calls are redialed periodically over a fixed interval of time, every ten minutes for a half hour, for example.
Since multiple messages are typically coursing through an S&F network at any point in time, it is important to have some mechanism to monitor the location and status of each. For example, in one known S&F network, a small data file called an envelope is created to track each fax document as it moves through the network. The source node creates the envelope after it receives an incoming fax document. As the fax document moves through the network, the envelope moves between the network devices and receives continuous updates regarding the status of the fax. This enables substantially real-time monitoring of the fax delivery process.
In the known S&F network, upon concluding the delivery attempt process, the destination network node declares the fax document either “delivered” or “not delivered”; it records the status in the corresponding envelope which is then returned to the source node. If the delivery was successful, the envelope is forwarded to a historical database (HD) which provides a basis for constructing customer bills. If the delivery was not successful, the envelope is forwarded to a delivery assist system (DAS) for further processing. DAS is a database management system which provides a human operator, the document delivery analyst, with the delivery history and options for resubmission to the network of the document in question. One of the possible actions the analyst may take is to assign the fax document to an alternate destination number, i.e., one provided by either the sender or the receiver.
While the delivery analyst (human operator) enables the network provider to arrange delivery of most fax documents, and to provide the customer with an on-going report on alternative delivery attempts, the cost of providing such services are substantial. Furthermore, as the amount of traffic on the network increases, the number of documents requiring assistance increases, and it becomes more and more difficult to provide such human-assisted delivery on a timely and cost-effective basis.
The alternative delivery attempts which the prior art network can make by itself to deliver a document are quite limited. For example, a fax card with hard coded instructions may be provided in each network node which, based on call progress tones heard on the network, may institute an automatic retry. The sounds which may be heard include:
Generally, the device automatically retries a number of times in a given time period or cycle, e.g., every five minutes for a half hour, and if delivery is still unsuccessful, the document is transferred to a human analyst.
Thus, the prior art automatic retry device provides a fixed response based on a response received to a first delivery attempt. This system has limited utility and most documents with delivery problems end up being sent to a human analyst. As previously indicated, the cost of providing such human analysts are escalating as the amount of network traffic steadily increases.
A method is provided for facilitating delivery of a fax document from a source to a destination over a network, when an initial delivery attempt has been unsuccessful. The method includes the step of initiating a rule-based process to determine a next action based on a time-variable set of input conditions to the process.
For example, the input conditions may be determined from one or more of:
Based on these input conditions, the process determines a next action which causes a next delivery attempt, and if the next delivery attempt is not successful, the process is repeated.
The next action may include one or more of:
Where additional information is requested, the responses received are a further input condition and the process then redetermines a next action based on the new input conditions.
In particular embodiments, the input conditions may include an identification of non-business days at the destination. The non-business days may be categorized by one or more of the country (in which the destination is located), region, and destination number. Input conditions may also include an identification of the non-business hours at the destination.
The process may be implemented by traversing a hierarchical decision tree, namely a data structure graph with one starting point known as a root and many end points known as leaves. Rules control the movement through the decision tree. The rules determine where to get information and how to compare that information to pick an appropriate path in the decision tree. An action is a leaf of the decision tree.
The rules consist of two parts, conditions and actions. Once a condition is met, an action or series of actions is triggered. The hierarchical tree may be constructed as a series of hierarchical tables, wherein each table will contain a set of rules with unique conditions. The top rule table will be searched first to determine if a corresponding condition can be matched. Once a match is found, the system will then handle the document accordingly.
The input conditions consist of attributes equaling a certain value. Some of the common attributes which may be used are response, cycle, destination business hours, delivery instruction, past action taken, and alternative network tried. Generally, the conditions will be unique such that each situation will find only one or no matches in a particular table.
In accordance with the apparatus disclosed herein, an integrated and automated document handling system is provided for facilitating delivery of a fax document from a source to a destination in a network. The system includes an automated rule-based process for determining an action based on a set of time-variable input conditions. The input conditions may be derived from one or more of the destination, the source, a database of past delivery attempts, and a human analyst. The output from the rule-based process causes a predetermined action based on a predetermined set of input conditions.
These and other features and benefits of the disclosure will be more particularly described in regard to the following detailed description and figures.
One of the major advantages of the delivery expert system (DES) and method disclosed herein is a reduction in the cost associated with documents requiring assisted delivery. The system is automated in that a document is owned by the delivery expert system and comes out of automation only for steps that need human intelligence and then automatically goes back to the control of the delivery expert system.
Decisions are made by the DES based on rules. Rules consist of two parts, conditions and actions. Once a condition is met an action or series of actions is triggered. Each table will contain a set of rules with unique conditions. When an envelope associated with a non-delivered document enters the DES, an attempt is made to match the non-delivery condition with a corresponding condition in the rule table. Once a match is found, the DES will then handle the document accordingly. If no match is found, DES passes the envelope on to a delivery analyst. The system is designed to be robust and scalable.
The left hand box 73, designated “Historical Knowledge”, is a database of past delivery information. This embodiment includes historical records which describe all past delivery attempts of all documents to a given destination number. Also included are records describing the past delivery attempt (cycle) of a given document. This information is provided by the envelope created for each fax document as it enters the network and which is updated with delivery information as the document travels through the network. The historical knowledge (prior delivery information) comprises the input conditions that are supplied to the rule-based process for determining a next action.
The lower box 74 designated “Delivery Communication” includes requests sent to the source or destination for additional delivery information, and the response thereto; again it provides additional input conditions to the rule-based process. For example, a source (customer) may be contacted via fax with a request to provide an alternative number for the destination of a non-delivered document. The rule-based process may automatically initiate the generation of such request when certain conditions are met during traversal of the rule-based decision tree. Generally, the response would be provided to both the rule-based process and to the historical knowledge database.
Another possible input condition is a new delivery instruction (NEW DI) sent by the customer (with or without a request). Upon receipt, the rule-based process may automatically flag all documents affected by the NEW DI so that the NEW DI is implemented as the next action (before traversing the other portions of the decision tree). For example, in a particular embodiment, a process scans a database of NEW DI records to determine if any new delivery information is available. The search is based on any delivery information that has been added to the system since the last time a given record was modified. If new information is found, the record is placed in a delivery assist envelope table with a status of “NEW DI”. This record is then sent to the DES rules engine to determine an appropriate action based on the new information.
The right-hand box 75 designated “DA Services” represents all steps that a delivery analyst (human operator) may take and then pass his/her findings to the DES. These steps may be separated into three categories: (1) steps which require the delivery analyst to gather more information so that the DES can make the next decision; (2) steps which require the delivery analyst to take action outside a document, such as sending a request to a destination or source for further information; and (3) steps where the conditions are not yet defined in the rules and a document will need to be manually handled.
In the first category, where the delivery analyst gathers more information which it passes back to the DES, there are included the process steps of investigation, research and technical. Investigation is a step where a fax number is called because an unidentified response needs clarification. For example, the DES passes a document to DA services because of a rule that says to pass back undefined voice responses with an investigate status. The delivery analyst then calls the fax number and if he/she hears an “out-of-service” recording, the delivery analyst redefines the response to be an “out-of-service” recording, and passes this response back to the DES. The DES then finds the appropriate rule to handle “out-of-service recording” documents and the DES takes the appropriate action.
In another example, the human operator may call a fax number and the response is a live person. In this case, the investigation turns into research and all information listed under research needs to be passed back to the DES.
Research is the step where a voice number is called to research the cause of a failed delivery attempt and determine a solution. For example, the delivery analyst may call a voice number and get an alternate fax number. The delivery analyst will then type in the alternate fax number and pass the information to the DES that research is complete with an alternate number. The DES would then immediately resubmit the document to the alternate number based on the rules. Alternatively, if the delivery analyst had called the voice number and got no answer, the delivery analyst would pass to the DES that he/she got no answer. The DES would then, based on the predefined rules, retry the document for a certain period until the next condition to research again or another action came up.
Note that investigate and research could happen at the same time so the system must be able to pass all appropriate information back no matter what was the original reason the DES contacted the delivery analyst.
Technical is the step where certain documents need to be passed to a delivery analyst for control because of technical problems. As an example, there are areas of India where a fax tone sounds like a busy signal (to a U.S. human analyst). This region should be flagged proactively so that any documents going to this region will be automatically retrieved and appropriate pauses added to the dialing pattern to insure delivery to the destination number.
In other cases, a document having a technical problem may be flagged on a reactive basis. For example, a document for which the network provider continually receives a broken connection response on the same page is often due to the page having too many scan lines. In this case, the delivery analyst will call the customer and request that the page be resent.
Returning to
By way of example, every December, the director of document delivery (human operator) executes a process to input holiday dates for the upcoming year. The default holidays, for which the dates do not vary from year to year, are automatically carried forward from the preceding year. Holidays that do not fall on the same date each year must be updated manually by entering the appropriate date. Each record, accessible by the user, may include the country, year, holiday date, and description, as set forth in the following examples; two records are shown, the first being the result of querying based on country, and the second being the result of querying based on date:
Similarly, the country business hours tracking system allows the document manager to specify standard business hours for every country (region and/or destination) in the world. The record specifies the hours the businesses are open versus closed for each day of the week. By specifying an open or closed value for each of the 24 hours, this provides flexibility in specifying odd closure hours (that is, lunch, half days on weekends, etc.). The user may query this information based on country and/or day of the week. A sample record is set forth below where O=open and C=closed:
In a particular embodiment described herein, the delivery expert system may be implemented as a network server with the following four components:
The service entity contains the rule-based process and is responsible for receiving envelopes, determining courses of action, and carrying out the actions. The supported actions may include:
As stated earlier, rules consist of two parts, conditions and actions. Once a condition is met an action or series of actions is triggered. Each table will contain a set of rules with unique conditions. When a failed delivery attempt occurs, the top rule table will be traversed until a corresponding condition can be matched. Once a match is found the DES will then handle the document accordingly.
Conditions consist of attributes or parameters equaling a certain value. Some of the common attributes include response, cycle, destination business hours, instruction type, past actions taken, and alternative network tried.
The values of attributes will use common logical signs such as equal, does not equal, greater than, less than, as well as a fixed number or text string. In addition, the value of attributes utilize true and false logic. For example, in order to see if a broken connection is occurring on the last page of a document, the condition can be set up as:
Conditions are unique in that each finds only one or no matches in a table. A met condition evokes an action. Some actions involve actual handling of the document such as resubmitting the document to the network or sending the document to a delivery analyst; other actions may trigger the sending of a request to a customer or destination for additional delivery information. Multiple actions may take place.
An action can be placed on “hold” (i.e., deferred) until it is triggered by a future event or time. This is illustrated in greater detail by the examples below.
Instead of putting all of the rules into one huge table, it is easier to organize the rules into several tables based on the conditions in that table. When a document needs delivery assistance, the process will search the first table to see if there is a matched condition. If the condition is matched the appropriate action would be taken; if not, the DES would then search the next table.
In one example, three types of tables may be provided in a hierarchy. For example, a standard rules table may be set proactively (i.e., prior to this specific document experiencing a problem). The table may be traversed in the following sequence:
As an additional example, a temporary rule table may be based on temporary instructions received from the customer on how to handle this particular document or destination number. Again, there may be three tables accessed in the following sequence:
The rules may be set up by the delivery analyst while servicing a document or while a document is out in the network.
As a further specific embodiment,
A document would go through these rules in the designated order. A more detailed description of the rules is set forth in
The following abbreviations are used in the figures described below:
BC broken connection
DAC delivery analyst center (i.e., human operator)
BH business hours
FBH first business hours
LBH lunch business hours
INV investigation
WFPB world fax phone book
HISTORICAL record of past delivery information of all documents to a given destination number
DOCTRACT record of delivery information for last delivery attempt (cycle) of this document
DOC fax document
DN document track notes
CST current standard time for destination
TECH technical
NW network
NBH non-business hours
BZ busy
NO ALT no alternative
PG page
DES delivery expert system
BC0/0 broken connection with no page delivered
ALT # alternate number
DA delivery analyst
PBH public business hours
OOS out of service
The retry tables are described in greater detail in
The functionality of the DES system can be achieved in software applications executing on standard PC platforms.
Various features of the invention may be implemented using a general purpose computer 161 as shown in
The invention may implement a memory, such as a floppy disk, compact disc, or hard drive, which contains a computer program or data structure, for providing to a general purpose computer instructions and data for carrying out the functions of the specific embodiment.
The parameters of the above process may be varied as desired to obtain a particular cycle time, i.e., a cycle being one pass through the rule-based process. For example, the number of retry attempts upon receipt of a busy, no answer, nonfax, voice or other signal can be predetermined and will effect the cycle time. Each cycle provides an opportunity for a change in the input conditions and thus a new rule-determined action (output) on the next cycle. A balance may be sought between the time per cycle and the number of cycles to maximize efficient resolution of most delivery assisted documents without extensive processing time or network overhead traffic.
In other embodiments, the system may be used for other message delivery such as delivery of E-mail (electronic mail) or voice mail over an S&F network. Thus, as used herein, the term document may include a fax document or other message such as an E-mail or voice mail message.
These and other modifications and improvements of the disclosure will be understood by a person skilled in the art and are intended to be included within the scope of the claimed invention.
This application is a continuation of U.S. patent application Ser. No. 12/535,606, filed Aug. 4, 2009 now U.S. Pat. No. 7,961,344, which is a continuation of U.S. patent application Ser. No. 11/828,408, filed Jul. 26, 2007, now abandoned, which is a continuation of U.S. patent application Ser. No. 10/662,858, filed Sep. 15, 2003, now U.S. Pat. No. 7,265,861, which is a continuation of U.S. patent application Ser. No. 08/739,625, filed Oct. 29, 1996, now U.S. Pat. No. 6,747,761, all of which are incorporated herein by reference.
Number | Name | Date | Kind |
---|---|---|---|
4991204 | Yamamoto et al. | Feb 1991 | A |
4994926 | Gordon et al. | Feb 1991 | A |
5065426 | Greenstein et al. | Nov 1991 | A |
5237428 | Tajitsu et al. | Aug 1993 | A |
5291302 | Gordon et al. | Mar 1994 | A |
5398277 | Martin, Jr. et al. | Mar 1995 | A |
5552901 | Kikuchi et al. | Sep 1996 | A |
5608786 | Gordon | Mar 1997 | A |
5782687 | Sniegocki | Jul 1998 | A |
5790639 | Ranalli et al. | Aug 1998 | A |
5802271 | Hashimoto et al. | Sep 1998 | A |
5805299 | Sakayama et al. | Sep 1998 | A |
5832455 | Hayashi et al. | Nov 1998 | A |
6215765 | McAllister et al. | Apr 2001 | B1 |
6747761 | Ranalli et al. | Jun 2004 | B1 |
7265861 | Ranalli et al. | Sep 2007 | B1 |
7961344 | Ranalli et al. | Jun 2011 | B2 |
20020176117 | Randalli et al. | Nov 2002 | A1 |
20030117667 | Eguchi | Jun 2003 | A1 |
Number | Date | Country |
---|---|---|
0 400 771 | Dec 1990 | EP |
64-081453 | Mar 1989 | JP |
04-212564 | Apr 1992 | JP |
05-048811 | Feb 1993 | JP |
05-083514 | Apr 1993 | JP |
05-304569 | Nov 1993 | JP |
05-327952 | Dec 1993 | JP |
08-079298 | Mar 1996 | JP |
10-329650 | Dec 1998 | JP |
11-308430 | Nov 1999 | JP |
Number | Date | Country | |
---|---|---|---|
20110216366 A1 | Sep 2011 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 12535606 | Aug 2009 | US |
Child | 13110795 | US | |
Parent | 11828408 | Jul 2007 | US |
Child | 12535606 | US | |
Parent | 10662858 | Sep 2003 | US |
Child | 11828408 | US | |
Parent | 08739625 | Oct 1996 | US |
Child | 10662858 | US |