 |
Description  |
|
|
FIELD OF THE INVENTION
This invention relates to systems for voice processing and methods of
providing voice processing functions over telecommunications lines. More
specifically, this invention relates to systems and methods for performing
integrated voice processing functions and transactions in a
general-purpose computer, common-platform environment.
BACKGROUND OF THE INVENTION
Computer-based telecommunications systems have proliferated in the last few
years along with the common proliferation of high-speed personal computers
and the generally lower costs of equipment now available for use in
complex telecommunications applications. With the use of high-speed
telephone switching lines, telecommunications applications are exhibiting
rapid advancements in technology and versatility. One of the areas in
which telecommunications has experienced rapid advancements is the "voice
processing" industry, wherein telephone lines provide communication links
between users calling in to obtain information from a computer-based
system that is adapted to provide information about a particular business
or organization.
The voice processing industry provides "voice-based" systems which interact
in varying degrees with users seeking information from the system.
Voice-based systems have evolved over the last several years into discrete
systems which accomplish specific tasks. Thus, the voice processing
industry is broken up into a series of sub-industries, each filling niche
technologies or "sub-technologies" which are occupied by particular
providers and which are further segregated according to the products and
services available in the specific sub-technology area. Generally, the
voice processing industry has developed the following sub-technology
areas: voice massaging ("VM") technology, call processing ("CP")
technology, interactive voice response ("IVR") technology, and a number of
other limited technologies which at the present are not large and do not
command significant market shares, such as for example, the "FAX voice
response" technology area. VM systems automatically answer calls and act
as "automated attendants" to direct the calls to the proper person or
department in an organization. These systems have in the past usually
comprised look-up databases that perform voice functions for the user as
the user accesses the system. VM technology can be adapted to read
electronic mail to a user or caller on a telephone, and may also provide
means for storing incoming facsimile messages for forwarding these
messages over TOUCHTONE telephones when so instructed. Systems that fall
under the VM category may also be adapted to recognize spoken phrases and
convert them into system usable data.
Previous VM systems are exemplified in U.S. Pat. No. 4,585,906, Matthews et
al. The Matthews et al. patent discloses an electronic voice messaging
system which is connected with a user's telephone communications network
or private branch exchange (PBX) to provide VM functions for the user. See
Matthews et al., col. 4, lines 49-66.
Another example of a VM system is disclosed in U.S. Pat. No. 4,926,462,
Ladd et al. The device of the Ladd et al. patent provides methods of
handling calls in a VM system based on information supplied by a PBX. See
Ladd et al., col. 4, lines 50-52. VM systems taught in the Ladd et al.
patent comprise a feature phone emulator interface which emulates known
PBX compatible feature phones having multiple line capability. The feature
phone emulator is interfaced to the PBX as an actual feature phone, and
the PBX is configured to assign a group of extension numbers to line
appearances on the feature phone. The VM systems disclosed in the Ladd et
al. patent answer the calls to these extensions by using the feature phone
emulator interface. See Ladd et al., col. 4, lines 53-65.
Yet another VM system is disclosed in U.S. Pat. No. 4,811,381, Woo et al.
The Woo et al. patent VM system which is connected to a trunk side of a
PBX in a business telephone system. The VM system described in Woo et al.
provides the feature of answering forwarded calls with a personal greeting
from the party whose phone is accessed by a user. See Woo et al., col. 2,
lines 37 through col. 2, lines 40-54.
If on the other hand a customer requires a voice processing system to
perform on-line transaction processing and interact with a caller to
answer routine questions about the status of an account, for example, the
customer's requirements are usually best addressed by an IVR system which
can be viewed as fulfilling requirements presented by a totally different
set of architectural problems. Essentially, in an IVR system the user
desires to talk to a central processing unit (CPU) to obtain database
information. IVRs are particular useful in the banking industry wherein
account holders can call a CPU to get account balances and other relevant
information. Generally, IVR systems must also interface to a TOUCHTONE
telephone to allow the caller to provide meaningful data to the IVR system
which then can return meaningful information to the user.
When a retail company wishes to sell large volumes of merchandise through a
"call-in ordering" system, it requires a call processing (CP) system. In
the past, before CP systems were available, such retail companies utilized
"agents" to handle incoming calls. The agents typically manned a
switchboard that allowed manual input of user orders to an ordering system
which could have been computer-based. CP technology today provides
automatic call distribution (ACD) which allows a company to nearly
eliminate the need for live agents handling phone calls, and replaces the
agents with an interactive telephone system through which products can be
ordered. The products can be paid for by credit cards having credit card
numbers which are input through a TOUCHTONE telephone to a computer
ordering system for billing purposes.
Other examples of CP technology are taught in U.S. Pat. No. 4,850,012,
Mehta et al. The Mehta et al. patent discloses a CP system for
intercepting incoming calls to a key telephone system, and returning a
message to a calling party. See Mehta et al., col. 2, lines 11-17. The
Mehta et al. system further provides an intercom line for providing voice
announcements or messages through the key telephone system to the called
parties. CP systems described in Mehta et al. comprise a call processor
which intercepts telephone calls wherein an instructional message is
returned to the calling party, thereby informing the calling party to
select a party associated with the key telephone system by dialing a
pseudoextension number associated with each party. See Mehta et al., col.
2, lines 18-28.
Other technologies have been developed to provide the particular services
and solutions to other niches and subtechnologies in the voice processing
industry. Interactive FAX voice processing is a burgeoning sub-technology
area and has required specialized technical advancements to provide
efficient voice-activated FAX systems. The technical advancements required
to make FAX voice processing and other advanced voice processing systems
feasible have not heretofore been adequately developed. There is a
long-felt need in the art for a general-purpose system which can
effectively, economically, and efficiently provide these technological
advancements and which will integrate the above-mentioned other
voice-based technologies in the voice processing field.
Examples of such systems for data reception and projection over telephone
lines are disclosed in U.S. Pat. Nos. 4,481,574, DeFino et al., and
4,489,438, Hughes. Both the DeFino et al. and Hughes patents teach
hard-wired systems which interface to telephone lines and computers to
provide telecommunications applications. However, the systems disclosed in
the DeFino and Hughes et al. patents generally perform the
telecommunications transactions in hardware, thus requiring expensive and
bulky equipment to accomplish these applications.
All of the above-referenced patents disclose voice-based systems which are
discrete and which perform narrow, limited voice-based transactions. If a
customer needs a voice messaging system, a device such as that disclosed,
for example, in the Matthews et al. patent could be purchased. However, if
the customer also needs a system to interact with callers and to answer
routine questions about the status of, for example, their bank accounts, a
separate IVR system would be necessary. Similarly, if a customer needs to
perform retail ordering and accounts management, a separate CP system such
as that disclosed in the Mehta et al. patent must be purchased. Thus, it
can be seen that the problem facing a customer who requires multiple voice
processing functions is that of the proliferation of a multitude of
special purpose systems that are expensive to purchase and to maintain,
and which potentially process telephone calls in separate and disjoint
manners.
An illustrative example will provide to those with skill in the art an
appreciation of the magnitude of this problem. Consider a bank that allows
its users to inquire about the balance of their accounts using an IVR
system, but must now transfer a call to a VM system if the caller wishes
to leave a message for an officer of the bank that could not be reached.
This creates several problems for both the bank and the user. First, the
bank must purchase and maintain at least two voice processing systems, an
IVR system and a VM system. Second, the user must wait while one system
addresses the other system to provide the new voice processing function.
Third, the bank has no way of getting a consolidated report of the
handling of a given call from start to finish. Fourth, if the user decides
that since the bank officer is not available and the IVR system can
provide additional information to answer a particular question, the
transfer back to the IVR takes a considerable amount of time and is
complicated since the user must usually enter the entire identification
password information again, thereby leaving the bank without any way to
trace a particular call as it is routed from one discrete voice processing
system to the other.
As more discrete voice processing systems proliferate in a single
environment, the problem of multiple disjoint systems becomes even more
complex. There is a longfelt need in the art for methods and systems which
integrate the various disparate voice processing functions to provide a
voice processing system which effectively and economically provides all of
the desired voice processing functions for a customer. This need has not
been fulfilled by any of the prior voice processing systems heretofore
discussed, which only focus narrowly on one particular sub-technology in
the complex and ever-growing array of voice processing sub-technologies.
A proposed solution to solve this long-felt need has been to connect a VM
and IVR system together through a signalling link that coordinates the two
systems. This link allows the systems to exchange calls with proper
information relating to each call and which generates consolidated
reports. However, the customer must still purchase discrete systems, and
this solution is akin to suggesting that the customer purchase a personal
computer with a word processing package of choice, another personal
computer with, for example, a spreadsheet program, and yet another
personal computer with a graphics program. Clearly, this is a cost
prohibitive and ineffective way of performing a plurality of voice
processing tasks and is not acceptable in light of the realities of
today's business markets.
Another proposed solution to the integration problem has been to package
two or more discrete systems in a larger cabinet. Usually, systems having
a large cabinet have nothing in common except the cabinet itself. The
systems may have their own separate consoles and keyboards, or they may
have an A/B switch to share a single console yet still retain their
individual keyboards. In all such "cabinet" systems, there is coexistence
of applications but not integration of applications. Furthermore, systems
which provide coexistence of applications usually provide hard-coded
software in C-language, while the rest of the application development
environment consists of C-language functions and programmer documentation
that can only be understood by an expert programmer, but not by a customer
who may require versatility and ease of use. Thus, the aforementioned
integration attempts do not solve the long-felt need in the art for a
truly integrated voice processing system.
Yet another attempted solution to the integration challenge has been to use
a fixed VM system or a fixed IVR system and modify the resultant composite
system to provide VM and IVR functions for execution in tandem on a common
computer. The results of such machinations have been mixed, and the
customer generally ends up with an inflexible VM or IVR system wherein the
limitations and problems of one half of the system dictate the abilities
and utility of the other half.
Examples of attempts at integration can be found in U.S. Pat. No. 4,792,968
to Katz. The Katz patent discloses a system of analysis selection and data
processing for operation and cooperation with a public communication
facility, for example a telephone system. See Katz, col. 1, lines 57-60.
The systems disclosed in Katz provide methods of selecting digital data to
develop records for further processing and allowing a caller to interface
directly with an operator. See Katz, col. 1, lines 62-68. Another example
of an attempt at integration may be found in U.S. Pat. No. 4,748,656 to
Gibbs et al. The Gibbs et al. patent discloses an interface arrangement
implemented on a personal computer to provide business communication
services. See Gibbs et al., col. 2, lines 8-12. The personal computer
interprets appropriate control signals which are then forwarded under
control resident software to activate a telephone station set and provide
communication services. See Gibbs et al., col. 2, lines 18-28.
Another integration attempt is disclosed in U.S. Pat. No. 4,893,335 to
Fuller et al., which teaches a telephone control system that produces
control signals which are programmable to provide a variety of control
functions to a remote user, including for example, conferencing and
transferring functions. See Fuller et al., col. 2, lines 7-44. However in
all of the above-referenced attempts at integration, only limited
applications are achievable and significant problems of interfacing the
different voice transactions are encountered. These aforementioned
attempts at integration simply do not provide high level and effective
voice transactions.
The inventor of the subject matter herein claimed and disclosed has also
recognized another problem facing the task of integrating VM with IVR and
other voice processing systems. Caller interfaces present a significant
problem in integration since VM systems generally have fixed, hard-coded
interfaces. In an integrated environment, this restricts the versatility
of the entire integrated system, since it confines the system to the
limitations of the original design of the VM interface. For example, if an
IVR system provides voice responses to an airline for crew scheduling, it
is unlikely that a IVR system could understand an employee number,
translate it to an extension, look up the caller's supervisor and
automatically transfer or drop the message in the supervisor's mailbox
without querying the caller. The VM interface is usually inadequate to
perform such complex tasking in an economical fashion. Thus, a fixed VM
system quickly dominates the more flexible IVR system when the two systems
attempt to operate together and the necessary VM caller interface is
introduced in a pseudo-integrated environment. Such pseudo-integration
schemes to put different voice processing applications together have
heretofore simply not been able to accomplish the multifarious complex
voice transactions required. Prior integrated systems do not solve the
long-felt need in the art for a truly universal integrated voice
processing telecommunications system.
During the evolution of the voice processing industry, VM systems have not
been customized to perform according to a particular customer's unique
specifications. Thus, VM-type systems were developed in mostly hard-coded
traditional programming languages such as the C-language or Pascal
language. In contrast, IVR systems were generally more sophisticated and
employed primitive customization for particular applications. The IVR
systems were thus generally designed in higher level programming language
known as "scripted languages." Scripted languages merely replace the
C-language or Pascal knowledge requirements of the system developer with
that of the Basic language.
The common problem which emerges with the use of scripted languages is a
disorientation of the system developer when designing the flow of the
particular application. Furthermore, most scripted languages require
several dozens of pages of basic code to accomplish even a simple
programming task. Even though scripted programs can be interpreted by a
programmer having less expertise than that which would be required if the
software programs were written in the more traditional C-language or
Pascal language, it will be recognized by those with skill in the art that
after even a few pages of the lengthy scripted code have been reviewed,
the entire flow of the application becomes disjoint and escapes the normal
comprehension of even the most expert programmers in scripted languages.
In order to devise ways of alleviating the problems extant in scripted
software voice processing systems, the concept of a state, event and
action to define applications having programming methodologies in, for
example, C-language or Pascal have been developed. Example of such systems
are disclosed in U.S. Pat. No. 4,747,127 to Hansen et al. The Hansen et
al. patent describes methods of implementing states, events, and actions
to control a real-time telecommunications switching system. The methods of
performing voice processing transactions in the Hansen et al. patent are
accomplished using a scripted base language similar to the "SHELL"
programming language used by the AT&T UNIX System V operating system. See
Hansen et al., col. 7, lines 15-35.
The methods and systems described in the Hansen et al. patent are strictly
limited to telecommunications switches on a PBX. While the implementation
of states, events and actions to perform higher level voice transactions
is desirable, the systems and methods disclosed in the Hansen et al.
patent do not fulfill the long-felt need in the art for integrated voice
processing systems adaptable to provide multiple functions in a single,
general-purpose computer environment and for varying customized
applications. Furthermore, the use of non-traditional script base
high-order programming language severely limits the adaptability of
systems taught in the Hansen et al. patent, and thus the systems and
methods disclosed in the Hansen et al. patent cannot be manipulated to
provide integrated voice processing transactions.
The aforementioned plethora of voice processing systems are generally
restricted to discrete sub-technology areas and accomplish narrow tasks of
specific voice transaction functions. The above systems are at best only
partially adapted to be customized for particular user applications and
are not practically integrated to provide multiple voice processing
transactions in a common computer platform. The aforementioned long-felt
needs in the art have therefore not been fulfilled by any of the voice
processing systems | | |
