JPH03144793A - Order data management system - Google Patents

Abstract

PURPOSE:To attain the high efficiency and labor saving of commodity order receiving work by recognizing a keyword such as a customer name, a commodity name and the number of units from voice contents in conversation between an operator and the customer, and based upon the recognized result directly accessing a customer management data base and updating the contents of the data base. CONSTITUTION:At the time receiving a customer's call, the operator hooks off a receiver to answer the call. In this case, the customer's name voiced by the operator is sent to a customer data base management part 6 as the recognition result of a voice collating part 5. When the operator repeats a commodity order in accordance with the commodity order from the customer, its contents are recognized by the collating part 5 again and sent and displayed to/on the management part 6. When the operator confirms the displayed contents, the confirmed contents are written in the data base as a new order data. Thus, the new order receiving data are entered in the customer data base at the end of speaking with the customer.

Description

[Detailed Description of the Invention] [Industrial Field of Application] This invention enables, when a customer calls to order a product, to identify the product type, product name, etc. from the audio of the store operator talking to the customer. By recognizing pre-registered keywords such as the quantity of products, and using the results to access the customer management database or update the contents of the database, the operator can receive product orders while having a conversation with the customer. The present invention relates to an order data management system that allows data to be input into a customer management database, thereby streamlining and labor-saving order processing at stores and the like.

[Conventional technology]

In small-scale retail stores such as liquor stores and grocery stores, and large-scale retail stores such as supermarkets and department stores, in addition to selling products face-to-face with customers, they also use a telephone ordering system in which they accept product orders from customers over the phone. Sales account for a large proportion. Therefore, streamlining and saving labor in these operations is extremely important for improving business performance.

Recently, customer databases are prepared on PCs and office computers, and when an order is received, the database is accessed and its contents are updated to automatically create order slips, sales slips, etc., and to track sales. Management has become popular.

The introduction of these new technologies has automated some of the work that was previously done entirely by hand, resulting in greater efficiency and labor savings. However, inputting order data into a database remains a manual process. In other words, in these operations, the operator first listens to the customer's order over the phone and creates a memo; after the telephone call is over, the operator accesses the database on a computer or office computer using the keyboard while looking at the memo;
You will go through a process of updating the contents.

[Problem to be solved by the invention]

Such a process has the following problems.

■It takes time because it is a two-step process.

■Errors are likely to occur as you type while looking at your notes.

■Entering data using a keyboard is not good for most Japanese people, so it is time-consuming and prone to errors.

Therefore, there is an urgent need to automate the process of inputting such telephone order data into a database in order to improve the efficiency and save labor of the entire business.

An object of the present invention is to provide an order data management system that can automate the process of inputting order data into a database.

[Means to solve the problem]

In order to achieve the above object, the present invention includes a voice analysis unit that, when receiving a product order from a customer by voice and the order taker responds to it by voice, extracts characteristic parameters from the voice by manually recording the exchanged voice; Comparing a standard pattern storage unit that stores keywords to be recognized as standard patterns such as words and phrases in advance with the feature parameters extracted by the speech analysis unit and the standard patterns read from the standard pattern storage unit. If a keyword is included in the human voice, there is a voice matching unit that detects and outputs it, and when the keyword detected by the voice matching unit is input, it is displayed and confirmed to the customer. When a request is made and confirmation is made, the order data management system is comprised of a customer management database management section that accesses the customer management database and updates the data based on the keyword.

[Function] The system according to the present invention, when an operator on the product order receiving side of a store or the like is answering a telephone call from a customer, checks keywords related to product ordering from the voice of the operator, and confirms the keyword related to the product order. ! This is a system that accesses and updates the customer management database based on the results of the analysis. Therefore, the database is accessed and the contents updated while the operator is answering the customer's phone call, so the inputting of order data into the database is completed by the time the phone call ends. It is possible to solve the disadvantages of ``processing,'' which took 0 hours, and ``prone to make mistakes,'' thereby realizing true efficiency and labor-saving in product ordering operations.

〔Example〕

FIG. 1 is a block diagram showing the configuration of one embodiment of the present invention. In the figure, reference numeral 1 denotes a telephone that receives orders for products. Reference numeral 2 denotes a voice analysis section which analyzes the voice input from the telephone 1, converts it into voice characteristic parameters such as LPG parameters, and outputs the same. 3 is a changeover switch. During normal operation, it lies toward the voice verification section 5.

When the operator registers a keyword that he wants to be recognized in advance, it is placed in the standard pattern storage section 4.

The standard pattern storage unit 4 stores keywords to be recognized. For example, if this system is installed at a liquor store, ■ customer name, ■ product type such as "beer" or "sake," ■ " Keywords include product names such as ``O○ Beer'' and ``ΔΔ Whiskey j, ■ quantity such as ``1 bottle'' and ``2 bottles.''

The voice matching unit 5 continuously compares the operator's continuous voice input from the voice analysis unit 2 with the keyword voice patterns stored in the standard pattern storage unit 4, and identifies keywords from the operator's continuous voice. Has the ability to extract.

Existing technology can be used as a function to extract keywords from continuous speech.For example, one example of such a technology is a technology called continuous DP that performs continuous P matching between input speech and a standard pattern. . An example of a method for extracting syllables and words from continuous speech using this technology is given in the literature: Noboru Kanmura, Akihiro Tomihisa, Chohei Nakatsu, rev, "Continuous speech recognition system based on word spotting," IEICE Speech. Study Group, 5P88-96, PP.

1-8 (1988-12).

Reference numeral 6 denotes a customer database management unit, which stores a corresponding database (not shown) based on the recognition result of the voice verification unit 5.
It has the ability to access and display its contents and input new data.

Next, an example of the operation of this embodiment will be described. First, prior to use, the operator speaks the above keywords one by one. The contents are converted into characteristic parameters by the voice analysis section 2 and stored in the standard pattern storage section 4.

Next, during actual use, the changeover switch 3 is tilted toward the voice matching unit 5, and the input voice is converted into characteristic parameters and then input to the voice matching unit 5, where the standard pattern storage unit 4 Continuous comparison is performed with standard patterns stored in , and keywords in the continuous speech are extracted.

An example of operation is shown in FIG. 2. First, when a customer calls, the operator picks up the telephone receiver and answers the phone. At this time, the customer's name spoken by the operator is recognized by the voice verification unit 5 and is sent to the customer database management unit 6. sent to. The customer database management unit 6 displays the customer name on the display and waits for the operator's confirmation, and the operator confirms by pressing a key, etc. (For example, if the first place in the recognition result is correct, it will set 1, and if the second place or lower) If there is a correct result, enter the number from the keyboard), the customer database management section 6 accesses the database using the customer name and displays the contents of the customer database.

Next, when the operator repeats the recitation in response to the customer's order for a product, the contents are again recognized by the voice verification section 5 and sent to the customer database management section 6 for display. When the operator performs the TI approval, the contents are written into the database as new order data.

After going through such a process, new order data will have been entered in the customer management database at the end of the telephone call with the customer.

[Effect of the invention] As explained above, in the conventional product ordering business,
While answering calls from customers, the operator creates a memo for the ordered product, and then inputs order data by accessing the customer management database on a computer, office computer, etc. from the keyboard and updating the contents while looking at the memo. This has the problem of being time consuming and prone to errors.

In contrast, in the present invention, keywords such as customer name, product name, quantity, etc. are recognized from the audio content of the operator's dialogue with the customer, and the results are used to directly access the customer management database and update the contents of the database. Since the product order data is already entered into the customer management database when the customer's order telephone call is over, this technology has the advantage of significantly streamlining and labor-saving the product order process compared to conventional technology.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing flIrIi in one embodiment of the present invention, and FIG. 2 is an explanatory diagram showing an example of the operation of the embodiment shown in FIG. Explanation of symbols 1... Telephone, 2... Voice analysis section, 3... Changeover switch, 4... Standard pattern storage section, 5... Verification section,
6...Customer database management department

Claims (1)

[Claims] 1) When a product order is received from a customer by voice and the order taker responds to it by voice, a voice analysis unit receives the voice exchanged and extracts characteristic parameters from the voice, and a voice analysis unit extracts the keywords and phrases to be recognized. Keywords are included in the input voice by comparing the standard pattern storage unit that stores standard patterns such as the above in advance with the feature parameters extracted by the voice analysis unit and the standard patterns read from the standard pattern storage unit. If the keyword is detected by the voice matching unit, the keyword detected by the voice matching unit is input, the keyword is displayed, and confirmation is requested from the contractor. , and a customer management database management section that accesses the customer management database and updates data based on the keyword.