JP2003337897A - Auction reputation system - Google Patents

Abstract

(57) [Summary] [Problem] In the prior art, reputation values are simply added, and there are cases where the reputation values do not match the latest trust data, and the evaluation value of a specific member is intentionally raised or lowered. It cannot be prevented even if a personal attack or group operation occurs. SOLUTION: In an auction system which can be connected to a plurality of user terminals via a network, accepts sales or bids from the user terminals, and provides an auction service to users, a buyer is a seller based on transaction results. It is possible to make a purchase decision by referring to the reputation data, and after the transaction, the system side determines whether to add the seller and the seller's evaluation value to the evaluation data based on the reputation permission data. It has a reputation function that controls the evaluation value in consideration of the transaction results and the time data. Based on the strength of the interaction between the users obtained from the post-transaction evaluation value and the number of transactions, the system performs the selection of the reputation.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a technique for providing a measure for fostering reliability among participants in online transactions.

[0002]

2. Description of the Related Art In online transactions, it is common for participants to exchange information while remaining anonymous and establish a transaction, but both the seller and the buyer gain a sense of trust in each other. It is indispensable for the trading system to have the means to do so. Conventionally, a method of creating an electronic feedback database regarding the integrity and reputation of potential trading partners among the members who participated in the transaction, checking the data during the transaction, and checking the reputation of other evaluators to confirm the reliability. Has already been put to practical use. There, as the evaluation value of the trust, the reputation data after the transaction, for example, "credible" is +1, "normal" is 0, "untrustworthy" is -1, simple addition is made for each transaction, and the average. Practical use of values as reputation data. In addition, at that time, not only the score but also the comments are combined to be used as reputation data, but since there is a danger that negative comments will worsen the evaluation of others, it is always necessary to use negative comments. Defensive measures have been taken to comment by setting a specific transaction number.

[0003]

In the conventional method, the reputation value is simply added, and when the transaction history is large, it does not always indicate the latest situation. Even if a person who had a high reputation recently got a bad reputation, the influence is not immediately reflected in the average value,
It is expected that there will be cases where the latest trust data is not met. Also, members who can enter reputation values and comments are unconditionally allowed as long as the member who completed the transaction,
Depending on the qualifications and conditions of the rating side, there are cases that are not suitable for the rating. For example, even if a personal attack or a group operation that intentionally raises or lowers the evaluation value of a specific member occurs, it cannot be prevented.

[0004]

The present invention has been made to solve the above-mentioned problems. In the reputation system of the present invention, each person participating in the trading market participates in the trading market. It is a system that evaluates the evaluation of other participating participants by themselves, the purchaser refers to the seller's reputation data based on the transaction record and decides to purchase, and after the transaction, the evaluation value of the seller and the seller is evaluated data. It is characterized in that the system side makes a decision based on the reputation permission data.
When the purchaser makes an evaluation based on the reputation value of the seller, the evaluation value (and evaluation comment) is a criterion. Previously, the evaluation value was given by the simple average value of all past data,
Based on the time information, an average value based only on the recent transaction record makes it possible to obtain data showing the latest reputation status. In addition, an operation that raises the reputation value only by dealing with a specific partner can be considered. In that case,
Based on the transaction record history of the seller and the buyer, the latest reputation value obtained above is weighted to reduce the influence of intentional reputation value operation.

[0005]

BEST MODE FOR CARRYING OUT THE INVENTION FIG. 1 shows the flow of the entire auction realized by the present invention. When the auction starts, the target product is first presented to all prospective buyers of the auction (1). The prospective purchaser who is interested in the presented product checks the reputation of the seller who is willing to sell the product in (2). The reputation value displayed at that time is the reputation DB (1 retrieved based on the seller's ID).
00), and if the result shows that the reputation value is OK (3), it purchases it, and if NO, it jumps to another product. In step 4, after the purchase, payment, and delivery, the buyer gives the seller a rating, and the seller gives a rating to the buyer when the buyer gives his rating. Things come out. At that time, on the reputation system side, whether the seller or the purchaser is qualified to give a reputation is evaluated by the reputation permission DB (20
The data of 0) is searched and judged in step 5. If YES, each result will be added to the new evaluation DB as an evaluation value (10). Then proceed to the auction for the next product. If the determination in step 5 is NO, the evaluation value is not added to the reputation DB, and the process proceeds to the auction for the next product. An algorithm is also conceivable in which the auction permission is given to the reputation permission DB at the same time as the reputation DB and the reputation value is also reflected when the auction is conducted.

FIG. 2 shows an algorithm for giving a reliable evaluation value as much as possible when giving the purchaser the reputation value of the seller described in FIG. Past transaction data is stored in the transaction history data (101). Here, a purchaser ID, a seller ID, a product ID, and time information related to the current transaction are input as input data. The database (101) extracts the transaction history of the purchaser and the seller and the period of time based on the above four data.
On the time series based on the time information, at 103, only the young reputation value is selected and the average thereof is calculated. Further, the depth of the relationship between the purchaser and the seller is estimated from the transaction data, and the weighting factor for the reputation value itself is calculated (102). This coefficient is 103
By multiplying the evaluation value that has been simply added with, it is possible to prevent the intentional operation of the evaluation value by intentionally raising the mutual reputation value by conducting transactions only with specific parties.

FIG. 3 illustrates an algorithm for permitting new addition of the reputation value described in FIG. Transaction history data (101) and weighting factor calculation (10) of FIG.
2) is applied as it is. Weighting factor calculation result (10
2) judges the randomness (202) and inputs the purchaser ID and the seller ID in the reputation permission data (200), and 2
When it is confirmed that the reputation permission is given from the past relation at 01, the final determination of the reputation permission is made at 203 and a new reputation value is processed. If the randomness is significantly impaired in the weighting factor calculation, and if it is found that the evaluation is prohibited in the past, the evaluation process is prohibited as in 204, and the purchaser as in 205 is If the seller's ID is verified and it is a new combination,
Evaluation permission data indicating that the combination is prohibited (200)
Write in.

The following method can be considered as an example of calculating the weighting coefficient. When the expected value of the number of auction participants per product is N, the probability that a member wins a product per transaction is 1 / N. If it is completely independent and has completed M transactions with a random partner, its normal probability is (1 / N) ^M. If the transaction partner is intentionally decided, the transaction is always with the same probability with a normal probability of 1. Therefore, it can be estimated to some extent according to how much the transaction occurrence rate with a specific counterparty is (1 / N) ^M.

As described above, the inventions made by the present inventor are
Although the specific description has been given based on the above-described embodiment, the present invention is not limited to the above-described embodiment, and needless to say, various modifications can be made without departing from the scope of the invention.

[0010]

According to the present invention, based on the correct reputation value,
Build a highly reliable market. Moreover, it is possible to prevent an arbitrary evaluation operation and conduct a healthy auction.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of the overall flow of an auction considering a reputation system according to the present invention.

FIG. 2 is a diagram showing an example of an algorithm for notifying a seller's reputation value according to the present invention.

FIG. 3 is a diagram showing an example of an algorithm for permitting participation in a reputation according to the present invention.

[Explanation of symbols]

1-6, 10 ... Each step of an auction flow considering a reputation system, 100 ... Reputation DB, 101 ... Transaction history data, 102, 103 ... Each step of algorithm for notifying a seller's reputation value, 200 ... Reputation permission DB,
201-205 ... Each step of the algorithm for permitting participation in reputation.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Hiroyuki Takahashi             2-3-1, Otemachi, Chiyoda-ku, Tokyo             Inside Telegraph and Telephone Corporation (72) Inventor Toshio Yamagishi             Hokkaido, Kita-ku, Sapporo, Kita 10 West 7

Claims (3)

[Claims] 1. In an auction system which can be connected to a plurality of user terminals via a network, accepts sales or bids from the user terminals, and provides an auction service to users, a buyer is a seller based on a transaction record. The auction system is characterized in that the system side determines, based on the reputation permission data, that the seller and the seller's evaluation value should be added to the evaluation data after making a purchase decision by referring to the reputation data of. 2. The auction system according to claim 1, further comprising a reputation function of controlling an evaluation value in consideration of transaction record and time data. 3. The auction system according to claim 1, wherein the system selects a reputable person based on the strength of mutual relation between the users obtained from the evaluation value after the transaction and the number of transactions. An auction system featuring.