Dynamic hotel price management method based on artificial intelligence

Abstract

The invention relates to the technical field of interconnection big data and artificial intelligence, in particular to a hotel price dynamic management method based on artificial intelligence, which comprises the following steps: S1, obtaining a hotel reservation demand, an expected check-in day number and an initial bid of a user; s2, the AI model carries out hotel matching to obtain a target hotel and a corresponding target room type; s3, the AI model calculates the recommended price of the target house type; s4, the AI model calculates the recommended check-in days meeting the initial bid of the user; s5, the AI model sends the recommended price and the number of recommended check-in days meeting the initial bid of the user to the user; s6, when the user accepts the recommended price, the AI model automatically reserves the target room type of the target hotel for the user based on the recommended price and the expected check-in days; and S7, when the user accepts the recommended check-in days, the AI model automatically reserves the target room type of the target hotel for the user based on the initial price and the recommended check-in days. According to the invention, the flexibility of hotel room reservation bargaining can be improved and the management cost of hotels can be reduced.

Description

Technical Field

[0001] The present invention relates to the field of interconnected big data and artificial intelligence technologies, and particularly relates to a method for dynamic management of hotel prices based on artificial intelligence. Background Art

[0002] With the rapid development of Internet technology and the booming rise of the tourism industry, the room reservation bargaining platform, as a bridge connecting consumers and hotel merchants, has gradually become an important part of the online travel market. These platforms integrate a large number of hotel resources, provide users with diverse accommodation options, and introduce a bargaining mechanism, aiming to achieve a win-win situation between users and hotels through price negotiation.

[0003] The working mode of the room reservation bargaining platform generally includes: users input room reservation requirements on the platform, including information such as check-in date, required room type, budget range, and special preferences. The platform screens and displays a list of eligible hotels for users to compare and select according to the user's requirements. After the user selects a desired hotel, they can initiate a bargaining request and propose their expected price. After receiving the bargaining request, the hotel merchant replies through the platform's artificial customer service or back-end management system, accepting or rejecting the user's bargaining, or proposing a new price plan. After both parties reach an agreement on the price, the reservation process is completed, the user pays a deposit or full payment, and the hotel confirms the reservation.

[0004] Currently, AI (Artificial Intelligence) technology has been introduced into the hotel room reservation bargaining platform. The AI model can provide more accurate and personalized recommendation services for users by analyzing user behavior data, market trends, and hotel operation conditions. For example, the AI model can intelligently recommend hotels and room types that meet the user's budget and needs based on the user's past reservation records, preferences, and current market supply and demand situations, and can even predict future price trends to help users make more informed reservation decisions. The AI model can also replace hotel staff to automatically process a large number of bargaining requests, improve the response speed, and reduce labor costs.

[0005] Although the introduction of AI models has brought many conveniences to hotel reservation bargaining platforms, there are still obvious shortcomings in the bargaining flexibility of current platforms. For example, when the initial bid proposed by a user is lower than the current hotel price, existing AI models often can only simply inform the user that the price is non-negotiable or suggest that the user choose another hotel, and cannot provide more flexible alternative solutions for the user. This rigid bargaining mode not only limits the user's bargaining space but also reduces the probability of the user successfully booking a satisfactory room, thus affecting the user's overall reservation experience. The applicant has found that sometimes the expected length of stay entered by the user is just an expected value, and the user is often willing to adjust the length of stay. Moreover, the change in the length of stay will directly bring about a change in price because for the hotel side, the longer the user stays, the lower the management costs required (such as the costs of cleaning the room, changing beddings and daily necessities). Therefore, how to design a method for dynamically adjusting hotel prices based on the user's length of stay is a technical problem that urgently needs to be solved. Summary of the Invention

[0006] Aiming at the deficiencies of the above-mentioned existing technologies, the technical problem to be solved by the present invention is: how to provide an AI-based dynamic hotel price management method that can calculate the recommended length of stay that meets the user's initial bid through real-time big data and hotel historical data, and provide a bargaining coordination strategy for the user to increase the length of stay to reduce the hotel price, thereby improving the flexibility of hotel reservation bargaining and reducing the management costs of the hotel.

[0007] To solve the above technical problems, the present invention adopts the following technical solutions:

[0008] An AI-based dynamic hotel price management method, comprising:

[0009] S1: Obtain the user's hotel reservation requirements, expected length of stay, and initial bid;

[0010] S2: The AI model matches hotels according to the user's hotel reservation requirements and expected length of stay to obtain the target hotel and the corresponding target room type;

[0011] S3: The AI model calculates the recommended price of the target room type according to real-time big data and the historical data of the target hotel;

[0012] S4: When the user's initial bid is lower than the recommended price, the AI model calculates the recommended length of stay that meets the user's initial bid based on the initial price, combined with real-time big data and the historical data of the target hotel;

[0013] S5: The AI model sends the recommended price and the recommended length of stay that meets the user's initial bid to the user, and obtains the feedback information on whether the user accepts the recommended price or the recommended length of stay;

[0014] S6: When the user accepts the recommended price, the AI model automatically books the target room type of the target hotel for the user based on the recommended price and the expected length of stay.

[0015] S7: When the user accepts the recommended length of stay, the AI model automatically books the target room type of the target hotel for the user based on the initial price and the recommended length of stay.

[0016] Preferably, in step S2, the AI model performs hotel matching based on the user location, booking date, room type requirement, and number of occupants in the user's hotel booking requirements in combination with the expected length of stay to obtain the target hotel that meets the user's hotel booking requirements and the corresponding target room type.

[0017] Preferably, in step S2, the AI model performs hotel matching through the following steps:

[0018] S201: Obtain the user location, booking date, room type requirement, and number of occupants in the user's hotel booking requirements.

[0019] S202: Calculate the expected check-in period based on the booking date and the expected length of stay, and use the hotels with remaining rooms during the expected check-in period as candidate hotels.

[0020] S203: Calculate the straight-line distance d between the user location and the candidate hotels, and calculate the distance attenuation factor δ of the candidate hotels in combination with the set attenuation coefficient k d =e -k·d ;

[0021] S204: Match the room type requirement with the existing room types of the candidate hotels to obtain the room type matching degree m of the candidate hotels: if the room type requirement is the same as the existing room types of the candidate hotels, then the room type matching degree m = 1; otherwise, m = 0.

[0022] S205: Calculate the occupancy suitability score of the candidate hotels based on the number of occupants in combination with the maximum capacity of a single room in the candidate hotels

[0023] S206: Weight the distance attenuation factor, room type matching degree, and occupancy suitability score of the candidate hotels to obtain the matching score S of the candidate hotels.

[0024] The calculation formula for the matching score is expressed as:

[0025] S = ω1·δ n +ω2·m+ω3·s n ;

[0026] In the formula: ω1, ω2, ω3 represent the set dynamic weights.

[0027] S207: Use the candidate hotels with matching scores higher than the score threshold as the target hotels, and use the room types in the target hotels that meet the user's needs as the target room types.

[0028] Preferably, in step S206, when the AI model calculates the matching score S of the candidate hotel, a user preference coefficient is introduced for calculation;

[0029] The calculation formula for the matching score is updated to:

[0030] S = (ω1·δ n + ω2·m + ω3·s n )·η;

[0031] In the formula: η represents the user preference coefficient, which is calculated based on the user's preferences and the facilities and services of the candidate hotel;

[0032] The user's preferences include any one or more of free network, free breakfast, free gym, free swimming pool, near the city center, near the subway station, and having high-quality restaurants;

[0033] When the facilities and services of the candidate hotel meet all the user's preferences, the user preference coefficient η = 1; when the number of items of the facilities and services of the candidate hotel that do not meet the user's preferences does not exceed one, the user preference coefficient η = 0.9; when the number of items of the facilities and services of the candidate hotel that do not meet the user's preferences is greater than or equal to two, the user preference coefficient η = 0.8.

[0034] Preferably, in step S3, the AI model calculates the recommended price through the following formula:

[0035] P t = (a·P c + b·P j )·θ h ·δ t ;

[0036] In the formula: P t represents the recommended price of the target room type; P c represents the historical lowest price of the target hotel for the target room type; P j represents the average price of the same type of room of the same type of hotel of the target hotel for the target room type; a, b represent the set weights; θ h represents the holiday premium factor, which is related to whether the user's booking date is a weekend or a holiday: when the user's booking date is a weekend, the holiday premium factor θ h = 1.2, when the user's booking date is a holiday, the holiday premium factor θ h = 1.5, when the user's booking date is not a weekend and a holiday, the holiday premium factor θ h = 1; δ t represents the supply and demand factor of the time period;

[0037]

[0038] Wherein: λ represents the time period sensitivity coefficient, which is associated with the time period;

[0039] One day is set to three time periods:

[0040] Time period 1: from 0:00 to 8:00, the time period sensitivity coefficient λ = 0.1;

[0041] Time period 2: from 8:00 to 16:00, the time period sensitivity coefficient λ = 0.5;

[0042] Time period 3: from 16:00 to 24:00, the time period sensitivity coefficient λ = 0.3.

[0043] Preferably, in step S4, the AI model calculates the recommended length of stay through the following steps:

[0044] S401: Obtain the historical transaction average price P h (w, d) and the corresponding average occupancy rate of different lengths of stay of the target room type at different times from the historical data of the target hotel, where w represents the time period and d represents the length of stay;

[0045] S402: Based on real-time big data, obtain the reference average price P c of the room type of the same type as the target room type in the same type of hotels as the target hotel at the current time period w m (w c ) and the price fluctuation index I f (w c ), and the price fluctuation index refers to the change range of the reference average price at the current time period w c relative to the previous time period; calculate the price adjustment coefficient K(w f (w c ) of the current time period w c ) = 1 + I c (w f (w c );

[0046] S403: Obtain the historical transaction average price P c of different lengths of stay at the current time period w from the historical data of the target hotel h (w c , d) and the corresponding average occupancy rate O h (w c , d), and calculate the historical average price P c (w a (w c ) of the current time period w;

[0047] The formula is expressed as:

[0048]

[0049] In the formula: D max represents the maximum occupancy days;

[0050] S404: Calculate the corresponding price benchmark value P c based on the historical average price P a (w c ) and the reference average price P m (w c ); b ;

[0051] The formula is expressed as:

[0052]

[0053] S405: Calculate the upper limit P c and the lower limit P c of the price reference range based on the price adjustment coefficient K(w b ) and the price benchmark value P max ; min ;

[0054] The formula is expressed as:

[0055] P max = P b ·K(w);

[0056] P min = P b / K(w);

[0057] S406: Select the target occupancy days d c that meet P h (w c , d) from the historical transaction average prices P max < P h (w c , d m ) < P min ; m ;

[0058] S407: Calculate the matching degree M(d i ) for each target occupancy days d m ; Combine with the average occupancy rate O m of the current period w c to calculate the average occupancy rate O[[ID=9,2]] h (w c , d) for each target occupancy days d mComprehensive evaluation value V(d m );

[0059] It is expressed by the formula as:

[0060]

[0061] V(d m ) = M(d m )·O h (w c , d m );

[0062] S408: Select the target check-in days d m with the highest comprehensive evaluation value V(d m as the recommended check-in days.

[0063] Preferably, in step S408, when there are multiple target check-in days with the highest comprehensive evaluation value V(d m ), the AI model selects the smallest target check-in days as the recommended check-in days.

[0064] Preferably, in step S406, after screening out the target check-in days d m , the AI model calculates the corresponding target check-in dates for each target check-in days d m based on the user's booking date; obtains the booked dates of the target room type in the target hotel, and determines whether the booked date of the target room type conflicts with the target check-in date corresponding to the target check-in days d m , and deletes the target check-in days d m corresponding to the conflicting target check-in dates.

[0065] Compared with the prior art, the hotel price dynamic management method based on artificial intelligence in the present invention has the following beneficial effects:

[0066] By analyzing the user's hotel reservation requirements, expected check-in days and initial bid through the AI model, and combining real-time big data and historical data of the target hotel, the present invention provides an innovative bargaining coordination scheme for users, that is, increasing the check-in days to reduce the room unit price, so that users can adjust the check-in days to match the hotel's quote under the limited budget, thereby greatly improving the flexibility and success rate of hotel reservation bargaining. When the user's initial bid is lower than the hotel's recommended price, the minimum check-in days that meet the user's budget can be calculated, enabling the user to have the opportunity to book the desired room at a more favorable price. The bargaining method not only improves the user's bargaining ability, but also enhances the user's reservation experience and satisfaction.

[0067] The bargaining coordination strategy of the present invention to reduce hotel prices by increasing the length of stay can provide the hotel with a longer length of stay. The longer the user stays, the lower the hotel management costs (such as the costs of cleaning the room, changing beddings and daily necessities). At the same time, the longer length of stay increases the long-term occupancy rate of hotel rooms, helps the hotel better plan room resources, reduces the vacancy rate, and can also reduce the hotel's operating costs, such as daily expenses for cleaning, maintenance, etc., thereby reducing the hotel's management costs.

[0068] The AI model predicts and recommends prices and recommended lengths of stay based on real-time big data and the hotel's historical data. On the one hand, by deeply mining and analyzing big data as well as the price strategies and market dynamics of competitors, the future price trend of hotel rooms can be predicted, and more reasonable and comprehensive recommended prices and recommended lengths of stay can be generated for users. On the other hand, by combining the hotel's historical data to calculate recommended prices and recommended lengths of stay, the hotel's costs and pricing strategies can be considered, and prices and lengths of stay that both users and the hotel can accept can be generated, thereby increasing the success rate of booking bargaining matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to make the objectives, technical solutions and advantages of the invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings, where:

[0070] Figure 1 It is a logic block diagram of a hotel price dynamic management method based on artificial intelligence. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] In order to make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but only represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0072] The following is a further detailed description through specific implementation manners:

[0073] Embodiment 1:

[0074] In this embodiment, a hotel price dynamic management method based on artificial intelligence is disclosed.

[0075] As Figure 1As shown in the figure, an AI-based dynamic hotel price management method includes:

[0076] S1: Obtain the hotel reservation requirements, expected length of stay, and initial bid of the user;

[0077] S2: The AI model matches hotels according to the user's hotel reservation requirements and expected length of stay to obtain the target hotel and the corresponding target room type;

[0078] S3: The AI model calculates the recommended price of the target room type based on real-time big data and the historical data of the target hotel;

[0079] S4: When the user's initial bid is lower than the recommended price, the AI model calculates the recommended length of stay that meets the user's initial bid based on the initial price, combined with real-time big data and the historical data of the target hotel; if the user's initial bid is higher than or equal to the recommended price, the AI model automatically reserves the target room type of the target hotel for the user;

[0080] S5: The AI model sends the recommended price and the recommended length of stay that meets the user's initial bid to the user (user terminal), and obtains the feedback information on whether the user accepts the recommended price or the recommended length of stay;

[0081] S6: When the user accepts the recommended price, the AI model automatically reserves the target room type of the target hotel for the user based on the recommended price and the expected length of stay;

[0082] S7: When the user accepts the recommended length of stay, the AI model automatically reserves the target room type of the target hotel for the user based on the initial price and the recommended length of stay.

[0083] Through the analysis of the user's hotel reservation requirements, expected length of stay, and initial bid by the AI model in the present invention, combined with real-time big data and the historical data of the target hotel, an innovative bargaining coordination solution is provided for the user, that is, increasing the length of stay to reduce the room unit price, so that the user can adjust the length of stay to match the hotel's quotation under the limited budget, thus greatly improving the flexibility and success rate of hotel room reservation bargaining. When the user's initial bid is lower than the hotel's recommended price, the minimum length of stay that meets the user's budget can be calculated, enabling the user to have the opportunity to book the desired room at a more favorable price. The bargaining method not only improves the user's bargaining ability but also enhances the user's reservation experience and satisfaction.

[0084] The bargaining coordination strategy of the present invention to reduce hotel prices by increasing the length of stay can provide a longer length of stay for the hotel. The longer the user stays, the lower the hotel management costs (such as the costs of cleaning rooms, changing beddings and daily necessities). At the same time, the longer length of stay increases the long-term occupancy rate of hotel rooms, helps the hotel better plan room resources, reduces the vacancy rate, and can also reduce the hotel's operating costs, such as daily expenses for cleaning, maintenance, etc., thereby reducing the hotel's management costs.

[0085] The AI model predicts and recommends prices and recommended lengths of stay based on real-time big data and the hotel's historical data. On the one hand, through in-depth mining and analysis of big data, as well as the price strategies and market dynamics of competitors, it can predict the future price trends of hotel rooms and generate more reasonable and comprehensive recommended prices and recommended lengths of stay for users; on the other hand, by combining the hotel's historical data to calculate recommended prices and recommended lengths of stay, it can consider the hotel's costs and pricing strategies, generate prices and lengths of stay that both users and hotels can accept, thereby increasing the success rate of booking bargaining matching.

[0086] To better introduce the technical solution of the present invention, this embodiment is described through the following several parts.

[0087] I. Hotel matching [[ID=*]] [[ID=*]]

[0088] In this embodiment, the AI model performs hotel matching based on the user's location (obtained through an authorized mobile device), booking date, room type requirements, and number of occupants in the hotel reservation requirements (input by the user) combined with the expected length of stay, and obtains the target hotel and the corresponding target room type that meet the user's hotel reservation requirements.

[0089] Specifically, the AI model performs hotel matching through the following steps:

[0090] S201: Obtain the user's location, booking date, room type requirements, and number of occupants in the hotel reservation requirements;

[0091] S202: Calculate the expected occupancy period based on the booking date and the expected length of stay, and use the hotels with remaining rooms during the expected occupancy period as candidate hotels;

[0092] S203: Calculate the straight-line distance d between the user's location and the candidate hotels, and calculate the distance attenuation factor δ of the candidate hotels in combination with the set attenuation coefficient k d = e -k·d ;

[0093] S204: Match the room type requirements with the existing room types of the candidate hotels to obtain the room type matching degree m of the candidate hotels: If the room type requirements are consistent with the existing room types of the candidate hotels, then the room type matching degree m = 1; otherwise, m = 0. For example, if the user's room type requirement is a double room, but a certain hotel only has standard rooms, then the room type matching degree m of this hotel is 0.

[0094] S205: Calculate the occupancy suitability score of the candidate hotels based on the number of occupants and the maximum capacity of a single room in the candidate hotels Among them, the maximum capacity of a single room in the candidate hotels can be dynamically adjusted according to the hotel policy. For example, if a certain hotel allows free extra beds for children under 6 years old, then when there are children under 6 years old in the number of occupants of the user, the maximum capacity of the rooms in this hotel can be increased by one.

[0095] S206: Weight the distance decay factor, room type matching degree, and occupancy suitability score of the candidate hotels to obtain the matching score S of the candidate hotels;

[0096] The calculation formula for the matching score is expressed as:

[0097] S = ω1·δ n + ω2·m + ω3·s n ;

[0098] In the formula: ω1, ω2, ω3 represent the set dynamic weights, which can be obtained through training with the user's historical behavior data.

[0099] S207: Use the candidate hotels with matching scores higher than the score threshold (set to 0.8) as the target hotels, and use the room types in the target hotels that meet the user's needs as the target room types.

[0100] The present invention performs hotel matching based on the user's location, booking date, room type requirements, number of occupants, and the user's expected length of stay. By comprehensively analyzing multi-dimensional requirements, it can quickly screen out the hotels and rooms that best meet the user's requirements, greatly shortening the user's search time, improving the booking efficiency. At the same time, through precise matching of multiple conditions, it reduces information overload and ineffective browsing, enabling the user to find the required housing sources faster, while also improving the matching degree between hotel housing sources and user needs, and promoting the conclusion of transactions.

[0101] Specifically, when the AI model calculates the matching score S of the candidate hotels, a user preference coefficient is introduced for calculation;

[0102] The calculation formula for the matching score is updated to:

[0103] S = (ω1·δ n + ω2·m + ω3·s n )·η;

[0104] Where: η represents the user preference coefficient, which is calculated based on user preferences and the facilities and services of candidate hotels.

[0105] User preferences include any one or more of free internet, free breakfast, free gym, free swimming pool, near the city center (distance from the city center less than 5KM), near the subway station (distance from the city center less than 500m), and having high-quality restaurants (restaurants with a Meituan or Dianping score over 4.8 within 1KM).

[0106] When the facilities and services of the candidate hotel meet all user preferences (for example, if the user preferences include five items and the facilities and services of the candidate hotel meet all five), the user preference coefficient η = 1; when the number of items that the facilities and services of the candidate hotel do not meet the user preferences does not exceed one (for example, if the user preferences include five items and the facilities and services of the candidate hotel meet four of them), the user preference coefficient η = 0.9; when the number of items that the facilities and services of the candidate hotel do not meet the user preferences is greater than or equal to two (for example, if the user preferences include five items and the facilities and services of the candidate hotel meet three or two or one of them), the user preference coefficient η = 0.8.

[0107] In calculating the matching score of the candidate hotel, the present invention introduces the user preference coefficient. By considering the weights of user preferences (such as free internet, free breakfast, fitness facilities, etc.), it can more accurately recommend hotels that meet the personalized needs of users. Personalized matching not only reduces the time cost for users to screen information but also increases the possibility for users to find the ideal accommodation, thereby assisting in improving the booking success rate and user loyalty.

[0108] II. Recommended price calculation

[0109] In this embodiment, the AI model calculates the recommended price through the following formula:

[0110] P t =(a·P c +b·P j )·θ h ·δ t ;

[0111] Where: P t represents the recommended price of the target room type; P c represents the historical lowest price of the target hotel for the target room type; P j represents the average price of the same type of room of the same type of hotel as the target hotel for the target room type; a, b represent the set weights; θ h represents the holiday premium factor, which is related to whether the user's booking date is a weekend or a holiday: when the user's booking date is a weekend, the holiday premium factor θ h=1.2, when the user's booking date is a holiday (legal holidays: New Year's Day, May Day, Dragon Boat Festival), the holiday premium factor θ h =1.5, when the user's booking date is not a weekend or holiday, the holiday premium factor θ h =1;δ t represents the supply and demand factors for the time period;

[0112]

[0113] Where: λ represents the time period sensitivity coefficient, which is associated with the time period;

[0114] Divide the day into three time periods:

[0115] Time period 1: 0:00 to 8:00, time period sensitivity coefficient λ = 0.1;

[0116] Time period 2: 8:00 to 16:00, time period sensitivity coefficient λ = 0.5;

[0117] Time period three: 16:00 to 24:00, time period sensitivity coefficient λ = 0.3.

[0118] By incorporating a holiday premium mechanism, the present invention can dynamically adjust prices to reflect the surge in demand during special periods, ensuring maximum hotel revenue while reasonably guiding user expectations. At the same time, combined with time period supply and demand analysis (such as price increases during the morning booking peak and price reductions during the nighttime trough), the recommended price can respond to market fluctuations more flexibly, which not only encourages users to book during off-peak hours to obtain more favorable prices, but also helps hotels balance occupancy rates in different time periods and optimize resource allocation, thereby slightly enhancing the market competitiveness of the system and promoting a win-win situation for users and hotels.

[0119] 3. Calculation of recommended length of stay

[0120] In this embodiment, the AI model calculates the recommended length of stay through the following steps:

[0121] S401: Obtain the historical average transaction price P of the target room type at different times and lengths of stay from the historical data of the target hotel. h (w, d) (average of the prices in previous years during this period) and the corresponding average occupancy rate O h (w,d) (average of the occupancy rates in the previous years for the time period), where w represents the time period (weeks) and d represents the number of days of stay;

[0122] In this embodiment, a year is divided into 52 weeks (the extra day or two days are included in the last week), and each week corresponds to a time period. max God, D max Can be set to 30.

[0123] S402: Obtain the reference average price P of the room type of the same type as the target room type in the same type of hotels of the target hotel during the current period w c (w m (w c ) and the price fluctuation index I f (w c ). The price fluctuation index refers to the change range of the reference average price during the current period w c relative to the previous period; Calculate the price adjustment coefficient K(w f (w c ) of the current period w c ) according to the price fluctuation index I c ) 1 + I f (w c );

[0124] In this embodiment, the same type of hotels refers to the hotels that are in the same area as the target hotel, have the same star rating, and the average price difference of the rooms does not exceed 20%.

[0125] S403: Obtain the historical transaction average price P c of different occupancy days during the current period w h (w c , d) and the corresponding average occupancy rate O h (w c , d) from the historical data of the target hotel, and calculate the historical average price P c of the current period w a (w c );

[0126] The formula is expressed as:

[0127]

[0128] In the formula: D max represents the maximum occupancy days;

[0129] S404: Calculate the corresponding price benchmark value P c according to the historical average price P a (w c ) and the reference average price P m (w c ) of the current period w b ;

[0130] The formula is expressed as:

[0131]

[0132] S405: According to the price adjustment coefficient K(w c ) of the current period wc ) and the price reference value P b Calculate the upper limit P of the price reference range max and the lower limit P min ;

[0133] The formula is expressed as:

[0134] P max = P b ·K(w);

[0135] P min = P b / K(w); If K(w) ≤ 1, P min can be set to P b ·(2 - K(w)) to avoid too low a lower limit;

[0136] S406: From the historical transaction average price P c for different occupancy days in the current period w h (w c , d) select the target occupancy days d max that satisfy P h (w c , d m ) < P min ; m ;

[0137] S407: Calculate the matching degree M(d i ) for each target occupancy day d m based on the user's initial bid P m ; Combine the average occupancy rate O c (w h , d) of the current period w c to calculate the comprehensive evaluation value V)d m ) for each target occupancy day d m ;

[0138] The formula is expressed as:

[0139]

[0140] V(d m ) = M(d m ) · O h (w c , d m );

[0141] S408: Select the target occupancy day d m with the highest comprehensive evaluation value V(d m ) as the recommended occupancy day.

[0142] Specifically, when there are multiple target occupancy days with the highest comprehensive evaluation value V(d m ), the AI model selects the smallest target occupancy day as the recommended occupancy day.

[0143] The present invention comprehensively considers the historical data and real-time big data of the target hotel, can accurately delimit the price reference range, and then calculates the comprehensive evaluation value by combining the price matching degree and the occupancy rate, so that the selected recommended occupancy days can not only fit the user's initial bid, but also fully consider the actual operation situation of the hotel, realizing the best match between the user's needs and the hotel's costs, thereby improving the transaction success rate of hotel bargaining and room reservation. At the same time, the present invention first screens the target occupancy days that meet the price range, and then sorts them according to the comprehensive evaluation value, and preferentially selects the plan with a smaller occupancy day, which helps the hotel to reasonably allocate room resources while meeting the user's price expectations, improve the room turnover rate, realize the efficient utilization of resources and the minimization of hotel operation costs; and fewer occupancy days can reduce the user's occupancy costs and improve the user's experience.

[0144] IV. Avoidance of Room Reservation Date Conflict

[0145] In this embodiment, after screening to obtain the target occupancy day d m , the AI model calculates the corresponding target occupancy date for each target occupancy day d m based on the user's room reservation date (for example: the room reservation date is April 20th, and the target occupancy day is 5 days, then the target occupancy date is from April 20th to April 24th); obtains the booked room dates of the target room type of the target hotel, and determines whether the booked room dates of the target room type conflict with the target occupancy date corresponding to the target occupancy day d m (for example: the user's target occupancy date is from April 20th to April 24th, and the booked room date of the target room type is April 22nd, then it is determined that there is a conflict), and deletes the target occupancy day d m corresponding to the conflicting target occupancy date.

[0146] When the present invention screens to obtain the target occupancy days, it considers the situation where the current user's occupancy time conflicts with the booked room dates of the target hotel, can avoid the problem of guest room conflicts while meeting the needs of the current user, thereby improving the management effect of the hotel and the experience of all users.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the technical solutions. Those of ordinary skill in the art should understand that those who modify or equivalently replace the technical solutions of the present invention without departing from the purpose and scope of the present technical solution should be covered by the scope of the claims of the present invention.

Claims

1. A hotel price dynamic management method based on artificial intelligence, characterized in that, Including: S1: Obtain the user's hotel reservation requirements, expected length of stay, and initial bid. S2: The AI model performs hotel matching based on the user's hotel reservation requirements and expected length of stay to obtain the target hotel and the corresponding target room type. S3: The AI model calculates the recommended price of the target room type based on real-time big data and the historical data of the target hotel. S4: When the user's initial bid is lower than the recommended price, the AI model calculates the recommended length of stay that meets the user's initial bid based on the initial price, combined with real-time big data and the historical data of the target hotel. S5: The AI model sends the recommended price and the recommended length of stay that meets the user's initial bid to the user, and obtains feedback information on whether the user accepts the recommended price or the recommended length of stay. S6: When the user accepts the recommended price, the AI model automatically reserves the target room type of the target hotel for the user based on the recommended price and the expected length of stay. S7: When the user accepts the recommended length of stay, the AI model automatically reserves the target room type of the target hotel for the user based on the initial price and the recommended length of stay.

2. The method for dynamic management of hotel prices based on artificial intelligence according to claim 1, wherein: In step S2, the AI model performs hotel matching based on the user location, reservation date, room type requirements, and number of occupants in the user's hotel reservation requirements, combined with the expected length of stay, to obtain the target hotel and the corresponding target room type that meet the user's hotel reservation requirements.

3. The method for dynamic management of hotel prices based on artificial intelligence according to claim 2, characterized in that: In step S2, the AI model performs hotel matching through the following steps: S201: Obtain the user location, reservation date, room type requirements, and number of occupants in the user's hotel reservation requirements. S202: Calculate the expected occupancy period based on the reservation date and the expected length of stay, and use the hotels with remaining rooms during the expected occupancy period as candidate hotels. S203: Calculate the straight-line distance d between the user's location and the candidate hotel, and calculate the distance attenuation factor δ of the candidate hotel in combination with the set attenuation coefficient k d = e -k·d ; S204: Match the room type requirements with the existing room types of the candidate hotels to obtain the room type matching degree m of the candidate hotels: if the room type requirements are the same as the existing room types of the candidate hotels, then the room type matching degree m = 1; otherwise, m = 0. S205: Calculate the occupancy fitness score of the candidate hotel based on the number of occupants and the maximum capacity of a single room in the candidate hotel S206: Weight the distance decay factor, room type matching degree, and number of occupants adaptability score of the candidate hotels to obtain the matching score S of the candidate hotels. The calculation formula of the matching score is expressed as: S = ω1·δ n + ω2·m + ω3·s n ; Where: ω1, ω2, ω3 represent the set dynamic weights. S207: Use the candidate hotels with a matching score higher than the score threshold as the target hotels, and use the room types that meet the user's requirements in the target hotels as the target room types.

4. The method for dynamically managing hotel prices based on artificial intelligence according to claim 3, characterized in that: In step S206, when the AI model calculates the matching score S of the candidate hotels, a user preference coefficient is introduced for calculation. The calculation formula of the matching score is updated to: S = (ω1·δ n + ω2·m + ω3ωs n )·η; Where: η represents the user preference coefficient, which is calculated based on the user preferences and the facilities and services of the candidate hotels. The user preferences include any one or more of free internet, free breakfast, free gym, free swimming pool, near the city center, near the subway station, and having high-quality restaurants. When the facilities and services of the candidate hotel meet all user preferences, the user preference coefficient η = 1; when the number of items that the facilities and services of the candidate hotel do not meet the user preferences does not exceed one item, the user preference coefficient η = 0.9; when the number of items that the facilities and services of the candidate hotel do not meet the user preferences is greater than or equal to two items, the user preference coefficient η = 0.

8.

5. The method for dynamic management of hotel prices based on artificial intelligence according to claim 1, characterized in that: In step S3, the AI model calculates the recommended price through the following formula: P t = (a·P c + b·P j )·θ h ·δ t ; Where: P t represents the recommended price of the target room type; P c represents the historical lowest price of the target hotel for the target room type; P j represents the average price of quotes for the same type of room as the target room type in hotels of the same type as the target hotel; a and b represent the set weights; θ h represents the holiday premium factor, which is related to whether the user's booking date is a weekend or a holiday: when the user's booking date is a weekend, the holiday premium factor θ h = 1.2, when the user's booking date is a holiday, the holiday premium factor θ h = 1.5, when the user's booking date is neither a weekend nor a holiday, the holiday premium factor θ h = 1; δ t represents the supply and demand factor for the time period; In the formula: λ represents the time period sensitivity coefficient, which is associated with the time period; One day is set as three time periods: Time period one: from 0:00 to 8:00, the time period sensitivity coefficient λ = 0.1; Time period two: from 8:00 to 16:00, the time period sensitivity coefficient λ = 0.5; Time period three: from 16:00 to 24:00, the time period sensitivity coefficient λ = 0.

3.

6. The method for dynamically managing hotel prices based on artificial intelligence according to claim 1, characterized in that: In step S4, the AI model calculates the recommended length of stay through the following steps: S401: Obtain the historical transaction average price P of the target room type for different lengths of stay at different times from the historical data of the target hotel h (w, d) and the corresponding average occupancy rate, where w represents the time period and d represents the length of stay; S402: Obtain the reference average price P of the room type of the same type as the target room type in hotels of the same type as the target hotel during the current period w c and the price fluctuation index I m (w c ), where the price fluctuation index refers to the change amplitude of the reference average price during the current period w f relative to the previous period; Calculate the price adjustment coefficient K(w c ) of the current period w c as 1 + I f (w c ); c c ) f (w c );​ S403: Obtain the historical transaction average price P for different lengths of stay d during the current period w c of the target hotel; h (w c , d) and the corresponding average occupancy rate O h (w c , d), and calculate the historical average price P c for the current period w a (w c ); The formula is expressed as: where: D max represents the maximum number of occupancy days; S404: Calculate the corresponding price reference value P c based on the historical average price P a (w c ) and the reference average price P m (w c ); b ; The formula is expressed as: S405: According to the current period w c 's price adjustment coefficient K(w c ) and the price reference value P b to calculate the upper limit P max and the lower limit P min of the price reference interval; The formula is expressed as: P max = P b ·K(w); P min = P b / K(w); S406: From the historical transaction average price P of different lengths of stay in the current period w c screen out the target length of stay d that satisfies P h (w c , d) such that P max < P h (w c , d m ) < P min ; m ; S407: According to the user's initial bid P i Calculate the matching degree M(d m ) for each target occupancy day d m ; Combine the average occupancy rate O c (w h , d) of the current period w c to calculate the comprehensive evaluation value V(d m ) for each target occupancy day d m ; The formula is expressed as: V(d m ) = M(d m )·O h (w c , d m ); S408: Select the target occupancy days d m ) with the highest comprehensive evaluation value V(d m as the recommended occupancy days.

7. The method for dynamic management of hotel prices based on artificial intelligence according to claim 6, wherein: In step S408, when there are multiple target occupancy days with the highest comprehensive evaluation value V(d m ), the AI model selects the smallest target occupancy day as the recommended occupancy day.

8. The method for dynamic management of hotel prices based on artificial intelligence according to claim 6, characterized in that: In step S406, the target length of stay d is screened out m After that, the AI model calculates the target lengths of stay d based on the user's room reservation date m The corresponding target check-in dates; obtain the booked dates of the target room type of the target hotel, and determine whether the booked date of the target room type conflicts with the target check-in date corresponding to the target length of stay d m Delete the target length of stay d corresponding to the target check-in date where a conflict occurs m .

Images