US20230385718A1 - Artificially Intelligent Computing Engine for Travel Itinerary Resolutions - Google Patents
Artificially Intelligent Computing Engine for Travel Itinerary Resolutions Download PDFInfo
- Publication number
- US20230385718A1 US20230385718A1 US18/365,819 US202318365819A US2023385718A1 US 20230385718 A1 US20230385718 A1 US 20230385718A1 US 202318365819 A US202318365819 A US 202318365819A US 2023385718 A1 US2023385718 A1 US 2023385718A1
- Authority
- US
- United States
- Prior art keywords
- itinerary
- travel
- network
- request
- objects
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013473 artificial intelligence Methods 0.000 claims description 12
- 238000003909 pattern recognition Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 125000002015 acyclic group Chemical group 0.000 claims description 5
- 238000000034 method Methods 0.000 abstract description 52
- 238000005516 engineering process Methods 0.000 description 37
- 230000008569 process Effects 0.000 description 17
- 230000001419 dependent effect Effects 0.000 description 11
- 230000004048 modification Effects 0.000 description 11
- 238000012986 modification Methods 0.000 description 11
- 230000015654 memory Effects 0.000 description 10
- 230000010006 flight Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 9
- 238000013528 artificial neural network Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000008901 benefit Effects 0.000 description 5
- 230000001427 coherent effect Effects 0.000 description 5
- 230000009471 action Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000002093 peripheral effect Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 241000219051 Fagopyrum Species 0.000 description 2
- 235000009419 Fagopyrum esculentum Nutrition 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 230000002441 reversible effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 241001481833 Coryphaena hippurus Species 0.000 description 1
- 241000287828 Gallus gallus Species 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000013070 change management Methods 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/02—Reservations, e.g. for tickets, services or events
- G06Q10/025—Coordination of plural reservations, e.g. plural trip segments, transportation combined with accommodation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/3407—Route searching; Route guidance specially adapted for specific applications
- G01C21/343—Calculating itineraries, i.e. routes leading from a starting point to a series of categorical destinations using a global route restraint, round trips, touristic trips
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/14—Travel agencies
Definitions
- the present technology relates generally to the processing and fulfilling of natural language travel requests, and more specifically, but not by way of limitation, to an exchange that allows suppliers to provide inventory records and customers to input travel itinerary requests in a natural language format, and fulfills the travel itinerary requests by applying pattern recognition artificial intelligence and/or semantic parsing to inventory records and travel itinerary requests to obtain matches therebetween.
- the ability to sell more inventory/content, sell current inventory more efficiently, and to differentiate product is extremely important and urgent to suppliers, especially in the travel and hospitality industries. Additionally, consumers want and need more choice and inventory/content.
- the current legacy supply chain for fulfilling travel related needs of consumers is complicated and remains under the control of various companies, most of which directly or indirectly compete with one another. Even if those within the supply chain are not hindered from cooperating by competition, the division of services/responsibilities within a single supplier may further hinder these legacy supply chains.
- current inventory may be maintained by one entity or department while flights are managed by another department and/or business.
- airline rules and pricing may be managed by yet another department and/or business. Business processes that interact with these legacy systems must be structured to correspond to these entities and their rules.
- a method for fulfilling travel requests may commence with receiving a travel request from a user and determining itinerary components based on the travel request.
- the method may further include generating an itinerary network based on the itinerary components.
- the itinerary network may be generated by creating a plurality of nodes and creating a plurality of edges within the itinerary network.
- Each of the plurality of nodes may represent information associated with the travel request.
- Each of the plurality of edges may connect two of the plurality of nodes.
- the plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes.
- the method may further include generating a travel itinerary responsive to the travel request.
- the travel itinerary may be consistent with the itinerary network.
- the method may continue with presenting the generated travel itinerary to the user on a user interface of a computing device associated with the user.
- the present technology may be directed to a system for fulfilling travel requests.
- the system may include a memory for storing executable instructions, a processor for executing the instructions, and a parser stored in the memory and executable by the processor.
- the parser may be configured to receive a travel request from a user and determine itinerary components based on the travel request.
- the parser may be further configured to generate an itinerary network by creating a plurality of nodes within the itinerary network and creating a plurality of edges within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request.
- Each of the plurality of edges may connect two of the plurality of nodes.
- the plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes.
- the parser may be further configured to generate a travel itinerary responsive to the travel request.
- the travel itinerary may be consistent with the itinerary network.
- the parser may be further configured to present the generated travel itinerary to the user on a user interface of a computing device associated with the user.
- FIG. 1 illustrates an exemplary architecture for practicing aspects of the present technology.
- FIG. 2 illustrates an exemplary itinerary processing system, constructed in accordance with the present technology.
- FIG. 3 illustrates flow diagram of events through an exchange system.
- FIG. 4 illustrates a flow diagram on an exemplary method for processing natural language travel requests.
- FIG. 5 illustrates an exemplary method for notifying suppliers of a natural language travel request.
- FIG. 6 illustrates an exemplary flow diagram of a process for fulfilling a schedule.
- FIG. 7 shows an itinerary network, according to an example embodiment.
- FIG. 8 shows a sub-path of nodes of an itinerary network, according to an example embodiment.
- FIG. 9 shows a parser network as a cover for travel conversations, according to an example embodiment.
- FIG. 10 shows an itinerary network, according to an example embodiment.
- FIG. 11 is a block diagram of an exemplary computing system for implementing embodiments of the present technology.
- the present technology comprises systems, methods, and media for processing natural language travel requests. More specifically, but not by limitation, the present technology may fulfill travel requests in the form of natural language expressions of a travel itinerary.
- An artificial intelligence engine used in methods and systems of the present disclosure may act as a priori artificial intelligence engine (i.e., may assume an a priori knowledge of certain structures).
- the present technology provides an efficient and simplified supply chain for the addition, organization, and consumption of inventory, together with a simplified distribution model. Additionally, the systems provided herein may also interact seamlessly with, and coexist with, legacy systems.
- the present technology provides increased efficiency and capabilities, allowing access to greater amounts of content that may be utilized to fulfill natural language travel requests.
- a URL is provided as a solution or a few thousand options for a single request or a component of a request
- the preset technology provides coherent solution(s) for natural language travel requests.
- the present technology may be implemented within the context of an exchange system that allows suppliers to provide inventory records and customers to input travel itinerary requests in a natural language format and fulfills the travel itinerary requests by applying pattern recognition artificial intelligence and/or semantic parsing to inventory records and travel itinerary requests to obtain matches therebetween.
- the present technology may facilitate an exchange that fulfills natural language travel requests.
- the present technology may be implemented within the context of an exemplary architecture 100 , hereinafter “architecture 100 ” as shown in FIG. 1 .
- the architecture 100 may be described as generally including an exchange 105 (also referred to herein as exchange system 105 ).
- Consumers 110 and third party suppliers 115 may communicate with either the exchange 105 , via a network 120 .
- the network 120 may include any one (or combination) of private or public communications networks such as the Internet.
- the consumers 110 may interact with the exchange 105 via end user client devices that access a web based interface or an application resident on the end user client device.
- the third party suppliers 115 may communicatively couple with the exchange 105 over the network 120 via an application programming interface (API). It is noteworthy that other methods/systems that allow the third party suppliers 115 and the exchange 105 to communicatively couple with one another, that would be known to one or ordinary skill in the art, are likewise contemplated for use in accordance with the present disclosure.
- API application programming interface
- the exchange 105 may include a cloud based computing environment.
- a cloud-based computing environment is a resource that typically combines the computational power of a large grouping of processors and/or that combines the storage capacity of a large grouping of computer memories or storage devices.
- systems that provide a cloud resource may be utilized exclusively by their owners, such as GoogleTM or Yahoo!TM; or such systems may be accessible to outside users who deploy applications within the computing infrastructure to obtain the benefit of large computational or storage resources.
- the cloud may be formed, for example, by a network of web servers, with each web server (or at least a plurality thereof) providing processor and/or storage resources. These servers may manage workloads provided by multiple users (e.g., cloud resource consumers or other users). Typically, each user places workload demands upon the cloud that vary in real-time, sometimes dramatically. The nature and extent of these variations typically depend on the type of business associated with the user.
- the exchange 105 may be generally described as a particular purpose computing environment that includes executable instructions that are configured to receive and fulfill natural language requests, such as travel itinerary requests.
- the exchange 105 may include executable instructions in the form of an itinerary processing and fulfillment application, hereinafter referred to as “application 200 ,” “a system for fulfilling travel requests,” or “a system.”
- application 200 an itinerary processing and fulfillment application
- FIG. 2 illustrates and exemplary schematic diagram of the application 200 .
- the application 200 is shown as generally comprising components such as a semantic parsing module, hereinafter referred to “a parser” or “parsing module 205 ,” a pattern recognition artificial intelligence engine, hereinafter “AI engine 210 ,” a scheduling module 215 (also referred to herein as scheduling module 215 ), and a modification module 220 . It is noteworthy that the application 200 may include additional modules, engines, or components, and still fall within the scope of the present technology.
- module and “engine” may also refer to any of an application-specific integrated circuit (ASIC), an electronic circuit, a processor (shared, dedicated, or group) that executes one or more software or firmware programs, a combinational logic circuit, and/or other suitable components that provide the described functionality.
- ASIC application-specific integrated circuit
- processor shared, dedicated, or group
- individual components of the application 200 may include separately configured web servers.
- FIG. 3 includes an exemplary flow diagram that illustrates the flow of data from a publishing environment into an exchange, along with the receipt of natural language travel requests and their fulfillment. While functional details regarding how the exchange 105 processes and fulfills natural language travel requests will be described with reference to additional figures described below (e.g., FIGS. 4 - 6 ), the overall operational flow of the exchange system 105 is shown in FIG. 3 .
- the scheduler module 215 may utilize the parsing module 205 to interpret the natural language queries.
- FIG. 4 illustrates a flowchart of an exemplary method for processing natural language travel requests.
- the parsing module 205 may assume an a priori knowledge of certain structures and intent over a class of information (for example, the hospitality and travel space).
- the natural language travel requests received by the parsing module 205 may comprise a textual request, a spoken (e.g., audio format) request, a location based request, an input based request (e.g., a click of an object on a map), a global positioning signal, and/or any combinations thereof.
- the request may comprise a non-natural language request, such as a keyword request, a Boolean phrase, and so forth.
- the parsing module 205 may infer a pre-determined set of information through a pattern recognition artificial intelligence module, such as the AI engine 210 .
- the parsing module 205 may first (Step 405 ) delimit the natural language query. For example, the parsing module 205 may determine inventory components in the query.
- the parsing module 205 may parse through each delimited string (Step 410 ) and transmit the delimited strings to the AI engine 210 .
- the AI engine 210 may act as an a priori AI engine (i.e., may assume an a priori knowledge of certain structures). While the a priori AI engine is discussed herein in terms of its application to travel, a person of ordinary skill in the art would understand that the engine can be similarly utilized in any domain.
- the a priori AI engine does not use evidentiary data required for the formulation of the AI.
- the knowledge and understanding have to exist before any data is provided. This is in stark contrast to current AI engines in the market, which use a posteriori AI that is more commonly prevalent and is entirely dependent on data and experience.
- the AI required for the a priori AI engine cannot be ascertained from the interaction between users and the system. Much of the intelligence is knowledge and understanding that is not part of this interaction but is contained in the structures and morphology behind the interface. The totality of the understanding of requests by the a priori AI engine is not contained within the conversation with a user.
- the a priori AI engine cannot be reduced to a set of rules. This may make the system entirely deterministic, which it is not, and may require the system to consider so many cases that it can be practically infinite in size.
- a rule in a rule-based system, a rule must be given for every case. In any non-trivial system, it is not possible to cover every potential scenario that may occur, as per Gödel's incompleteness theorem. In the a priori AI engine, the AI can ascertain the rule, intention, and action for all cases.
- the a priori AI engine may essentially deal with the conversation itself, not just the content of the conversation.
- each conversation initiated by either the user or the system may have a number of essential parts.
- the conversation may have an objective or goal.
- the objective may be the building or modification of an itinerary.
- the objective may be a response from the user to a question posed by the system to the user dealing with errors, or the like.
- each of the conversations may have a state, such as awaiting answer, completed, and the like. As would be understood by a person of ordinary skill in the art, many types of conversations can exist at the same time for a given user.
- the AI engine 210 may employ a combination of phraseology and keyword inference (Step 415 ) to decode what type of request is being made. Specifically, the AI engine 210 may determine the itinerary components by decoding the itinerary components from the travel request by determining phrases and tokens of the travel request. The AI engine 210 may reference the metadata database and the equivalence class database. Keywords included in an AI pattern recognition database may direct the AI engine 210 to appropriate content categories for the itinerary components included in the request (Step 420 ). The AI engine 210 may employ additional inferential methods as well as statistical methods and frequency to determine where and how to match content to the request.
- the parsing module 205 may evaluate each word of the sentence. If no keywords are found, nothing is constructed. However, the AI engine 210 may employ a “similar to” inference functionality which allows for variation among the phraseology to account for different ways that natural language queries may be structured such as incorrect spelling, grammar, and similar contingencies.
- the parsing module 205 determines a node type for each of the itinerary components and ascertain dependencies between each of the itinerary components based upon respective node types. It will be understood that the parsing module may effectuate construction of itineraries in a variety of manners. For example, the parsing module 205 may parse the words of the request in a sequential manner. The parsing module 205 may also parse the request to determine categories of itinerary components included in the request. In other instances, the parsing module 205 may delimit the request.
- the method for fulfilling travel requests of the present disclosure utilizes an itinerary network graph with nodes and dependencies.
- the parsing module 205 may utilize a directed acyclic graph (DAG), also referred to as an “itinerary network,” to interpret natural language queries.
- DAG directed acyclic graph
- the information extracted by the parsing module 205 may be utilized to generate an itinerary network that provides a further dynamic intelligence to the parsing module 205 in understanding the requested, parsed information, and assist the parsing module 205 in determining possible logical and logistics connections (e.g., location, time, and traveler preference based dependencies). Therefore, the parsing module 205 may generate the itinerary network based on the itinerary components of the travel request.
- the parsing module 205 may create a plurality of nodes within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request.
- the parsing module 205 may further create a plurality of edges within the itinerary network. Each of the plurality of edges may connect two of the plurality of nodes. The plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes.
- the parsing module 205 may generate a travel itinerary responsive to the travel request.
- the travel itinerary may be consistent with the itinerary network.
- the parsing module 205 may present the generated travel itinerary to the user on a user interface of a computing device associated with the user.
- the nodes may be of different types and may store information about the itinerary.
- the nodes may represent cities, hotels, flights, and destination content (such as a concert or car rental).
- the nodes can contain information specific to the type of node. For instance, a city node may contain an airport, whereas a flight node may contain the class of service.
- Edges connect nodes to show dependencies between nodes and may contain information themselves.
- the edges are directed to represent the order of nodes in time.
- edges can be air transport, car transport, start to start, or start to finish.
- the edges can present information that may be important when scheduling the itinerary.
- an itinerary network can be used to represent this travel request of the user.
- two city nodes may be created first: one for the source city and one for the destination city expressed by the user.
- the city nodes may be populated with the names of the cities in the user request and possibly names of airports.
- these cities may be connected with a directed edge (dependency).
- the edge may contain information for a flight, rail, bus, or other mode of transport between the two cities, as well as other information, such as the scheduling type.
- the mode of transport may be presented as the node type and may be further qualified in the dependencies between the city and the transport node.
- the first condition is that each passenger must have a path-connected graph.
- the second condition is that the graph is acyclic.
- the a priori AI engine understands that the itinerary network is a compact path connected topological space, with all the mathematical tools and theorems that follow from that at disposal of the a priori AI engine. This allows the AI to fully understand the properties of the itinerary network and to know if the path connectedness property is broken and how to resolve it by communication with the user. This understanding is naturally inherent within the systems and the methods of the present disclosure and it has been built into the a priori AI engine.
- the itinerary network provides a coherent consistent entity that can then be treated by a scheduler for scheduling of a travel request. More than this, the systems and the methods of the present disclosure cover every case one could glean from data or experience, without having to input a multitude of individual rules to teach the a priori AI engine.
- the resultant itinerary network contains all of the information necessary for the system to supply congruent content (logistic and curated data).
- the characteristics of the itinerary are the characteristics of a path-connected, directed acyclic, compact topological space or combinations of such spaces. This provides the capability of scheduling the itinerary in a consistent and coherent manner. It also allows the parser to understand all the aspects of the itinerary.
- the itinerary network may describe the itinerary, such as flights, hotels, cars, events, reservations, and other disparate content.
- the itinerary network also describes the temporal and dependent relationships of the passengers and their content.
- the understanding of the itinerary depends on the scheduler, which instantaneously understands the logistics and content of the network.
- the parser may be imbued with the lexicon and capabilities of the itinerary network and as such can understand and interact with the itinerary network independently or through conversations with the user.
- a priori AI engine directed to not simply understanding language.
- the language is an insufficient condition to the AI of the system.
- the a priori AI engine may inform what language is to be understood and what language is not relevant.
- the a priori AI engine may use the parser, the scheduler, and so forth, to seek to fulfill an itinerary request. For this purpose, the a priori AI engine may build an itinerary network and schedule the itinerary network with appropriate curated content.
- the parser may use the tree of nodes (parser trees) to address any conversation that seeks to instruct the system to build or modify any itinerary.
- itinerary components may comprise travel or non-travel node types.
- the parsing module 205 may obtain source and destination information from relevant itinerary components (Steps 425 and 430 ). If they do not exist on the itinerary network, the parsing module 205 may add them to the itinerary network.
- the parsing module 205 may determine if the node has a time or location dependency to another node (Step 435 ). If the node does have a dependency, the parsing module 205 checks to see if the dependent node exists. If it does not, the parsing module 205 will create the node and populate the node with any necessary attributes (Step 440 ).
- the parsing module may also identify traveler preferences.
- Traveler preferences can include general or specific preferences and are requested or ordered in natural language. Some examples include: “give me cheapest flight,” “do not book me into any Hilton hotels,” “provide me four-star hotels or better,” and “If I am in San Francisco book me into the San Mateo Sofitel hotel.”
- the process of identifying nodes for itinerary components and interrelating these nodes may be referred to as generating an itinerary network.
- the itinerary network may be utilized by the scheduler module 215 to generate an unconstrained schedule for the natural language request, as will be described in greater detail herein.
- the characteristics of an itinerary may include a stage, a state, and a context. These characteristics allow the parser to be sentient in the conversation with the user through the life cycle of the itinerary (pre- to post-booking and change management) and partial completion of the itinerary.
- the sentience includes understanding the meaning of the request (based on from context, stage and state).
- the goal orientation of the system may be to serve an itinerary request.
- the stage may include a pre-booking state and a post-booking stage. The stage is important for establishing the context, as well as for communicating relevant information to a user, such as change and cancellation fees or other disruptions to the itinerary as a result of a change or delay.
- the state is broadly defined by two elements, namely the itinerary network and the conversation.
- the state includes determining, for example, whether the system is conversing with the user about the itinerary, whether the user has booked a trip, whether there are flags on the itinerary that need to be addressed (such as incomplete information or infeasible solution), and so forth.
- Context may be built into an itinerary object. This may provide sentience to the AI engine as the AI engine interprets new requests or changes for the itinerary.
- the context may be dynamic as the itinerary object changes its stage, state and topology. As one can easily see, traditional machine learned artificial intelligence simply cannot cope with dynamic context because the number of experiences is an NP-Complete decision problem, even a higher order of complexity than NP-Hard decision problem.
- the neural networks and other statistical-based inference models will pick a 5 out 99% of the time correctly when shown numbers.
- the neural networks may still infer, with the same confidence, that a static page before a user is a number 2 .
- the neural networks cannot answer the user.
- the structure of conventional neural networks is incapable of doing anything other than returning a probability.
- the a priori AI engine described herein changes the problems complexity class from NP-complete to order kN.
- the a priori AI engine is imbued with knowledge of travel, and language does not need to be inferred, language needs only be matched as the set of language patterns are generated by the capabilities and understandings of the underlying morphology/referential framework. In this sense, the behavior of the a priori AI engine is totally unique in the ecosystem of natural language human machine interfaces.
- the parsing module 205 may generate an itinerary network in any order, allowing itinerary components to be inserted into the itinerary network when a starting/ending reference point has been established, such as when the source and destination itinerary components are identified.
- An exemplary itinerary network 500 is illustrated in FIG. 5 , and is constructed from the natural language travel request, “From Toronto to Seattle. From Seattle to Tokyo. Stay at any preferred hotel with a buckwheat pillow. Reservations at Daniel's Broiler and a well-known sushi restaurant near my hotel in Tokyo.”
- traveler preferences that were received in natural language format include: “Give me lowest cost tickets,” “Exclude Hilton chain,” “Route me through Cincinnati on route to Seattle,” “Integrate my calendar and exclude red category,” as well as many other traveler preferences, which would be known to one of ordinary skill in the art with the present disclosure before them.
- parsing module 205 may populate each itinerary component with attributes identified by the AI engine 210 , such as node type and dependencies.
- the parsing module 205 may then establish dependencies between appropriate itinerary components. There is an extended set of dependencies that extend from the normal start-start, start-finish, finish-start, and finish-finish to parent-child, local dependency, and so forth. Other exemplary dependencies may include, but are not limited to: Air-Connect, Local-Connect, Activity, Location, Time, Time and Location, Logical-Connect, and dependencies that relate to the travel data of another traveler such as “Travel Together” and “Travel Meet At.”
- Time dependencies may be utilized to generate itinerary schedules in reverse order, based upon an end point. For example, using a scheduled meeting as an end point, the present technology may create and fulfill a travel itinerary for a customer that ensures that the customer arrives in the proper location and at the proper point in time to allow them to attend the scheduled meeting.
- the parsing module 205 may generate an adjacency matrix using the itinerary components and their respective dependencies. Utilizing the adjacency matrix, the parsing module may create an itinerary network using the adjacency matrix.
- the parsing module 205 may determine a topological ordering of itinerary components using the itinerary network. It is noteworthy that the topological ordering of itinerary components may comprise an arrangement of the itinerary components using their respective location and time dependencies used by the scheduling module 215 to generate an unconstrained schedule, as will be discussed in greater detail below.
- the parsing module 205 and AI engine 210 may utilize the itinerary network to inform the scheduling module 215 in generating schedules and allocating inventory to the schedules. For example, if an itinerary node includes an activity, or location dependent node such as a theater, restaurant, hotel, conference, or the like, the parsing module 205 will understand the activity must take place in a city. Thus, depending on the phraseology encountered by the AI engine 210 , the AI engine 210 may loop through the admissible ways of saying “I'm here” and compare the location against a city dictionary list. If the city is valid, the AI engine 210 may look for the city name in the itinerary network, creating a node if the AI engine 210 does not find an appropriate node or adding the activity node with a time/location dependency underneath.
- Dependent activities may have their own dependencies as well (for example, local transportation between a restaurant and a conference). Moreover, preferences associated with each dependent node may appear as another level of dependency (for example, a buckwheat pillow in a hotel room).
- the parsing module 205 may check to see if a desired node is present in the itinerary network and create nodes as needed. Since each city and activity has a time dependency as well as a location dependency, in complex itineraries with multiple cities being visited multiple times by multiple people, the parsing module 205 may prevent confusion relative to a dependent node's dependencies relative to location and time. The parsing module 205 may also inform the consumer that he has asked for a hotel in a city to which the consumer is not traveling.
- the parsing module 205 may infer there must be a source and destination and a mode of travel therebetween. The parsing module 205 may further infer what kind of travel is most appropriate, so a consumer will not find himself driving or taking the train from Miami to Manchester, U.K.
- the parsing module 205 may not dictate a mode of travel; however, a consumer may choose to take any form of transportation desired.
- the parsing module 205 may send the phrase to the AI engine 210 , extract the source and destination cities, match them against the city list dictionary, and check the network for the nodes existence and add them if necessary.
- the AI engine 210 may then add the travel node and a travel dependency between the travel node and the two cities to the itinerary network.
- a consumer may ask for any itinerary, in any order, and the present technology may produce a correctly networked schedule.
- the present technology may take the natural language phrase, “I want to go from Seattle to Dallas, Miami to Atlanta, Dallas to Miami, Toronto to Seattle.”
- the parsing module 205 may create an itinerary network which linked Toronto to Seattle to Dallas to Miami to Atlanta. As before, additional content nodes and dependencies may be added as required.
- the parsing module 205 may understand the different types of dependencies that occur. For instance, in Toronto there may be an Italian restaurant called Pizza Banfi. If a traveler preference indicates a hometown of Toronto, or location-based data from a consumers' cellphone indicates that the consumer is in Toronto, and consumer requests “From Pizza Banfi to Seattle,” the AI engine 210 may understand that the consumer requires transport between two points, but that one point is a city, and the other is a dependent node belonging to another city. The AI engine 210 may create the Toronto node, place the restaurant as a dependent node, arrange for transport to the airport, which is local dependency, a flight dependency between the two cities right after it creates the Seattle node.
- the scheduling module 215 may be executed to generate an unconstrained schedule from the itinerary network (e.g., DAG).
- the itinerary network e.g., DAG
- the generation of an unconstrained schedule establishes the earliest start and latest finish for all nodes and hence the initial starting point for all requests pertinent to the content represented by the nodes.
- the scheduling module 215 then employs one of several methods to resolve the allocation of content (e.g., inventory) to the requests for content and fill the itinerary.
- the scheduling module 215 may apply an Adaptive Method that “levels” the itinerary. For example, the scheduling module 215 may search content within the topological ordering. Each line item in the topology may be considered, the exchange searched, and/or offers obtained from the suppliers. The content request is established by the scheduling module 215 from the node type and its attributes as filled out by the parsing module 205 . These attributes also include general and specific preferences. A set of valid options may be obtained and ordered by the traveler preferences.
- the scheduling module 215 may employ additional methods to allocate inventory to the request.
- a best alternative e.g., available inventory
- This sets the starting conditions for successor nodes in the topology and the topology is then recursed by the scheduling module 215 using only the best client alternatives.
- a specific best path itinerary can be identified.
- the itinerary can be optimized with respect to an equivalence class of airline tickets, where the result from selecting a specific airline ticket does not impact the remainder of the itinerary.
- each alternative (up to some arbitrary limit) of the sorted list of nodes by client preferences may be considered by the scheduling module 215 and a separate itinerary developed for each.
- the scheduling module 215 processes each line item in the topology by applying a recursion algorithm.
- the results of this modal process may generate many different itineraries whose costs and time frames can vary substantially. These itineraries may be sorted in different ways using multiple sorting criteria; (shortest, lowest cost); (lowest cost, shortest); and so forth.
- the scheduling module 215 can dynamically schedule robustness into the schedule in the sense that it can maintain specific times required between flights; these can be in minutes, hours or days. The scheduler will automatically extend hotel stays if the flights do not leave on the same day as the hotel checkout.
- the scheduling module 215 may create time and space dependent solutions to the logical schedule dynamically, based on the offers made to the requested itinerary from suppliers.
- the scheduling module 215 maintains the dependencies so that requests remain accurate with respect to the current solution. In this manner, the logistics of travel are maintained and their constraints adhered to.
- the scheduling module 215 may be configured to always return a solution, even if the constraints cannot be met. This solution may comprise the closest available under the constraints and options that have been requested. It is noteworthy that when inventories for content are tight, it could take an extremely long time to find any solution. Therefore an “approximate fit” schedule may be preferred to no schedule.
- the scheduling module 215 may be configured to generate a leveled solution where the scheduling module 215 may allow requests to level out in time across the itinerary, showing when solutions are available. Thus, if a customer books a flight today to San Francisco, the scheduling module 215 may allow a solution for tomorrow if that is the only alternative.
- the scheduling module 215 may also provide one or more possible schedules (solutions) to the exchange 105 ( FIG. 1 ) in either a sequential or leveled manner. In the sequential method, all dependencies for a specific aspect of the itinerary may be filled before it is submitted to the exchange 105 . An alternative method allows the scheduling module 215 to maintain the times and dates specified, and only offers that match these times and dates are allowed.
- the scheduling module 215 may transmit itinerary components and/or entire itinerary schedules (such as the itinerary network) to the exchange 105 .
- the exchange 105 may employ a listener that immediately picks up the new requested line item or itinerary (Step 605 ).
- Line items or an itinerary may also be referred to as a “request.”
- the listener may identify the itinerary components (nodes) (Step 610 ) and/or a buyer profile or content profile associated with the itinerary (Step 615 ).
- the listener may compose a list of the suppliers that have indicated that they want to bid on these types of line items or itineraries (Step 620 ).
- Suppliers may be notified of these requests and can then analyze them and bid on them or entire itinerary (Step 625 ).
- the transactions made available to the suppliers contain the entire content, inferential information, and/or semantics of the request together with a framework for interpreting the same.
- the supplier can respond based on review of its inventory and availability. In other instances, the supplier can dynamically decide what to do with the content and price through its own legacy systems.
- the exchange 105 makes available APIs to interrogate the system for any requests that the supplier may want to look at (For example, City-Pair for flights and/or Activity Keyword or partial Keyword).
- the default listener is the exchange itself that will process, search, and respond to every request.
- Offers may be written back to the exchange in the form of a response. Additionally, suppliers can respond with any additional content they desire, together with pricing for itinerary components. For example, an airline can offer a golf bag at $100 with the air ticket at a reduced price. Other similar types of vouchers may be exchanged or facilitated utilizing the present technology.
- offers are written to the exchange 105 , they are matched against the line items and itinerary generated by the scheduling module 215 . In some instances, before being considered, the offers may be passed through a set of filters that describe the traveler's restrictions and preferences.
- An exemplary flow diagram of a process 600 for fulfilling a schedule is depicted in FIG. 6 .
- the scheduling module 215 may selectively adjust the allocation of inventory based upon various constraints such as available/dynamic inventory. In other embodiments, the scheduling module 215 may adjust the schedule provided to the consumer based upon inferential modeling of the consumer's request (for example, when the consumer expresses a traveler preference that is new or contradictory to a known traveler preference for that particular consumer).
- the modification module 220 may be executed to process modifications to travel itineraries.
- the modification module 220 may receive a modification to the travel itinerary from a traveler who has previously input a natural language travel request that has been processed using the aforementioned methods to generate an itinerary schedule.
- the modification module 220 may adjust the allocation of available inventory for each itinerary component remaining in the travel itinerary based upon one or more dependency adjustments associated with the plurality of nodes and the plurality of edges and caused by modification of the travel itinerary. That is, because the parsing module 205 appreciates the dependencies between the current itinerary components in the schedule, along with the dependencies of the modification, the parsing module 205 may insert the modification into the schedule and adjust other itinerary components, as necessary. Therefore, even for an itinerary that is currently being executed (e.g., traveler has already completed at least a portion of their itinerary), the parsing module 205 may adjust the schedule to ensure that traveler preferences are maintained. For example, if cost is an important traveler preference, the parsing module 205 may adjust the schedule to cause the least impact from a cost perspective.
- the parser trees may define a priori AI engine that can deal with any conversation seeking to build or change itineraries through successive iterations of the conversation—all in place before the first ever conversation takes place. Furthermore, no rules are required; rather, the understanding is fundamental and complete in advance of the conversations or requests with users.
- the itinerary network may handle the language interface between the user and the system.
- the itinerary network also facilitate interactions between various elements of the system itself.
- the network nodes may be comprised of tokens.
- the paths (sequence of tokens) in the itinerary network are called ‘phrases’. Together, the tokens and the phrases may make a normalized travel language or meta language. This is strictly a pattern (or phrase) recognition process; no traditional linguistic grammars are used in this process. This is a departure from traditional Natural Language Processing methods which rely heavily on the linguistic syntactic composition of a sentence and data to learn relationships between the various syntactic elements. The system is able to construct the phrases without ever having to have experienced them.
- Tokens are the categorization of information related to travel.
- the tokens may include travel attributes related to the travel request, such as passengers, source and destination cities, times and dates, time and date restrictions, hotel names, addresses, and other descriptive elements, such as dependencies between the various itinerary network elements.
- a phrase is an ordered set (or sequence) of tokens that describes an intention, request, or like communication with the system (parser). In effect, the entire conversation related to the itinerary may be pre-defined in this manner. All travel conversations and their intent, regardless of context, and the like, can be reduced to a normalized set of phrases in the parser trees.
- FIG. 7 shows an itinerary network 700 , according to an example embodiment.
- the itinerary network 700 may include nodes 705 and edges 710 .
- FIG. 8 shows a sub-path of nodes of an example itinerary network 800 .
- the nodes may have annotations.
- a node 805 may have an annotation 810
- a node 815 may have an annotation 820 .
- Each node in the itinerary network 700 or 800 can have one or more annotations 810 and 820 associated with it.
- the annotations 810 and 820 may describe specific information or understanding.
- the node 805 may have the annotation 810 ‘flight’. This annotation may mean that there is sufficient information (within the path or network subtree reaching the node 805 ) to determine that there is an intentional request for a flight.
- the AI engine understands this information together with the data associated with the flight.
- Another sub-path may add additional information to an existing node or define a global requirement or preference or a change to the existing itinerary network.
- the sequence may be matched against a sub-path in the parser tree.
- the system may be configured to perform error handling and the ability to anticipate “next” likely token(s). Specifically, at any stage in matching the ordered set of tokens obtained from the request, the parser already knows the set of next possible valid tokens of the travel request. This is significant because the parser can detect insufficient information to fully understand a request, as well as missing information (tokens) necessary for generating the travel itinerary. Based on the possible next tokens, the system may detect insufficient information associated with the travel request and request the insufficient information from the user.
- An example of building a set of phrases using the parser trees Once a match for a sub-path with an annotation is obtained, there is created an immediate and absolute “understanding” and a potential actionable event.
- the nodes in the sub-path may not have any physical resemblance to what was understood or inferred or the actionability thereof.
- the a priori AI engine has understanding before there is any specific request and is general in nature, and the a priori AI engine requires no data for this understanding; rather, it can handle any such request associated with the sub-path. Indeed, the capabilities of the itinerary network and scheduler define the phrases in their totality. Additional phrases, when encountered, merely constitute an equivalence class relationship with an already extant phrase.
- the node 805 may have the attribution of a flight that means that the parser has intentionality to define a flight.
- the node 815 may have the attribution of a flight that means that the parser has intentionality to define a hotel.
- the system understands that what is requested is a flight and a hotel. The set of nodes leading up to this provides sufficient and necessary logical information to create a flight.
- an actionable phrase may start. This phrase invokes an action, for example, called “Build Flight and Hotel.” This action builds a travel itinerary and the itinerary network, which represents the travel itinerary in a unique manner.
- the system determines content nodes, city nodes, city children content, the relationships between the nodes, the dependencies between nodes, the ability to represent multiple travelers, and so forth.
- the Parser Complexity Class The cardinality of the set of equivalence classes is finite, countable, and has N members. This set may form a compact cover of the set of actions, requests, information, and so forth defined by the system.
- the equivalence classes which provide equivalent ways of stating any phrase, form a larger open cover 905 as shown on FIG. 9 and discussed below.
- the complexity class of the a priori engine may be ⁇ O(kN), where N is the finite number of equivalence classes (i.e., the finite number of statements about the travel itinerary possible to be made) and k is the maximum number of the equivalence classes of the statements. This holds true regardless of how many times phrases are matched in the parser trees or the size of the context in which they are made.
- FIG. 9 shows a parser network 900 as a cover for travel conversations, according to an example embodiment.
- the normalized set of phrases form a compact cover 910 of the intentionality space for travel
- each phrase can have a set of like phrases that are equivalent to the single member phrase of the compact set and form an open cover 905 .
- the set of phrases is O(N) complexity class
- the compact cover 910 together with the equivalence classes form a complexity class of O(kN). This is extremely important since most other AI technologies are essentially NP-complete.
- the natural language request of the user may have the following form: “I want to fly from Toronto to New York staying 3 days upper west side in a 5 star hotel then go to Miami staying in South Beach for 2 nights then to San Francisco staying at the Sofitel on Twin Dolphin drive for 2 nights then home. Business class tickets on all flights. Jonathan will join me in San Francisco stay at the same hotel and we will share a room. He will then go to Seattle stay at the Woodmark Hotel in Kirkland for 3 days then fly home.”
- the request may be processed into meta patterns (phrases) and these phrases may be used to build a directed, acyclic, path connected network, which in turn may be used to build a resource independent (logical and resource unconstrained) schedule.
- the goal of the system includes finding suitable content for this itinerary as defined by the itinerary network.
- the itinerary network may provide the context both when constructing the itinerary network for requests from the conversation, as well as when creating the goal for the system, namely, to fulfill this travel itinerary network.
- travel requests and successive requests may define the itinerary network and the system may always act to fulfill the itinerary network, whether by creating a new itinerary network or changing an existing itinerary network.
- This iterative process can continue indefinitely without human intervention.
- the system may be automatically goal seeking and the AI engine may be synthetic and analytic a priori (i.e., no data and/or experience may be required for the AI engine).
- the itinerary network 1000 may have a plurality of nodes 1005 , 1010 , 1015 associated with flights, hotels, cities, and so forth, and a plurality of edges 1020 , 1025 that connect the nodes.
- the AI engine may consequently deal with multiple states that occur over the lifespan of the travel itinerary.
- the AI engine also understands the properties of an itinerary object.
- the AI engine may understand that the itinerary object is a compact path connected topological space. When the path connectedness property is broken, the AI engine understands the implications of this and how to deal with it by communicating with the user. This understanding is naturally inherent to the AI engine and the methods of the present disclosure.
- the itinerary network may provide a coherent consistent entity that can be treated by the scheduler with a single set of methods and provide a scheduling AI ecosystem for the scheduling of the travel request.
- the AI engine is aware of the context of the travel itinerary, namely a stage, state, and all the methods and attributes the itinerary network represents.
- the resultant itinerary network contains all the information necessary for the system to supply congruent content (logistic and curated data).
- the characteristics of the itinerary are that of a path-connected, directed acyclic, compact topological space, or combinations of such spaces. This provides the capability of scheduling the itinerary in a coherent manner and allows the parser to understand all the aspects of an itinerary with respect to nodes of the itinerary network.
- the parsing may be fully integrated into the system, methods, and all the conversations that exist between the modules of the system, the state, and stage of the itinerary, the error conditions together with the conversations that may exist between the user and the system.
- the a priori engine may be, in fact, built in reverse order to the traditional model. All the goal seeking methods may be defined, given context in terms of the itinerary object, and then the specific phrases (meta language) may be defined in these terms.
- the a priori engine may include the itinerary network, the parser network and the phrases and the lexicon required to drive them.
- the set of phrases may be a compact cover of the truthful statements that can be made within the tautology of the a priori engine.
- the AI engine can be easily extended to cover new methods, lexicon, and phrases, without experimentation or any data sources.
- the a priori engine may comprise a set of phrases which are understood explicitly in terms of their respective goals/objectives, information content, queries, and the like. The phrases are understood in context of the current state of the itinerary, supply chain, and so forth.
- Embodiments of a computing system discussed herein include providing end to end digital service, where the AI engine extends through understanding of a request from a user, the goal seeking properties of constructing a congruent itinerary, modifying the itinerary, curating the content for each user, maintaining context and sentience, and seeking to complete goals.
- the AI engine discussed herein can be implemented via a specially-programmed and special-purpose computer.
- the computing engine may utilize one or more processors storing instructions, static and/or dynamic memory, one or more databases or other data structures, and network interface(s).
- FIG. 11 illustrates an exemplary computing system 1100 that may be used to implement an embodiment of the present technology.
- the system 1100 of FIG. 11 may be implemented in the contexts of the likes of computing systems, networks, exchanges, servers, or combinations thereof disclosed herein.
- the computing system 1100 of FIG. 11 includes one or more processors 1110 and main memory 1120 .
- Main memory 1120 stores, in part, instructions and data for execution by processor 1110 .
- Main memory 1120 may store the executable code when in operation.
- the system 1100 of FIG. 11 further includes a mass storage device 1130 , portable storage medium drive(s) 1140 , output devices 1150 , user input devices 1160 , a graphics display 1170 , and peripheral devices 1180 .
- FIG. 11 The components shown in FIG. 11 are depicted as being connected via a single bus 1190 .
- the components may be connected through one or more data transport means.
- Processor unit 1110 and main memory 1120 may be connected via a local microprocessor bus, and the mass storage device 1130 , peripheral device(s) 1180 , portable storage device 1140 , and display system 1170 may be connected via one or more input/output (I/O) buses.
- I/O input/output
- Mass storage device 1130 which may be implemented with a magnetic disk drive or an optical disk drive, is a non-volatile storage device for storing data and instructions for use by processor unit 1110 . Mass storage device 1130 may store the system software for implementing embodiments of the present technology for purposes of loading that software into main memory 1120 .
- Portable storage device 1140 operates in conjunction with a portable non-volatile storage medium, such as a floppy disk, compact disk (CD), digital video disc (DVD), or USB storage device, to input and output data and code to and from the computer system 1100 of FIG. 11 .
- a portable non-volatile storage medium such as a floppy disk, compact disk (CD), digital video disc (DVD), or USB storage device
- the system software for implementing embodiments of the present technology may be stored on such a portable medium and input to the computer system 1100 via the portable storage device 1140 .
- Input devices 1160 provide a portion of a user interface.
- Input devices 1160 may include an alphanumeric keypad, such as a keyboard, for inputting alphanumeric and other information, or a pointing device, such as a mouse, a trackball, stylus, or cursor direction keys, or voice to text.
- the system 1100 as shown in FIG. 11 includes output devices 1150 . Suitable output devices include speakers, printers, network interfaces, and monitors.
- Display system 1170 may include a liquid crystal display (LCD) or other suitable display device. Display system 1170 receives textual and graphical information and processes the information for output to the display device.
- LCD liquid crystal display
- Peripherals devices 1180 may include any type of computer support device to add additional functionality to the computer system.
- Peripheral device(s) 1180 may include a modem or a router.
- the components provided in the computer system 1100 of FIG. 11 are those typically found in computer systems that may be suitable for use with embodiments of the present technology and are intended to represent a broad category of such computer components that are well known in the art.
- the computer system 1100 of FIG. 11 may be a personal computer, hand held computing system, telephone, mobile computing system, workstation, server, minicomputer, mainframe computer, or any other computing system.
- the computer may also include different bus configurations, networked platforms, multi-processor platforms, and so forth.
- Various operating systems may be used including Unix, Linux, Windows, Macintosh OS, Palm OS, Android, iPhone OS and other suitable operating systems.
- Computer-readable storage media refer to any medium or media that participate in providing instructions to a central processing unit (CPU), a processor, a microcontroller, or the like. Such media may take forms including, but not limited to, non-volatile and volatile media such as optical or magnetic disks and dynamic memory, respectively. Common forms of computer-readable storage media include a floppy disk, a flexible disk, a hard disk, magnetic tape, any other magnetic storage medium, a CD-ROM disk, DVD, any other optical storage medium, RAM, PROM, EPROM, a FLASHEPROM, any other memory chip or cartridge.
Landscapes
- Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Automation & Control Theory (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Systems and methods for fulfilling travel requests are described herein. A method for fulfilling travel requests may commence with receiving a travel request from a user and determining itinerary components based on the travel request. The method may further include generating an itinerary network based on the itinerary components. The itinerary network may be generated by creating a plurality of nodes and creating a plurality of edges within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request. The plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes. The method may further include generating a travel itinerary responsive to the travel request. The method may continue with presenting the generated travel itinerary to the user on a user interface of a computing device associated with the user.
Description
- This application is a Continuation of U.S. non-provisional patent application Ser. No. 16/396,487, filed Apr. 26, 2019, which claims the priority benefit of U.S. provisional patent application No. 62/747,088, filed Oct. 18, 2017 and is a Continuation-in-part and claims the priority benefit of U.S. non-provisional patent application Ser. No. 13/420,179, filed Mar. 14, 2012, which in turn claims the priority benefit of U.S. provisional patent application Ser. No. 61/452,633, filed Mar. 14, 2011. The present application also claims the priority benefit of U.S. provisional patent application Ser. No. 62/747,088 filed on Oct. 17, 2018.
- U.S. non-provisional patent application Ser. No. 13/420,179, filed Mar. 14, 2012 is related to the Applicants' co-pending U.S. non-provisional patent application Ser. No. 13/419,989, filed Mar. 14, 2012 and issued Mar. 15, 2016, as U.S. Pat. No. 9,286,629, and to the Applicants' co-pending U.S. non-provisional patent application Ser. No. 13/420,433, filed Mar. 14, 2012 and issued Sep. 18, 2018, as U.S. Pat. No. 10,078,855. All of the above referenced applications are hereby incorporated by reference herein in their entirety.
- The present technology relates generally to the processing and fulfilling of natural language travel requests, and more specifically, but not by way of limitation, to an exchange that allows suppliers to provide inventory records and customers to input travel itinerary requests in a natural language format, and fulfills the travel itinerary requests by applying pattern recognition artificial intelligence and/or semantic parsing to inventory records and travel itinerary requests to obtain matches therebetween.
- The ability to sell more inventory/content, sell current inventory more efficiently, and to differentiate product is extremely important and urgent to suppliers, especially in the travel and hospitality industries. Additionally, consumers want and need more choice and inventory/content. The current legacy supply chain for fulfilling travel related needs of consumers is complicated and remains under the control of various companies, most of which directly or indirectly compete with one another. Even if those within the supply chain are not hindered from cooperating by competition, the division of services/responsibilities within a single supplier may further hinder these legacy supply chains. For example, with respect to an airline, current inventory may be maintained by one entity or department while flights are managed by another department and/or business. Moreover, airline rules and pricing may be managed by yet another department and/or business. Business processes that interact with these legacy systems must be structured to correspond to these entities and their rules. For each entity, a completely different set of requirements may be imposed upon business processes that depend upon these entities. In sum, the structures of these legacy supply chain systems make it extremely difficult, if not impractical, to properly aggregate offerings and/or add new inventory/content that would be recognized and accepted by the legacy systems.
- Furthermore, conventional artificial intelligence engines available in the market use of a posteriori artificial intelligence that is entirely dependent on data and experience and, hence, requires large amounts of resources to collect data and analyze experience.
- This disclosure is directed to systems and methods or fulfilling travel requests. According to some embodiments, a method for fulfilling travel requests may commence with receiving a travel request from a user and determining itinerary components based on the travel request. The method may further include generating an itinerary network based on the itinerary components. The itinerary network may be generated by creating a plurality of nodes and creating a plurality of edges within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request. Each of the plurality of edges may connect two of the plurality of nodes. The plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes. The method may further include generating a travel itinerary responsive to the travel request. The travel itinerary may be consistent with the itinerary network. The method may continue with presenting the generated travel itinerary to the user on a user interface of a computing device associated with the user.
- According to other embodiments, the present technology may be directed to a system for fulfilling travel requests. The system may include a memory for storing executable instructions, a processor for executing the instructions, and a parser stored in the memory and executable by the processor. The parser may be configured to receive a travel request from a user and determine itinerary components based on the travel request. The parser may be further configured to generate an itinerary network by creating a plurality of nodes within the itinerary network and creating a plurality of edges within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request. Each of the plurality of edges may connect two of the plurality of nodes. The plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes. The parser may be further configured to generate a travel itinerary responsive to the travel request. The travel itinerary may be consistent with the itinerary network. The parser may be further configured to present the generated travel itinerary to the user on a user interface of a computing device associated with the user.
- Certain embodiments of the present technology are illustrated by the accompanying figures. It will be understood that the figures are not necessarily to scale and that details not necessary for an understanding of the technology or that render other details difficult to perceive may be omitted. It will be understood that the technology is not necessarily limited to the particular embodiments illustrated herein.
-
FIG. 1 illustrates an exemplary architecture for practicing aspects of the present technology. -
FIG. 2 illustrates an exemplary itinerary processing system, constructed in accordance with the present technology. -
FIG. 3 illustrates flow diagram of events through an exchange system. -
FIG. 4 illustrates a flow diagram on an exemplary method for processing natural language travel requests. -
FIG. 5 illustrates an exemplary method for notifying suppliers of a natural language travel request. -
FIG. 6 illustrates an exemplary flow diagram of a process for fulfilling a schedule. -
FIG. 7 shows an itinerary network, according to an example embodiment. -
FIG. 8 shows a sub-path of nodes of an itinerary network, according to an example embodiment. -
FIG. 9 shows a parser network as a cover for travel conversations, according to an example embodiment. -
FIG. 10 shows an itinerary network, according to an example embodiment. -
FIG. 11 is a block diagram of an exemplary computing system for implementing embodiments of the present technology. - While this technology is susceptible of embodiment in many different forms, there is shown in the drawings and will herein be described in detail several specific embodiments with the understanding that the present disclosure is to be considered as an exemplification of the principles of the technology and is not intended to limit the technology to the embodiments illustrated.
- It will be understood that like or analogous elements and/or components, referred to herein, may be identified throughout the drawings with like reference characters. It will be further understood that several of the figures are merely schematic representations of the present technology. As such, some of the components may have been distorted from their actual scale for pictorial clarity.
- Generally speaking, the present technology comprises systems, methods, and media for processing natural language travel requests. More specifically, but not by limitation, the present technology may fulfill travel requests in the form of natural language expressions of a travel itinerary. An artificial intelligence engine used in methods and systems of the present disclosure may act as a priori artificial intelligence engine (i.e., may assume an a priori knowledge of certain structures). The present technology provides an efficient and simplified supply chain for the addition, organization, and consumption of inventory, together with a simplified distribution model. Additionally, the systems provided herein may also interact seamlessly with, and coexist with, legacy systems.
- Advantageously, the present technology provides increased efficiency and capabilities, allowing access to greater amounts of content that may be utilized to fulfill natural language travel requests. Unlike most systems or search engines, where a URL is provided as a solution or a few thousand options for a single request or a component of a request, the preset technology provides coherent solution(s) for natural language travel requests.
- Additionally, the present technology may be implemented within the context of an exchange system that allows suppliers to provide inventory records and customers to input travel itinerary requests in a natural language format and fulfills the travel itinerary requests by applying pattern recognition artificial intelligence and/or semantic parsing to inventory records and travel itinerary requests to obtain matches therebetween.
- Referring to the collective drawings (e.g.,
FIGS. 1-11 ), the present technology may facilitate an exchange that fulfills natural language travel requests. The present technology may be implemented within the context of anexemplary architecture 100, hereinafter “architecture 100” as shown inFIG. 1 . Thearchitecture 100 may be described as generally including an exchange 105 (also referred to herein as exchange system 105).Consumers 110 andthird party suppliers 115 may communicate with either theexchange 105, via anetwork 120. It is noteworthy to mention that thenetwork 120 may include any one (or combination) of private or public communications networks such as the Internet. Theconsumers 110 may interact with theexchange 105 via end user client devices that access a web based interface or an application resident on the end user client device. - In some embodiments, the
third party suppliers 115 may communicatively couple with theexchange 105 over thenetwork 120 via an application programming interface (API). It is noteworthy that other methods/systems that allow thethird party suppliers 115 and theexchange 105 to communicatively couple with one another, that would be known to one or ordinary skill in the art, are likewise contemplated for use in accordance with the present disclosure. - For the purposes of brevity and clarity, certain functional and/or structural aspects of the
exchange 105 will be described in greater detail herein. More specifically, but not by way of limitation, the present disclosure will address the processing and fulfillment of natural language travel requests. Additional details regarding theexchange 105 may be found in co-pending U.S. non-provisional patent application Ser. No. 13/420,433, filed Mar. 14, 2012 and issued Sep. 18, 2018, as U.S. Pat. No. 10,078,855, which is hereby incorporated by reference herein in its entirety. - According to some embodiments, the
exchange 105 may include a cloud based computing environment. In general, a cloud-based computing environment is a resource that typically combines the computational power of a large grouping of processors and/or that combines the storage capacity of a large grouping of computer memories or storage devices. For example, systems that provide a cloud resource may be utilized exclusively by their owners, such as Google™ or Yahoo!™; or such systems may be accessible to outside users who deploy applications within the computing infrastructure to obtain the benefit of large computational or storage resources. - The cloud may be formed, for example, by a network of web servers, with each web server (or at least a plurality thereof) providing processor and/or storage resources. These servers may manage workloads provided by multiple users (e.g., cloud resource consumers or other users). Typically, each user places workload demands upon the cloud that vary in real-time, sometimes dramatically. The nature and extent of these variations typically depend on the type of business associated with the user.
- The
exchange 105 may be generally described as a particular purpose computing environment that includes executable instructions that are configured to receive and fulfill natural language requests, such as travel itinerary requests. - In some embodiments, the
exchange 105 may include executable instructions in the form of an itinerary processing and fulfillment application, hereinafter referred to as “application 200,” “a system for fulfilling travel requests,” or “a system.” The application provides various functionalities that will be described in greater detail herein.FIG. 2 illustrates and exemplary schematic diagram of theapplication 200. - The
application 200 is shown as generally comprising components such as a semantic parsing module, hereinafter referred to “a parser” or “parsingmodule 205,” a pattern recognition artificial intelligence engine, hereinafter “AI engine 210,” a scheduling module 215 (also referred to herein as scheduling module 215), and amodification module 220. It is noteworthy that theapplication 200 may include additional modules, engines, or components, and still fall within the scope of the present technology. As used herein, the terms “module” and “engine” may also refer to any of an application-specific integrated circuit (ASIC), an electronic circuit, a processor (shared, dedicated, or group) that executes one or more software or firmware programs, a combinational logic circuit, and/or other suitable components that provide the described functionality. In other embodiments, individual components of theapplication 200 may include separately configured web servers. -
FIG. 3 includes an exemplary flow diagram that illustrates the flow of data from a publishing environment into an exchange, along with the receipt of natural language travel requests and their fulfillment. While functional details regarding how theexchange 105 processes and fulfills natural language travel requests will be described with reference to additional figures described below (e.g.,FIGS. 4-6 ), the overall operational flow of theexchange system 105 is shown inFIG. 3 . - Referring now to
FIGS. 2 and 4 collectively, thescheduler module 215 may utilize theparsing module 205 to interpret the natural language queries.FIG. 4 illustrates a flowchart of an exemplary method for processing natural language travel requests. - According to some embodiments, the
parsing module 205 may assume an a priori knowledge of certain structures and intent over a class of information (for example, the hospitality and travel space). - Initially, it is noteworthy to mention that the natural language travel requests received by the
parsing module 205 may comprise a textual request, a spoken (e.g., audio format) request, a location based request, an input based request (e.g., a click of an object on a map), a global positioning signal, and/or any combinations thereof. Moreover, in some instances, the request may comprise a non-natural language request, such as a keyword request, a Boolean phrase, and so forth. - In this sense, the information requested by the end user in natural language may not be parsed by the
parsing module 205 for grammar in the sense that a normal parser would operate. Rather, theparsing module 205 may infer a pre-determined set of information through a pattern recognition artificial intelligence module, such as theAI engine 210. - More specifically, the
parsing module 205 may first (Step 405) delimit the natural language query. For example, theparsing module 205 may determine inventory components in the query. - The
parsing module 205 may parse through each delimited string (Step 410) and transmit the delimited strings to theAI engine 210. - The
AI engine 210 may act as an a priori AI engine (i.e., may assume an a priori knowledge of certain structures). While the a priori AI engine is discussed herein in terms of its application to travel, a person of ordinary skill in the art would understand that the engine can be similarly utilized in any domain. - The a priori AI engine does not use evidentiary data required for the formulation of the AI. The knowledge and understanding have to exist before any data is provided. This is in stark contrast to current AI engines in the market, which use a posteriori AI that is more commonly prevalent and is entirely dependent on data and experience.
- The AI required for the a priori AI engine cannot be ascertained from the interaction between users and the system. Much of the intelligence is knowledge and understanding that is not part of this interaction but is contained in the structures and morphology behind the interface. The totality of the understanding of requests by the a priori AI engine is not contained within the conversation with a user.
- The a priori AI engine cannot be reduced to a set of rules. This may make the system entirely deterministic, which it is not, and may require the system to consider so many cases that it can be practically infinite in size.
- Additionally, in a rule-based system, a rule must be given for every case. In any non-trivial system, it is not possible to cover every potential scenario that may occur, as per Gödel's incompleteness theorem. In the a priori AI engine, the AI can ascertain the rule, intention, and action for all cases.
- It is asserted that concepts are known and understood first and that language evolves to describe these concepts. Consequently, we do not begin with language at all, but rather, we imbue the system with an understanding of travel logistics.
- The a priori AI engine may essentially deal with the conversation itself, not just the content of the conversation. Thus, each conversation initiated by either the user or the system may have a number of essential parts. Specifically, the conversation may have an objective or goal. For example, the objective may be the building or modification of an itinerary. The objective may be a response from the user to a question posed by the system to the user dealing with errors, or the like. Furthermore, each of the conversations may have a state, such as awaiting answer, completed, and the like. As would be understood by a person of ordinary skill in the art, many types of conversations can exist at the same time for a given user.
- The
AI engine 210 may employ a combination of phraseology and keyword inference (Step 415) to decode what type of request is being made. Specifically, theAI engine 210 may determine the itinerary components by decoding the itinerary components from the travel request by determining phrases and tokens of the travel request. TheAI engine 210 may reference the metadata database and the equivalence class database. Keywords included in an AI pattern recognition database may direct theAI engine 210 to appropriate content categories for the itinerary components included in the request (Step 420). TheAI engine 210 may employ additional inferential methods as well as statistical methods and frequency to determine where and how to match content to the request. - The
parsing module 205 may evaluate each word of the sentence. If no keywords are found, nothing is constructed. However, theAI engine 210 may employ a “similar to” inference functionality which allows for variation among the phraseology to account for different ways that natural language queries may be structured such as incorrect spelling, grammar, and similar contingencies. - Once the
parsing module 205 has determined the itinerary components included in the natural language travel request, theparsing module 205 determines a node type for each of the itinerary components and ascertain dependencies between each of the itinerary components based upon respective node types. It will be understood that the parsing module may effectuate construction of itineraries in a variety of manners. For example, theparsing module 205 may parse the words of the request in a sequential manner. Theparsing module 205 may also parse the request to determine categories of itinerary components included in the request. In other instances, theparsing module 205 may delimit the request. - The method for fulfilling travel requests of the present disclosure utilizes an itinerary network graph with nodes and dependencies. Specifically, according to some embodiments, the
parsing module 205 may utilize a directed acyclic graph (DAG), also referred to as an “itinerary network,” to interpret natural language queries. The information extracted by theparsing module 205 may be utilized to generate an itinerary network that provides a further dynamic intelligence to theparsing module 205 in understanding the requested, parsed information, and assist theparsing module 205 in determining possible logical and logistics connections (e.g., location, time, and traveler preference based dependencies). Therefore, theparsing module 205 may generate the itinerary network based on the itinerary components of the travel request. Specifically, theparsing module 205 may create a plurality of nodes within the itinerary network. Each of the plurality of nodes may represent information associated with the travel request. Theparsing module 205 may further create a plurality of edges within the itinerary network. Each of the plurality of edges may connect two of the plurality of nodes. The plurality of edges may represent an order of the plurality of nodes in time based on dependencies between the plurality of nodes. - The
parsing module 205 may generate a travel itinerary responsive to the travel request. The travel itinerary may be consistent with the itinerary network. Theparsing module 205 may present the generated travel itinerary to the user on a user interface of a computing device associated with the user. - The nodes may be of different types and may store information about the itinerary. The nodes may represent cities, hotels, flights, and destination content (such as a concert or car rental). The nodes can contain information specific to the type of node. For instance, a city node may contain an airport, whereas a flight node may contain the class of service.
- Edges connect nodes to show dependencies between nodes and may contain information themselves. The edges are directed to represent the order of nodes in time. In one example embodiment, edges can be air transport, car transport, start to start, or start to finish. In other words, the edges can present information that may be important when scheduling the itinerary.
- If a user expresses a desire to travel between two cities, an itinerary network can be used to represent this travel request of the user. To create the itinerary network, two city nodes may be created first: one for the source city and one for the destination city expressed by the user. The city nodes may be populated with the names of the cities in the user request and possibly names of airports. Then, these cities may be connected with a directed edge (dependency). The edge may contain information for a flight, rail, bus, or other mode of transport between the two cities, as well as other information, such as the scheduling type. In some cases, for example, in multi-passenger scenarios and situations where dependencies between passengers arise, the mode of transport may be presented as the node type and may be further qualified in the dependencies between the city and the transport node.
- If the same user expresses a desire to travel between two city pairs, this process may be done twice. It may be incorrect logistically and logically to put one person on two flights departing at the same time. In view of this, the first condition is that each passenger must have a path-connected graph. The second condition is that the graph is acyclic. The a priori AI engine understands that the itinerary network is a compact path connected topological space, with all the mathematical tools and theorems that follow from that at disposal of the a priori AI engine. This allows the AI to fully understand the properties of the itinerary network and to know if the path connectedness property is broken and how to resolve it by communication with the user. This understanding is naturally inherent within the systems and the methods of the present disclosure and it has been built into the a priori AI engine.
- The itinerary network provides a coherent consistent entity that can then be treated by a scheduler for scheduling of a travel request. More than this, the systems and the methods of the present disclosure cover every case one could glean from data or experience, without having to input a multitude of individual rules to teach the a priori AI engine.
- The resultant itinerary network contains all of the information necessary for the system to supply congruent content (logistic and curated data). The characteristics of the itinerary are the characteristics of a path-connected, directed acyclic, compact topological space or combinations of such spaces. This provides the capability of scheduling the itinerary in a consistent and coherent manner. It also allows the parser to understand all the aspects of the itinerary.
- The itinerary network may describe the itinerary, such as flights, hotels, cars, events, reservations, and other disparate content. The itinerary network also describes the temporal and dependent relationships of the passengers and their content. The understanding of the itinerary depends on the scheduler, which instantaneously understands the logistics and content of the network. The parser may be imbued with the lexicon and capabilities of the itinerary network and as such can understand and interact with the itinerary network independently or through conversations with the user.
- All of these described elements are part of the a priori AI engine directed to not simply understanding language. In fact, the language is an insufficient condition to the AI of the system. The a priori AI engine may inform what language is to be understood and what language is not relevant.
- Goal Orientation—Seeking. The a priori AI engine may use the parser, the scheduler, and so forth, to seek to fulfill an itinerary request. For this purpose, the a priori AI engine may build an itinerary network and schedule the itinerary network with appropriate curated content. The parser may use the tree of nodes (parser trees) to address any conversation that seeks to instruct the system to build or modify any itinerary.
- In some instances, itinerary components may comprise travel or non-travel node types. For travel node types, the
parsing module 205 may obtain source and destination information from relevant itinerary components (Steps 425 and 430). If they do not exist on the itinerary network, theparsing module 205 may add them to the itinerary network. For non-travel nodes, theparsing module 205 may determine if the node has a time or location dependency to another node (Step 435). If the node does have a dependency, theparsing module 205 checks to see if the dependent node exists. If it does not, theparsing module 205 will create the node and populate the node with any necessary attributes (Step 440). - According to some embodiments, the parsing module may also identify traveler preferences. Traveler preferences can include general or specific preferences and are requested or ordered in natural language. Some examples include: “give me cheapest flight,” “do not book me into any Hilton hotels,” “provide me four-star hotels or better,” and “If I am in San Francisco book me into the San Mateo Sofitel hotel.”
- The process of identifying nodes for itinerary components and interrelating these nodes may be referred to as generating an itinerary network. The itinerary network may be utilized by the
scheduler module 215 to generate an unconstrained schedule for the natural language request, as will be described in greater detail herein. - The characteristics of an itinerary may include a stage, a state, and a context. These characteristics allow the parser to be sentient in the conversation with the user through the life cycle of the itinerary (pre- to post-booking and change management) and partial completion of the itinerary. The sentience includes understanding the meaning of the request (based on from context, stage and state). The goal orientation of the system may be to serve an itinerary request. The stage may include a pre-booking state and a post-booking stage. The stage is important for establishing the context, as well as for communicating relevant information to a user, such as change and cancellation fees or other disruptions to the itinerary as a result of a change or delay.
- The state is broadly defined by two elements, namely the itinerary network and the conversation. The state includes determining, for example, whether the system is conversing with the user about the itinerary, whether the user has booked a trip, whether there are flags on the itinerary that need to be addressed (such as incomplete information or infeasible solution), and so forth.
- Context may be built into an itinerary object. This may provide sentience to the AI engine as the AI engine interprets new requests or changes for the itinerary. The context may be dynamic as the itinerary object changes its stage, state and topology. As one can easily see, traditional machine learned artificial intelligence simply cannot cope with dynamic context because the number of experiences is an NP-Complete decision problem, even a higher order of complexity than NP-Hard decision problem.
- It should be noted that all of the machine learning/neural network type processes around natural language by their construction must infer the semantic meaning from language data that is input into the neural network and trained on the semantic meaning. In “The Language Complexity Game (Artificial Intelligence)” by Eric Ristad, The MIT Press; First Edition (Mar. 17, 1993), it is argued that language is the process of constructing linguistic representations from the forms produced by other cognitive modules and that this process is NP-complete. However, neural networks and other statistical-based inference models as a technology do not rise to the complexity class required to solve NP-complete problems. It is undeniable that they can make inferences with some accuracy, e.g., when being shown one million examples of a
number 5, the neural networks and other statistical-based inference models will pick a 5 out 99% of the time correctly when shown numbers. When noise is introduced into the neural networks, the neural networks may still infer, with the same confidence, that a static page before a user is anumber 2. When being asked by a user to draw a 5, the neural networks cannot answer the user. Thus, the structure of conventional neural networks is incapable of doing anything other than returning a probability. - The a priori AI engine described herein changes the problems complexity class from NP-complete to order kN. The a priori AI engine is imbued with knowledge of travel, and language does not need to be inferred, language needs only be matched as the set of language patterns are generated by the capabilities and understandings of the underlying morphology/referential framework. In this sense, the behavior of the a priori AI engine is totally unique in the ecosystem of natural language human machine interfaces.
- It will be understood that the
parsing module 205 may generate an itinerary network in any order, allowing itinerary components to be inserted into the itinerary network when a starting/ending reference point has been established, such as when the source and destination itinerary components are identified. Anexemplary itinerary network 500 is illustrated inFIG. 5 , and is constructed from the natural language travel request, “From Toronto to Seattle. From Seattle to Tokyo. Stay at any preferred hotel with a buckwheat pillow. Reservations at Daniel's Broiler and a well-known sushi restaurant near my hotel in Tokyo.” - Additionally, the following traveler preferences that were received in natural language format include: “Give me lowest cost tickets,” “Exclude Hilton chain,” “Route me through Cincinnati on route to Seattle,” “Integrate my calendar and exclude red category,” as well as many other traveler preferences, which would be known to one of ordinary skill in the art with the present disclosure before them.
- Additionally, the
parsing module 205 may populate each itinerary component with attributes identified by theAI engine 210, such as node type and dependencies. - The
parsing module 205 may then establish dependencies between appropriate itinerary components. There is an extended set of dependencies that extend from the normal start-start, start-finish, finish-start, and finish-finish to parent-child, local dependency, and so forth. Other exemplary dependencies may include, but are not limited to: Air-Connect, Local-Connect, Activity, Location, Time, Time and Location, Logical-Connect, and dependencies that relate to the travel data of another traveler such as “Travel Together” and “Travel Meet At.” - Time dependencies may be utilized to generate itinerary schedules in reverse order, based upon an end point. For example, using a scheduled meeting as an end point, the present technology may create and fulfill a travel itinerary for a customer that ensures that the customer arrives in the proper location and at the proper point in time to allow them to attend the scheduled meeting.
- Once node types and dependencies have been established for the itinerary components of the natural language request, the
parsing module 205 may generate an adjacency matrix using the itinerary components and their respective dependencies. Utilizing the adjacency matrix, the parsing module may create an itinerary network using the adjacency matrix. - Next, the
parsing module 205 may determine a topological ordering of itinerary components using the itinerary network. It is noteworthy that the topological ordering of itinerary components may comprise an arrangement of the itinerary components using their respective location and time dependencies used by thescheduling module 215 to generate an unconstrained schedule, as will be discussed in greater detail below. - Conceptually, the
parsing module 205 andAI engine 210 may utilize the itinerary network to inform thescheduling module 215 in generating schedules and allocating inventory to the schedules. For example, if an itinerary node includes an activity, or location dependent node such as a theater, restaurant, hotel, conference, or the like, theparsing module 205 will understand the activity must take place in a city. Thus, depending on the phraseology encountered by theAI engine 210, theAI engine 210 may loop through the admissible ways of saying “I'm here” and compare the location against a city dictionary list. If the city is valid, theAI engine 210 may look for the city name in the itinerary network, creating a node if theAI engine 210 does not find an appropriate node or adding the activity node with a time/location dependency underneath. - Dependent activities may have their own dependencies as well (for example, local transportation between a restaurant and a conference). Moreover, preferences associated with each dependent node may appear as another level of dependency (for example, a buckwheat pillow in a hotel room).
- At each level, the
parsing module 205 may check to see if a desired node is present in the itinerary network and create nodes as needed. Since each city and activity has a time dependency as well as a location dependency, in complex itineraries with multiple cities being visited multiple times by multiple people, theparsing module 205 may prevent confusion relative to a dependent node's dependencies relative to location and time. Theparsing module 205 may also inform the consumer that he has asked for a hotel in a city to which the consumer is not traveling. - If the
parsing module 205 determines a travel phrase or keyword, theparsing module 205 may infer there must be a source and destination and a mode of travel therebetween. Theparsing module 205 may further infer what kind of travel is most appropriate, so a consumer will not find himself driving or taking the train from Miami to Manchester, U.K. - The
parsing module 205 may not dictate a mode of travel; however, a consumer may choose to take any form of transportation desired. Theparsing module 205 may send the phrase to theAI engine 210, extract the source and destination cities, match them against the city list dictionary, and check the network for the nodes existence and add them if necessary. TheAI engine 210 may then add the travel node and a travel dependency between the travel node and the two cities to the itinerary network. - Therefore, a consumer may ask for any itinerary, in any order, and the present technology may produce a correctly networked schedule. For example, the present technology may take the natural language phrase, “I want to go from Seattle to Dallas, Miami to Atlanta, Dallas to Miami, Toronto to Seattle.” The
parsing module 205 may create an itinerary network which linked Toronto to Seattle to Dallas to Miami to Atlanta. As before, additional content nodes and dependencies may be added as required. - The
parsing module 205 may understand the different types of dependencies that occur. For instance, in Toronto there may be an Italian restaurant called Pizza Banfi. If a traveler preference indicates a hometown of Toronto, or location-based data from a consumers' cellphone indicates that the consumer is in Toronto, and consumer requests “From Pizza Banfi to Seattle,” theAI engine 210 may understand that the consumer requires transport between two points, but that one point is a city, and the other is a dependent node belonging to another city. TheAI engine 210 may create the Toronto node, place the restaurant as a dependent node, arrange for transport to the airport, which is local dependency, a flight dependency between the two cities right after it creates the Seattle node. - The
scheduling module 215 may be executed to generate an unconstrained schedule from the itinerary network (e.g., DAG). - The generation of an unconstrained schedule establishes the earliest start and latest finish for all nodes and hence the initial starting point for all requests pertinent to the content represented by the nodes. The
scheduling module 215 then employs one of several methods to resolve the allocation of content (e.g., inventory) to the requests for content and fill the itinerary. - The
scheduling module 215 may apply an Adaptive Method that “levels” the itinerary. For example, thescheduling module 215 may search content within the topological ordering. Each line item in the topology may be considered, the exchange searched, and/or offers obtained from the suppliers. The content request is established by thescheduling module 215 from the node type and its attributes as filled out by theparsing module 205. These attributes also include general and specific preferences. A set of valid options may be obtained and ordered by the traveler preferences. - Further, the
scheduling module 215 may employ additional methods to allocate inventory to the request. In a “best alternative” mode, a best alternative (e.g., available inventory) is selected that comprises the content selection that is at the top of the list sorted by traveler preferences This then sets the starting conditions for successor nodes in the topology and the topology is then recursed by thescheduling module 215 using only the best client alternatives. In some instances, a specific best path itinerary can be identified. - Additionally, the itinerary can be optimized with respect to an equivalence class of airline tickets, where the result from selecting a specific airline ticket does not impact the remainder of the itinerary.
- In an “all possible” mode, each alternative (up to some arbitrary limit) of the sorted list of nodes by client preferences may be considered by the
scheduling module 215 and a separate itinerary developed for each. Thescheduling module 215 processes each line item in the topology by applying a recursion algorithm. - The results of this modal process may generate many different itineraries whose costs and time frames can vary substantially. These itineraries may be sorted in different ways using multiple sorting criteria; (shortest, lowest cost); (lowest cost, shortest); and so forth. The
scheduling module 215 can dynamically schedule robustness into the schedule in the sense that it can maintain specific times required between flights; these can be in minutes, hours or days. The scheduler will automatically extend hotel stays if the flights do not leave on the same day as the hotel checkout. - The
scheduling module 215 may create time and space dependent solutions to the logical schedule dynamically, based on the offers made to the requested itinerary from suppliers. Thescheduling module 215 maintains the dependencies so that requests remain accurate with respect to the current solution. In this manner, the logistics of travel are maintained and their constraints adhered to. - The
scheduling module 215 may be configured to always return a solution, even if the constraints cannot be met. This solution may comprise the closest available under the constraints and options that have been requested. It is noteworthy that when inventories for content are tight, it could take an extremely long time to find any solution. Therefore an “approximate fit” schedule may be preferred to no schedule. - The
scheduling module 215 may be configured to generate a leveled solution where thescheduling module 215 may allow requests to level out in time across the itinerary, showing when solutions are available. Thus, if a customer books a flight today to San Francisco, thescheduling module 215 may allow a solution for tomorrow if that is the only alternative. - The
scheduling module 215 may also provide one or more possible schedules (solutions) to the exchange 105 (FIG. 1 ) in either a sequential or leveled manner. In the sequential method, all dependencies for a specific aspect of the itinerary may be filled before it is submitted to theexchange 105. An alternative method allows thescheduling module 215 to maintain the times and dates specified, and only offers that match these times and dates are allowed. - Referring now to
FIGS. 2 and 6 collectively, thescheduling module 215 may transmit itinerary components and/or entire itinerary schedules (such as the itinerary network) to theexchange 105. Theexchange 105 may employ a listener that immediately picks up the new requested line item or itinerary (Step 605). Line items or an itinerary may also be referred to as a “request.” The listener may identify the itinerary components (nodes) (Step 610) and/or a buyer profile or content profile associated with the itinerary (Step 615). The listener may compose a list of the suppliers that have indicated that they want to bid on these types of line items or itineraries (Step 620). Suppliers may be notified of these requests and can then analyze them and bid on them or entire itinerary (Step 625). The transactions made available to the suppliers contain the entire content, inferential information, and/or semantics of the request together with a framework for interpreting the same. The supplier can respond based on review of its inventory and availability. In other instances, the supplier can dynamically decide what to do with the content and price through its own legacy systems. Alternatively, theexchange 105 makes available APIs to interrogate the system for any requests that the supplier may want to look at (For example, City-Pair for flights and/or Activity Keyword or partial Keyword). In some embodiments, the default listener is the exchange itself that will process, search, and respond to every request. - Offers may be written back to the exchange in the form of a response. Additionally, suppliers can respond with any additional content they desire, together with pricing for itinerary components. For example, an airline can offer a golf bag at $100 with the air ticket at a reduced price. Other similar types of vouchers may be exchanged or facilitated utilizing the present technology.
- As offers are written to the
exchange 105, they are matched against the line items and itinerary generated by thescheduling module 215. In some instances, before being considered, the offers may be passed through a set of filters that describe the traveler's restrictions and preferences. An exemplary flow diagram of aprocess 600 for fulfilling a schedule (e.g., request) is depicted inFIG. 6 . - According to some embodiments, the
scheduling module 215 may selectively adjust the allocation of inventory based upon various constraints such as available/dynamic inventory. In other embodiments, thescheduling module 215 may adjust the schedule provided to the consumer based upon inferential modeling of the consumer's request (for example, when the consumer expresses a traveler preference that is new or contradictory to a known traveler preference for that particular consumer). - According to some embodiments, the
modification module 220 may be executed to process modifications to travel itineraries. Generally speaking, themodification module 220 may receive a modification to the travel itinerary from a traveler who has previously input a natural language travel request that has been processed using the aforementioned methods to generate an itinerary schedule. - The
modification module 220 may adjust the allocation of available inventory for each itinerary component remaining in the travel itinerary based upon one or more dependency adjustments associated with the plurality of nodes and the plurality of edges and caused by modification of the travel itinerary. That is, because theparsing module 205 appreciates the dependencies between the current itinerary components in the schedule, along with the dependencies of the modification, theparsing module 205 may insert the modification into the schedule and adjust other itinerary components, as necessary. Therefore, even for an itinerary that is currently being executed (e.g., traveler has already completed at least a portion of their itinerary), theparsing module 205 may adjust the schedule to ensure that traveler preferences are maintained. For example, if cost is an important traveler preference, theparsing module 205 may adjust the schedule to cause the least impact from a cost perspective. - The parser trees may define a priori AI engine that can deal with any conversation seeking to build or change itineraries through successive iterations of the conversation—all in place before the first ever conversation takes place. Furthermore, no rules are required; rather, the understanding is fundamental and complete in advance of the conversations or requests with users.
- The itinerary network may handle the language interface between the user and the system. The itinerary network also facilitate interactions between various elements of the system itself. The network nodes may be comprised of tokens. The paths (sequence of tokens) in the itinerary network are called ‘phrases’. Together, the tokens and the phrases may make a normalized travel language or meta language. This is strictly a pattern (or phrase) recognition process; no traditional linguistic grammars are used in this process. This is a departure from traditional Natural Language Processing methods which rely heavily on the linguistic syntactic composition of a sentence and data to learn relationships between the various syntactic elements. The system is able to construct the phrases without ever having to have experienced them.
- Tokens are the categorization of information related to travel. For example, the tokens may include travel attributes related to the travel request, such as passengers, source and destination cities, times and dates, time and date restrictions, hotel names, addresses, and other descriptive elements, such as dependencies between the various itinerary network elements. A phrase is an ordered set (or sequence) of tokens that describes an intention, request, or like communication with the system (parser). In effect, the entire conversation related to the itinerary may be pre-defined in this manner. All travel conversations and their intent, regardless of context, and the like, can be reduced to a normalized set of phrases in the parser trees.
-
FIG. 7 shows anitinerary network 700, according to an example embodiment. Theitinerary network 700 may includenodes 705 and edges 710.FIG. 8 shows a sub-path of nodes of anexample itinerary network 800. The nodes may have annotations. Anode 805 may have anannotation 810, and anode 815 may have anannotation 820. Each node in theitinerary network more annotations annotations node 805 may have the annotation 810 ‘flight’. This annotation may mean that there is sufficient information (within the path or network subtree reaching the node 805) to determine that there is an intentional request for a flight. The AI engine understands this information together with the data associated with the flight. Another sub-path may add additional information to an existing node or define a global requirement or preference or a change to the existing itinerary network. As these tokens from the conversation are processed (for this example, in sequence), the sequence may be matched against a sub-path in the parser tree. - The system may be configured to perform error handling and the ability to anticipate “next” likely token(s). Specifically, at any stage in matching the ordered set of tokens obtained from the request, the parser already knows the set of next possible valid tokens of the travel request. This is significant because the parser can detect insufficient information to fully understand a request, as well as missing information (tokens) necessary for generating the travel itinerary. Based on the possible next tokens, the system may detect insufficient information associated with the travel request and request the insufficient information from the user.
- An example of building a set of phrases using the parser trees. Once a match for a sub-path with an annotation is obtained, there is created an immediate and absolute “understanding” and a potential actionable event. The nodes in the sub-path may not have any physical resemblance to what was understood or inferred or the actionability thereof. This constitutes an a priori AI engine. The a priori AI engine has understanding before there is any specific request and is general in nature, and the a priori AI engine requires no data for this understanding; rather, it can handle any such request associated with the sub-path. Indeed, the capabilities of the itinerary network and scheduler define the phrases in their totality. Additional phrases, when encountered, merely constitute an equivalence class relationship with an already extant phrase.
- For example, as shown on
FIG. 8 , at the stage B, thenode 805 may have the attribution of a flight that means that the parser has intentionality to define a flight. At stage C, thenode 815 may have the attribution of a flight that means that the parser has intentionality to define a hotel. The system understands that what is requested is a flight and a hotel. The set of nodes leading up to this provides sufficient and necessary logical information to create a flight. Now an actionable phrase may start. This phrase invokes an action, for example, called “Build Flight and Hotel.” This action builds a travel itinerary and the itinerary network, which represents the travel itinerary in a unique manner. To build the itinerary network, the system determines content nodes, city nodes, city children content, the relationships between the nodes, the dependencies between nodes, the ability to represent multiple travelers, and so forth. - The Parser Complexity Class. The cardinality of the set of equivalence classes is finite, countable, and has N members. This set may form a compact cover of the set of actions, requests, information, and so forth defined by the system. The equivalence classes, which provide equivalent ways of stating any phrase, form a larger
open cover 905 as shown onFIG. 9 and discussed below. The complexity class of the a priori engine may be <O(kN), where N is the finite number of equivalence classes (i.e., the finite number of statements about the travel itinerary possible to be made) and k is the maximum number of the equivalence classes of the statements. This holds true regardless of how many times phrases are matched in the parser trees or the size of the context in which they are made. -
FIG. 9 shows aparser network 900 as a cover for travel conversations, according to an example embodiment. While the normalized set of phrases form acompact cover 910 of the intentionality space for travel, each phrase can have a set of like phrases that are equivalent to the single member phrase of the compact set and form anopen cover 905. For example, “I want to go from YYZ to MIA” is equivalent to “I want to go to MIA from YYZ”. While the set of phrases is O(N) complexity class, thecompact cover 910 together with the equivalence classes form a complexity class of O(kN). This is extremely important since most other AI technologies are essentially NP-complete. - Building the Itinerary Network. The natural language request of the user may have the following form: “I want to fly from Toronto to New York staying 3 days upper west side in a 5 star hotel then go to Miami staying in South Beach for 2 nights then to San Francisco staying at the Sofitel on Twin Dolphin drive for 2 nights then home. Business class tickets on all flights. Jonathan will join me in San Francisco stay at the same hotel and we will share a room. He will then go to Seattle stay at the Woodmark Hotel in Kirkland for 3 days then fly home.” The request may be processed into meta patterns (phrases) and these phrases may be used to build a directed, acyclic, path connected network, which in turn may be used to build a resource independent (logical and resource unconstrained) schedule.
- The goal of the system includes finding suitable content for this itinerary as defined by the itinerary network. The itinerary network may provide the context both when constructing the itinerary network for requests from the conversation, as well as when creating the goal for the system, namely, to fulfill this travel itinerary network. In this manner, travel requests and successive requests may define the itinerary network and the system may always act to fulfill the itinerary network, whether by creating a new itinerary network or changing an existing itinerary network. This iterative process can continue indefinitely without human intervention. The system may be automatically goal seeking and the AI engine may be synthetic and analytic a priori (i.e., no data and/or experience may be required for the AI engine).
- An
example itinerary network 1000 is shown inFIG. 10 . Theitinerary network 1000 may have a plurality ofnodes edges - Properties of the Itinerary Network. The AI engine may consequently deal with multiple states that occur over the lifespan of the travel itinerary. The AI engine also understands the properties of an itinerary object. For example, the AI engine may understand that the itinerary object is a compact path connected topological space. When the path connectedness property is broken, the AI engine understands the implications of this and how to deal with it by communicating with the user. This understanding is naturally inherent to the AI engine and the methods of the present disclosure.
- The itinerary network may provide a coherent consistent entity that can be treated by the scheduler with a single set of methods and provide a scheduling AI ecosystem for the scheduling of the travel request. The AI engine is aware of the context of the travel itinerary, namely a stage, state, and all the methods and attributes the itinerary network represents. The resultant itinerary network contains all the information necessary for the system to supply congruent content (logistic and curated data). The characteristics of the itinerary are that of a path-connected, directed acyclic, compact topological space, or combinations of such spaces. This provides the capability of scheduling the itinerary in a coherent manner and allows the parser to understand all the aspects of an itinerary with respect to nodes of the itinerary network.
- The Construction of Parser Network Trees. Normally, one considers natural language as the mechanism to interface with a user and the machine. Natural language parsing is complicated and requires correct linguistic grammar to interpret the request. This is difficult enough and subject to the ambiguities of language, as well as because people do not often speak grammatically correctly. Regardless of the ability for the natural language parsing to parse a request into a grammatical structure (subject, objects, verb, clauses, etc.), the ability for the system to actually understand the context and intent of the spoken word is complex. Traditional methods for AI use language through inference and large amounts of data (experiential) to determine what the intent actually is. Complex statistical methods in the end just give probabilities that the spoken sentence means a specific thing. For example, in conventional natural language parsing systems, in response to “I want to fly to Montreal next Tuesday” the conventional natural language parsing system replies, “There is a 94% likelihood that you want to take a flight to Montreal next Tuesday.” Furthermore, no context of the conversation is determined by the conventional natural language parsing system.
- Building a priori parser tree(s). The parsing may be fully integrated into the system, methods, and all the conversations that exist between the modules of the system, the state, and stage of the itinerary, the error conditions together with the conversations that may exist between the user and the system.
- The a priori engine may be, in fact, built in reverse order to the traditional model. All the goal seeking methods may be defined, given context in terms of the itinerary object, and then the specific phrases (meta language) may be defined in these terms. The a priori engine may include the itinerary network, the parser network and the phrases and the lexicon required to drive them. The set of phrases may be a compact cover of the truthful statements that can be made within the tautology of the a priori engine. The AI engine can be easily extended to cover new methods, lexicon, and phrases, without experimentation or any data sources.
- Since the AI engine has to be constructed from known structures and methods, the AI engine cannot be learned via experience or data model. The a priori engine may comprise a set of phrases which are understood explicitly in terms of their respective goals/objectives, information content, queries, and the like. The phrases are understood in context of the current state of the itinerary, supply chain, and so forth.
- Embodiments of a computing system discussed herein include providing end to end digital service, where the AI engine extends through understanding of a request from a user, the goal seeking properties of constructing a congruent itinerary, modifying the itinerary, curating the content for each user, maintaining context and sentience, and seeking to complete goals.
- The AI engine discussed herein can be implemented via a specially-programmed and special-purpose computer. The computing engine may utilize one or more processors storing instructions, static and/or dynamic memory, one or more databases or other data structures, and network interface(s).
-
FIG. 11 illustrates anexemplary computing system 1100 that may be used to implement an embodiment of the present technology. Thesystem 1100 ofFIG. 11 may be implemented in the contexts of the likes of computing systems, networks, exchanges, servers, or combinations thereof disclosed herein. Thecomputing system 1100 ofFIG. 11 includes one ormore processors 1110 andmain memory 1120.Main memory 1120 stores, in part, instructions and data for execution byprocessor 1110.Main memory 1120 may store the executable code when in operation. Thesystem 1100 ofFIG. 11 further includes amass storage device 1130, portable storage medium drive(s) 1140,output devices 1150,user input devices 1160, agraphics display 1170, andperipheral devices 1180. - The components shown in
FIG. 11 are depicted as being connected via asingle bus 1190. The components may be connected through one or more data transport means.Processor unit 1110 andmain memory 1120 may be connected via a local microprocessor bus, and themass storage device 1130, peripheral device(s) 1180,portable storage device 1140, anddisplay system 1170 may be connected via one or more input/output (I/O) buses. -
Mass storage device 1130, which may be implemented with a magnetic disk drive or an optical disk drive, is a non-volatile storage device for storing data and instructions for use byprocessor unit 1110.Mass storage device 1130 may store the system software for implementing embodiments of the present technology for purposes of loading that software intomain memory 1120. -
Portable storage device 1140 operates in conjunction with a portable non-volatile storage medium, such as a floppy disk, compact disk (CD), digital video disc (DVD), or USB storage device, to input and output data and code to and from thecomputer system 1100 ofFIG. 11 . The system software for implementing embodiments of the present technology may be stored on such a portable medium and input to thecomputer system 1100 via theportable storage device 1140. -
Input devices 1160 provide a portion of a user interface.Input devices 1160 may include an alphanumeric keypad, such as a keyboard, for inputting alphanumeric and other information, or a pointing device, such as a mouse, a trackball, stylus, or cursor direction keys, or voice to text. Additionally, thesystem 1100 as shown inFIG. 11 includesoutput devices 1150. Suitable output devices include speakers, printers, network interfaces, and monitors. -
Display system 1170 may include a liquid crystal display (LCD) or other suitable display device.Display system 1170 receives textual and graphical information and processes the information for output to the display device. -
Peripherals devices 1180 may include any type of computer support device to add additional functionality to the computer system. Peripheral device(s) 1180 may include a modem or a router. - The components provided in the
computer system 1100 ofFIG. 11 are those typically found in computer systems that may be suitable for use with embodiments of the present technology and are intended to represent a broad category of such computer components that are well known in the art. Thus, thecomputer system 1100 ofFIG. 11 may be a personal computer, hand held computing system, telephone, mobile computing system, workstation, server, minicomputer, mainframe computer, or any other computing system. The computer may also include different bus configurations, networked platforms, multi-processor platforms, and so forth. Various operating systems may be used including Unix, Linux, Windows, Macintosh OS, Palm OS, Android, iPhone OS and other suitable operating systems. - It is noteworthy that any hardware platform suitable for performing the processing described herein is suitable for use with the technology. Computer-readable storage media refer to any medium or media that participate in providing instructions to a central processing unit (CPU), a processor, a microcontroller, or the like. Such media may take forms including, but not limited to, non-volatile and volatile media such as optical or magnetic disks and dynamic memory, respectively. Common forms of computer-readable storage media include a floppy disk, a flexible disk, a hard disk, magnetic tape, any other magnetic storage medium, a CD-ROM disk, DVD, any other optical storage medium, RAM, PROM, EPROM, a FLASHEPROM, any other memory chip or cartridge.
- While various embodiments have been described above, it should be understood that they have been presented by way of example only, and not limitation. The descriptions are not intended to limit the scope of the technology to the particular forms set forth herein. Thus, the breadth and scope of a preferred embodiment should not be limited by any of the above-described exemplary embodiments. It should be understood that the above description is illustrative and not restrictive. To the contrary, the present descriptions are intended to cover such alternatives, modifications, and equivalents as may be included within the spirit and scope of the technology as defined by the appended claims and otherwise appreciated by one of ordinary skill in the art. The scope of the technology should, therefore, be determined not with reference to the above description, but instead should be determined with reference to the appended claims along with their full scope of equivalents.
Claims (11)
1. A system for fulfilling travel requests, the system comprising:
a network of travel content suppliers for providing:
supplier-defined record types; and
travel content records;
a content database for storing the travel content records;
a metadata database for storing the supplier-defined record types;
a parsing module configured to:
receive a travel itinerary request from an exchange;
in response to the travel itinerary request, determine associated travel components expressed as metadata records and equivalence classes assembled as itinerary networks;
an artificial intelligence engine configured to impose a structure of travel logistics on the travel itinerary request, the artificial intelligence engine comprising:
a pattern recognition module for determining a pattern of the metadata records and the equivalence classes; and
a scheduling module configured to:
receive the itinerary networks;
search the content database to retrieve itinerary objects;
level the itinerary networks to select a subset of the itinerary objects; and
return the subset of the itinerary objects to the network of travel content suppliers.
2. The system of claim 1 , wherein the scheduling module is further configured to generate a constrained itinerary network based upon the subset of the itinerary objects.
3. The system of claim 2 , wherein the scheduling module is further configured to return the constrained itinerary network to the network of travel content suppliers.
4. The system of claim 1 , wherein each of the itinerary networks is associated with at least one of a stage, a state or a context.
5. The system of claim 1 , wherein the itinerary networks are expressed as directed acyclic graphs.
6. The system of claim 1 , wherein the supplier-defined record types and the travel content records are provided via an application programming interface.
7. The system of claim 1 , wherein the returning the subset of the itinerary objects to the network of travel content suppliers comprises returning the subset of the itinerary objects via the exchange, wherein the exchange employs a listener configured to identify the itinerary objects associated with the travel request.
8. The system of claim 1 , wherein the returning the subset of the itinerary objects to the network of travel content suppliers comprises returning the subset of the itinerary objects via the exchange to a subset of the travel content suppliers within the network of travel content suppliers, the subset of the travel content suppliers identified by the exchange based on the subset of the itinerary objects, wherein the exchange employs a listener configured to identify the itinerary objects associated with the travel request.
9. The system of claim 1 , wherein the returning the subset of the itinerary objects to the network of travel content suppliers comprises writing an offer expressed as a response.
10. The system of claim 9 wherein the network of travel content suppliers is configured to, in response to receiving the offer, provide a pricing for the subset of the itinerary objects.
11. The system of claim 10 , wherein the network of travel content suppliers is further configured to provide additional content relating to the offer.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US18/365,819 US20230385718A1 (en) | 2011-03-14 | 2023-08-04 | Artificially Intelligent Computing Engine for Travel Itinerary Resolutions |
Applications Claiming Priority (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201161452633P | 2011-03-14 | 2011-03-14 | |
US13/420,179 US10275810B2 (en) | 2011-03-14 | 2012-03-14 | Processing and fulfilling natural language travel requests |
US201862747088P | 2018-10-17 | 2018-10-17 | |
US16/396,487 US11763212B2 (en) | 2011-03-14 | 2019-04-26 | Artificially intelligent computing engine for travel itinerary resolutions |
US18/365,819 US20230385718A1 (en) | 2011-03-14 | 2023-08-04 | Artificially Intelligent Computing Engine for Travel Itinerary Resolutions |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/396,487 Continuation US11763212B2 (en) | 2011-03-14 | 2019-04-26 | Artificially intelligent computing engine for travel itinerary resolutions |
Publications (1)
Publication Number | Publication Date |
---|---|
US20230385718A1 true US20230385718A1 (en) | 2023-11-30 |
Family
ID=69161943
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/396,487 Active 2034-05-11 US11763212B2 (en) | 2011-03-14 | 2019-04-26 | Artificially intelligent computing engine for travel itinerary resolutions |
US18/365,819 Pending US20230385718A1 (en) | 2011-03-14 | 2023-08-04 | Artificially Intelligent Computing Engine for Travel Itinerary Resolutions |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/396,487 Active 2034-05-11 US11763212B2 (en) | 2011-03-14 | 2019-04-26 | Artificially intelligent computing engine for travel itinerary resolutions |
Country Status (1)
Country | Link |
---|---|
US (2) | US11763212B2 (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9659099B2 (en) * | 2011-03-14 | 2017-05-23 | Amgine Technologies (Us), Inc. | Translation of user requests into itinerary solutions |
US11049047B2 (en) | 2015-06-25 | 2021-06-29 | Amgine Technologies (Us), Inc. | Multiattribute travel booking platform |
CA2988975C (en) | 2015-06-18 | 2022-09-27 | Amgine Technologies (Us), Inc. | Scoring system for travel planning |
US11941552B2 (en) | 2015-06-25 | 2024-03-26 | Amgine Technologies (Us), Inc. | Travel booking platform with multiattribute portfolio evaluation |
US11210756B2 (en) * | 2018-09-28 | 2021-12-28 | Ford Global Technologies, Llc | Ride request interactions |
US20210264546A1 (en) * | 2020-02-26 | 2021-08-26 | Airbnb, Inc. | Accommodation furnishing based on subscription living user preferences |
US11783388B2 (en) | 2020-02-26 | 2023-10-10 | Airbnb, Inc. | Detecting user preferences of subscription living users |
Family Cites Families (229)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5021953A (en) | 1988-01-06 | 1991-06-04 | Travelmation Corporation | Trip planner optimizing travel itinerary selection conforming to individualized travel policies |
US5557524A (en) | 1991-10-18 | 1996-09-17 | Maki; Stanley C. | GPS/GLONASS travel recorder |
US5948040A (en) | 1994-06-24 | 1999-09-07 | Delorme Publishing Co. | Travel reservation information and planning system |
US5832452A (en) | 1996-01-31 | 1998-11-03 | Electronic Data Systems Corporation | Hotel database inquiry system |
US20020178034A1 (en) | 1996-04-10 | 2002-11-28 | Christopher W. Gardner | Airline travel technologies |
US5797127A (en) | 1996-12-31 | 1998-08-18 | Walker Asset Management Limited Partnership | Method, apparatus, and program for pricing, selling, and exercising options to purchase airline tickets |
US6059724A (en) | 1997-02-14 | 2000-05-09 | Biosignal, Inc. | System for predicting future health |
US8332247B1 (en) | 1997-06-12 | 2012-12-11 | G. William Bailey | Methods and systems for optimizing network travel costs |
US6307572B1 (en) | 1998-07-02 | 2001-10-23 | Ita Software, Inc. | Graphical user interface for travel planning system |
US6275808B1 (en) | 1998-07-02 | 2001-08-14 | Ita Software, Inc. | Pricing graph representation for sets of pricing solutions for travel planning system |
US6477520B1 (en) | 1999-02-22 | 2002-11-05 | Yatra Corporation | Adaptive travel purchasing optimization system |
AU5596700A (en) | 1999-06-03 | 2000-12-28 | Charles H. CELLA | Contingency-based options and futures for contingent travel accommodations |
US7181438B1 (en) | 1999-07-21 | 2007-02-20 | Alberti Anemometer, Llc | Database access system |
US7219073B1 (en) | 1999-08-03 | 2007-05-15 | Brandnamestores.Com | Method for extracting information utilizing a user-context-based search engine |
JP2003527664A (en) | 1999-09-10 | 2003-09-16 | ポストレル,リチャード | System and method for generating travel coupons |
US6836537B1 (en) | 1999-09-13 | 2004-12-28 | Microstrategy Incorporated | System and method for real-time, personalized, dynamic, interactive voice services for information related to existing travel schedule |
US20030187705A1 (en) | 1999-12-03 | 2003-10-02 | Schiff Martin R. | Systems and methods of comparing product information |
US7092892B1 (en) | 2000-03-01 | 2006-08-15 | Site59, Inc. | System and method for grouping and selling products or services |
EP1264264A1 (en) | 2000-03-10 | 2002-12-11 | Flighttime.Com, Inc. | Dynamic-risk pricing for air-charter services |
US7567958B1 (en) | 2000-04-04 | 2009-07-28 | Aol, Llc | Filtering system for providing personalized information in the absence of negative data |
US9026472B2 (en) | 2000-05-08 | 2015-05-05 | Smart Options, Llc | Method and system for reserving future purchases of goods and services |
US8650114B2 (en) | 2000-05-08 | 2014-02-11 | Smart Options, Llc | Method and system for reserving future purchases of goods or services |
US20030055772A1 (en) | 2000-06-06 | 2003-03-20 | Michael Goldstein | Method and apparatus to facilitate transactions between buyers and sellers |
US20030217052A1 (en) | 2000-08-24 | 2003-11-20 | Celebros Ltd. | Search engine method and apparatus |
US6553310B1 (en) | 2000-11-14 | 2003-04-22 | Hewlett-Packard Company | Method of and apparatus for topologically based retrieval of information |
JP2002163590A (en) | 2000-11-28 | 2002-06-07 | Sony Corp | Surrogate system, surrogate method, service surrogate server, corporate server, and recording medium |
US20020069133A1 (en) | 2000-12-04 | 2002-06-06 | Kirk Currie | System and method for selling fitted putters |
US7644032B2 (en) | 2000-12-19 | 2010-01-05 | H.J.J. Inc. | Adaptive matching program and system for corporate meeting planners and hospitality providers |
US6795710B1 (en) | 2001-01-05 | 2004-09-21 | Palmone, Inc. | Identifying client patterns using online location-based derivative analysis |
US20030018499A1 (en) | 2001-02-09 | 2003-01-23 | Miller Rodney D. | System and methods for continuous fare shopping and virtual grouping of itinerary requests |
US20020143587A1 (en) | 2001-04-02 | 2002-10-03 | Microsoft Corporation | Optimized system and method for finding best fares |
US20020147619A1 (en) | 2001-04-05 | 2002-10-10 | Peter Floss | Method and system for providing personal travel advice to a user |
US20030023463A1 (en) | 2001-04-16 | 2003-01-30 | Frank Dombroski | Method and system for automatically planning, booking, and calendaring travel arrangements |
AU2002258901B2 (en) | 2001-04-20 | 2007-03-29 | American Express Travel Related Services Company, Inc. | System and method for travel carrier contract management and optimization |
US20020173978A1 (en) | 2001-05-17 | 2002-11-21 | International Business Machines Corporation | Method and apparatus for scoring travel itineraries in a data processing system |
US7305356B2 (en) | 2001-05-25 | 2007-12-04 | Amadeus Americas, Inc. | Travel value index |
US7783506B2 (en) | 2001-08-17 | 2010-08-24 | Expedia, Inc. | System and method for managing reservation requests for one or more inventory items |
US20030055690A1 (en) | 2001-09-19 | 2003-03-20 | Garback Brent J. | Internet-based computer travel planning system |
JP4062910B2 (en) | 2001-11-29 | 2008-03-19 | 株式会社日立製作所 | HEALTH MANAGEMENT SUPPORT METHOD AND DEVICE AND HEALTH LIFE LIFE PREDICTION DATA GENERATION METHOD AND DEVICE |
AU2003213716A1 (en) | 2002-03-06 | 2003-09-22 | Cw Government Travel, Inc. | System, method and computer program product for on-line travel and expense management |
US8209200B2 (en) | 2002-03-13 | 2012-06-26 | Orbitz Llc | System and method for synchronizing passenger name record data |
US7080059B1 (en) | 2002-05-13 | 2006-07-18 | Quasm Corporation | Search and presentation engine |
US7398209B2 (en) | 2002-06-03 | 2008-07-08 | Voicebox Technologies, Inc. | Systems and methods for responding to natural language speech utterance |
EP1518201A4 (en) | 2002-06-19 | 2007-11-07 | Sabre Inc | Method system and computer program product for dynamic construction of packages and optimal assignement of generated packages to shopping categories |
US20040111255A1 (en) | 2002-12-10 | 2004-06-10 | International Business Machines Corporation | Graph-based method for design, representation, and manipulation of NLU parser domains |
US9818136B1 (en) | 2003-02-05 | 2017-11-14 | Steven M. Hoffberg | System and method for determining contingent relevance |
US7383239B2 (en) | 2003-04-30 | 2008-06-03 | Genworth Financial, Inc. | System and process for a fusion classification for insurance underwriting suitable for use by an automated system |
US7962354B2 (en) | 2003-06-06 | 2011-06-14 | Orbitz Llc | Booking engine for booking airline tickets on multiple host environments |
US8200775B2 (en) | 2005-02-01 | 2012-06-12 | Newsilike Media Group, Inc | Enhanced syndication |
US20060106655A1 (en) | 2003-08-05 | 2006-05-18 | Ladislav Lettovsky | System and method for coordinating travel itineraries |
US7050987B2 (en) | 2003-08-05 | 2006-05-23 | Sabre Inc. | System and method for coordinating travel itineraries |
WO2005017767A1 (en) | 2003-08-15 | 2005-02-24 | Silverbrook Research Pty Ltd | Natural language recognition using distributed processing |
US20050043940A1 (en) | 2003-08-20 | 2005-02-24 | Marvin Elder | Preparing a data source for a natural language query |
US7983956B1 (en) | 2003-10-24 | 2011-07-19 | Sachin Goel | System and method for providing options on products including flights |
US7424449B2 (en) | 2003-10-24 | 2008-09-09 | Sachin Goel | Computer-implemented method to provide options on products to enhance customer experience |
US20050108068A1 (en) | 2003-11-14 | 2005-05-19 | Marcken Carl D. | Generating flight schedules using fare routings and rules |
US8600845B2 (en) | 2006-10-25 | 2013-12-03 | American Express Travel Related Services Company, Inc. | System and method for reconciling one or more financial transactions |
US20050144162A1 (en) | 2003-12-29 | 2005-06-30 | Ping Liang | Advanced search, file system, and intelligent assistant agent |
US20050267651A1 (en) | 2004-01-15 | 2005-12-01 | Guillermo Arango | System and method for knowledge-based emergency response |
CN103398719B (en) | 2004-03-23 | 2017-04-12 | 咕果公司 | Digital mapping system |
US20050288973A1 (en) | 2004-06-24 | 2005-12-29 | Taylor Steven F | System and method for changing a travel itinerary |
US8600784B1 (en) | 2004-07-30 | 2013-12-03 | Kayak Software Corporation | Indentifying information sources to search based on rules |
US7640162B2 (en) | 2004-12-14 | 2009-12-29 | Microsoft Corporation | Semantic canvas |
US20060178931A1 (en) | 2005-02-05 | 2006-08-10 | Summerbrook Media Incorporated | Controling customer demand to match capicity |
US20070198442A1 (en) | 2005-02-05 | 2007-08-23 | Summerbrook Media Incorporated | Sales method for mobile media |
US9092523B2 (en) | 2005-02-28 | 2015-07-28 | Search Engine Technologies, Llc | Methods of and systems for searching by incorporating user-entered information |
US7979457B1 (en) | 2005-03-02 | 2011-07-12 | Kayak Software Corporation | Efficient search of supplier servers based on stored search results |
US8762160B2 (en) | 2005-03-23 | 2014-06-24 | Amadeus S.A.S. | Purchaser value optimization system |
US20060241983A1 (en) | 2005-04-21 | 2006-10-26 | Valerie Viale | Customer centric travel system |
US20060247954A1 (en) | 2005-04-29 | 2006-11-02 | Us Airways, Inc. | Method and system for scheduling travel ltineraries through an online interface |
US20060265508A1 (en) | 2005-05-02 | 2006-11-23 | Angel Franklin J | System for administering a multiplicity of namespaces containing state information and services |
US20070100962A1 (en) | 2005-05-10 | 2007-05-03 | Barth Brian E | Dynamic flight packaging |
US7813485B2 (en) | 2005-05-26 | 2010-10-12 | International Business Machines Corporation | System and method for seamlessly integrating an interactive visual menu with an voice menu provided in an interactive voice response system |
US8825370B2 (en) | 2005-05-27 | 2014-09-02 | Yahoo! Inc. | Interactive map-based travel guide |
US8005685B1 (en) | 2005-06-13 | 2011-08-23 | Amazon Technologies, Inc. | Ranking air travel search results based upon user criteria |
US20060293930A1 (en) | 2005-06-23 | 2006-12-28 | Rodgers Edward B | Sales call management |
US7949529B2 (en) | 2005-08-29 | 2011-05-24 | Voicebox Technologies, Inc. | Mobile systems and methods of supporting natural language human-machine interactions |
US20070073562A1 (en) | 2005-09-28 | 2007-03-29 | Sabre Inc. | System, method, and computer program product for providing travel information using information obtained from other travelers |
US7712670B2 (en) | 2005-09-28 | 2010-05-11 | Sauerwein Jr James T | Data collection device and network having radio signal responsive mode switching |
US8131574B2 (en) | 2005-09-29 | 2012-03-06 | Amadeus S.A.S. | Air travel system and method for planning roundtrip flights using one way combinable fares |
US20070260495A1 (en) | 2005-10-21 | 2007-11-08 | Scott Mace | Software Architecture and Database for Integrated Travel Itinerary and Related Reservation System Components |
US7627466B2 (en) | 2005-11-09 | 2009-12-01 | Microsoft Corporation | Natural language interface for driving adaptive scenarios |
US20070143154A1 (en) | 2005-12-20 | 2007-06-21 | Unisys Corporation | System and method for managing customer-based availability for a transportation carrier |
US20070156469A1 (en) | 2005-12-29 | 2007-07-05 | Bird Thomas K | Airline management system generating routings based on stored customer preference data |
US7860808B2 (en) | 2006-01-05 | 2010-12-28 | International Business Machines Corporation | System and method for hybrid conservation of fossil fuel |
US8005696B2 (en) | 2006-01-18 | 2011-08-23 | Ita Software, Inc. | Incremental searching in multi-passenger multi-route travel planning |
US20070168854A1 (en) | 2006-01-18 | 2007-07-19 | De Marcken Carl G | User interface for presentation of solutions in multi-passenger multi-route travel planning |
US8306835B2 (en) | 2006-01-18 | 2012-11-06 | Google Inc. | User interface for inputting multi-passenger multi-route travel planning query |
US20070185744A1 (en) | 2006-02-09 | 2007-08-09 | Steven Robertson | System and method for providing customized travel guides and itineraries over a distributed network |
EP1818867A1 (en) | 2006-02-14 | 2007-08-15 | Amadeus S.A.S. | Automated service fees assessment methods and system |
US20070192186A1 (en) | 2006-02-16 | 2007-08-16 | American Express Travel Related Services Co., Inc. A New York Corporation | Search, transfer, and booking tool for multiple rewards programs |
US8200514B1 (en) | 2006-02-17 | 2012-06-12 | Farecast, Inc. | Travel-related prediction system |
US20070203735A1 (en) | 2006-02-28 | 2007-08-30 | Commonwealth Intellectual Property Holdings, Inc. | Transaction Enabled Information System |
US20070208503A1 (en) | 2006-03-02 | 2007-09-06 | James Harnsberger | System and method for documenting a travel event |
WO2007124456A2 (en) | 2006-04-20 | 2007-11-01 | 10Best, Inc. | Ststem and method for providing travel-related products and services |
US9031853B2 (en) | 2006-05-06 | 2015-05-12 | Irody, Inc. | Apparatus and method for obtaining an identification of drugs for enhanced safety |
US10520325B2 (en) | 2006-05-25 | 2019-12-31 | Rideshark Corporation | Method of selective ride-sharing among multiple users along an optimized travel route |
US20080109232A1 (en) | 2006-06-07 | 2008-05-08 | Cnet Networks, Inc. | Evaluative information system and method |
US8407104B2 (en) | 2006-06-09 | 2013-03-26 | Campusi, Inc. | Catalog based price search |
GB2440958A (en) | 2006-08-15 | 2008-02-20 | Tomtom Bv | Method of correcting map data for use in navigation systems |
US8005832B2 (en) | 2006-08-29 | 2011-08-23 | Switchbook, Inc. | Search document generation and use to provide recommendations |
US20080091525A1 (en) | 2006-10-13 | 2008-04-17 | Laurent Kretz | Virtual online community with geographically targeted advertising |
US8972268B2 (en) | 2008-04-15 | 2015-03-03 | Facebook, Inc. | Enhanced speech-to-speech translation system and methods for adding a new word |
US7797187B2 (en) | 2006-11-13 | 2010-09-14 | Farecast, Inc. | System and method of protecting prices |
US8589211B2 (en) | 2006-11-15 | 2013-11-19 | Amadeus S.A.S. | Airline ticket change constrainer |
US7603436B2 (en) | 2006-11-17 | 2009-10-13 | Microsoft Corporation | Data capture and fusion from a population of device users |
US8200506B2 (en) | 2006-12-19 | 2012-06-12 | Accenture Global Services Limited | Integrated health management platform |
US20080201178A1 (en) | 2007-02-20 | 2008-08-21 | Yuri Vizitei | On-demand travel management service and platform |
US20080262995A1 (en) | 2007-04-19 | 2008-10-23 | Microsoft Corporation | Multimodal rating system |
US20080319803A1 (en) | 2007-06-20 | 2008-12-25 | Amadeus S.A.S. | Method and system for booking travel products online on the basis of up-to-date availability data displayed on a map-based client interface |
US20090006143A1 (en) | 2007-06-26 | 2009-01-01 | Rearden Commerce, Inc. | System and Method for Interactive Natural Language Rebooking or Rescheduling of Calendar Activities |
US20090005650A1 (en) | 2007-06-29 | 2009-01-01 | Robert Lee Angell | Method and apparatus for implementing digital video modeling to generate a patient risk assessment model |
US8035511B2 (en) | 2007-07-03 | 2011-10-11 | 3M Innovative Properties Company | Methods for providing services and information based upon data collected via wireless network sensors |
US20090063359A1 (en) | 2007-08-27 | 2009-03-05 | Connors Laurence A | Method of presenting predictive data of financial securities |
US20090070322A1 (en) | 2007-08-31 | 2009-03-12 | Powerset, Inc. | Browsing knowledge on the basis of semantic relations |
WO2009033167A1 (en) | 2007-09-07 | 2009-03-12 | Miller David A | Alternative route monitoring |
US20090112639A1 (en) | 2007-10-31 | 2009-04-30 | Robinson Beaver Nancy J | Combined Rewards System and Process Providing Variable Travel Redemption |
WO2009064390A2 (en) | 2007-11-09 | 2009-05-22 | Insidetrip, Inc. | Method and system for attribute-based evaluation of travel-related products and services |
EP2068276A1 (en) | 2007-12-04 | 2009-06-10 | Sony Corporation | Information processing device and method, program, and recording medium |
US20090157664A1 (en) | 2007-12-13 | 2009-06-18 | Chih Po Wen | System for extracting itineraries from plain text documents and its application in online trip planning |
US20090157312A1 (en) | 2007-12-14 | 2009-06-18 | Microsoft Corporation | Social network based routes |
US20090193352A1 (en) | 2008-01-26 | 2009-07-30 | Robert Stanley Bunn | Interface for assisting in the construction of search queries |
US20090210262A1 (en) | 2008-02-15 | 2009-08-20 | Remotian Systems, Inc. (Delaware Corporation) | Methods and apparatus for automated travel |
US20090216633A1 (en) | 2008-02-26 | 2009-08-27 | Travelocity.Com Lp | System, Method, and Computer Program Product for Assembling and Displaying a Travel Itinerary |
US20090313055A1 (en) | 2008-06-13 | 2009-12-17 | Natalie Martin | Computer-based system and method for facilitating travel planning for a prospective traveler |
US8224665B2 (en) | 2008-06-26 | 2012-07-17 | Archimedes, Inc. | Estimating healthcare outcomes for individuals |
US20090327148A1 (en) | 2008-06-27 | 2009-12-31 | Microsoft Corporation | Mechanisms and architecture for mobile opportunistic commerce |
US8175918B2 (en) | 2008-07-09 | 2012-05-08 | Orbitz Worldwide, L.L.C. | System and method for automatically determining travel product price rebates |
US20100010978A1 (en) | 2008-07-11 | 2010-01-14 | Amadeus S.A.S. | Method and system to search for travel products |
US20100030594A1 (en) | 2008-07-29 | 2010-02-04 | Garret Frederick Swart | Method for User Driven Multi-objective Optimization of Travel Plans |
US20100082241A1 (en) | 2008-09-30 | 2010-04-01 | Verizon Business Network Services Inc. | Method and system for providing navigational services |
US8631007B1 (en) | 2008-12-09 | 2014-01-14 | Google Inc. | Disambiguating keywords and other query terms used to select sponsored content |
US20100153292A1 (en) | 2008-12-11 | 2010-06-17 | Microsoft Corporation | Making Friend and Location Recommendations Based on Location Similarities |
US8438072B2 (en) | 2009-02-20 | 2013-05-07 | Consumercartel, Llc | Online exchange system and method with reverse auction |
US8583511B2 (en) | 2009-05-19 | 2013-11-12 | Bradley Marshall Hendrickson | Systems and methods for storing customer purchasing and preference data and enabling a customer to pre-register orders and events |
US20100324927A1 (en) | 2009-06-17 | 2010-12-23 | Tinsley Eric C | Senior care navigation systems and methods for using the same |
WO2011049612A1 (en) | 2009-10-20 | 2011-04-28 | Lisa Morales | Method and system for online shopping and searching for groups of items |
CN102655930B (en) | 2009-12-15 | 2014-12-24 | 因温斯特技术公司 | Emulsion compositions and a method for selecting surfactants |
US8805711B2 (en) | 2009-12-22 | 2014-08-12 | International Business Machines Corporation | Two-layer data architecture for reservation management systems |
US9558520B2 (en) | 2009-12-31 | 2017-01-31 | Hartford Fire Insurance Company | System and method for geocoded insurance processing using mobile devices |
US10284679B2 (en) | 2010-01-07 | 2019-05-07 | Microsoft Technology Licensing, Llc | Maintaining privacy during personalized content delivery |
US8612273B2 (en) | 2010-04-01 | 2013-12-17 | The Crawford Group, Inc. | Method and system for managing vehicle travel |
TWI463423B (en) | 2010-05-28 | 2014-12-01 | Poynt Corp | Method of using location information for advertising system based on 3-dimensional shapes |
US20110307280A1 (en) | 2010-06-15 | 2011-12-15 | Triptility, LLC | Apparatus and method for searching and booking a complete travel itinerary |
US8694537B2 (en) | 2010-07-29 | 2014-04-08 | Soundhound, Inc. | Systems and methods for enabling natural language processing |
US9009145B2 (en) | 2010-08-04 | 2015-04-14 | Amadeus S.A.S. | Travel booking method and system |
US20120054001A1 (en) | 2010-08-25 | 2012-03-01 | Poynt Corporation | Geo-fenced Virtual Scratchcard |
JP5071536B2 (en) | 2010-08-31 | 2012-11-14 | 株式会社デンソー | Information providing apparatus and information providing system |
US8775426B2 (en) | 2010-09-14 | 2014-07-08 | Microsoft Corporation | Interface to navigate and search a concept hierarchy |
US20120089406A1 (en) | 2010-10-11 | 2012-04-12 | Steven Ladd Huffman | System and method for grouping trip itineraries |
US8635010B2 (en) | 2011-02-15 | 2014-01-21 | Telenav, Inc. | Navigation system with accessory control mechanism and method of operation thereof |
US9053449B2 (en) | 2011-02-22 | 2015-06-09 | Theatrolabs, Inc. | Using structured communications to quantify social skills |
US9659099B2 (en) | 2011-03-14 | 2017-05-23 | Amgine Technologies (Us), Inc. | Translation of user requests into itinerary solutions |
US9286629B2 (en) | 2011-03-14 | 2016-03-15 | Amgine Technologies (Us), Inc. | Methods and systems for transacting travel-related goods and services |
US20120239584A1 (en) | 2011-03-20 | 2012-09-20 | Microsoft Corporation | Navigation to dynamic endpoint |
US9070100B2 (en) | 2011-03-31 | 2015-06-30 | United Parcel Service Of America, Inc. | Calculating speed and travel times with travel delays |
EP2509035A1 (en) | 2011-04-05 | 2012-10-10 | Amadeus S.A.S. | Reservation method and system with improved PNR handling |
US20120265696A1 (en) | 2011-04-12 | 2012-10-18 | Teletech Holdings, Inc. | Methods for providing dynamic and proactive support services |
WO2012142443A2 (en) | 2011-04-13 | 2012-10-18 | Douglas Krone | Systems and methods for facilitating the sale of goods and/or services via incentives |
US20130073323A1 (en) | 2011-04-22 | 2013-03-21 | Kayak Software Corporation | Travel exploration methods and apparatus |
US9449288B2 (en) | 2011-05-20 | 2016-09-20 | Deem, Inc. | Travel services search |
EP2541439A1 (en) | 2011-06-27 | 2013-01-02 | Amadeus s.a.s. | Method and system for processing a search request |
US20130041696A1 (en) | 2011-08-10 | 2013-02-14 | Postrel Richard | Travel discovery and recommendation method and system |
US20130054375A1 (en) | 2011-08-24 | 2013-02-28 | Accenture Global Services Limited | Personalized travel experience with social media integration |
US20130073325A1 (en) | 2011-09-21 | 2013-03-21 | Warren Ross | Method of Event and Venue Planning and Delivering System |
US20140279196A1 (en) | 2013-03-15 | 2014-09-18 | Nara Logics, Inc. | System and methods for providing spatially segmented recommendations |
US8170971B1 (en) | 2011-09-28 | 2012-05-01 | Ava, Inc. | Systems and methods for providing recommendations based on collaborative and/or content-based nodal interrelationships |
US20130090959A1 (en) | 2011-10-06 | 2013-04-11 | Seatme, Inc. | Restaurant management and reservation systems and methods |
WO2013074868A1 (en) | 2011-11-16 | 2013-05-23 | Flextronics Ap, Llc | Complete vehicle ecosystem |
US20130132128A1 (en) | 2011-11-17 | 2013-05-23 | Us Airways, Inc. | Overbooking, forecasting and optimization methods and systems |
US9323834B2 (en) | 2011-12-05 | 2016-04-26 | International Business Machines Corporation | Semantic and contextual searching of knowledge repositories |
US20130151291A1 (en) | 2011-12-08 | 2013-06-13 | Sumant Salway | System and method for building on-demand aviation trip |
US20130159023A1 (en) | 2011-12-16 | 2013-06-20 | Neela SRINIVAS | System and method for evidence based differential analysis and incentives based heal thcare policy |
US8612350B2 (en) | 2011-12-16 | 2013-12-17 | Ebay Inc. | Travel account |
EP2610801A1 (en) | 2011-12-27 | 2013-07-03 | Amadeus | Social network travel inspiration engine and method of same |
DE202013001048U1 (en) | 2012-02-20 | 2013-03-04 | Toll Collect Gmbh | Vehicle equipment, mobile device, computer program product, data processing device and charging system for booking and canceling user authorizations for alternatively usable tolled traffic areas |
AU2012100465B4 (en) | 2012-02-23 | 2012-12-06 | Uniloc Usa, Inc. | Health assessment by remote physical examination |
US8818715B2 (en) | 2012-03-29 | 2014-08-26 | Yahoo! Inc. | Systems and methods to suggest travel itineraries based on users' current location |
CA2873210A1 (en) | 2012-04-09 | 2013-10-17 | Vivek Ventures, LLC | Clustered information processing and searching with structured-unstructured database bridge |
US20140012659A1 (en) | 2012-07-09 | 2014-01-09 | Rong Yan | Modifying targeting criteria for an advertising campaign based on advertising campaign budget |
US20140067483A1 (en) | 2012-08-30 | 2014-03-06 | Electronics And Telecommunications Research Institute | Apparatus and method for constructing radar chart |
US20140074746A1 (en) | 2012-09-07 | 2014-03-13 | Hand Held Products Inc. doing business as (d.b.a) Honeywell Scanning & Mobility | Package source verification |
US9141975B2 (en) | 2012-09-23 | 2015-09-22 | Intel Corporation | Inferring user risk profile from travel patterns |
US20140089036A1 (en) | 2012-09-26 | 2014-03-27 | Xerox Corporation | Dynamic city zoning for understanding passenger travel demand |
US20140089020A1 (en) | 2012-09-27 | 2014-03-27 | Suitest IP Group, Inc. | Systems and methods for optimizing markets for temporary living space |
EP2912624A4 (en) | 2012-10-23 | 2016-04-27 | Olset Inc | Methods and systems for making travel arrangements |
US10083462B2 (en) | 2012-12-05 | 2018-09-25 | Capital One Services, Llc | Methods and systems for dynamically providing content |
US10185917B2 (en) | 2013-01-31 | 2019-01-22 | Lf Technology Development Corporation Limited | Computer-aided decision systems |
US9117182B2 (en) | 2013-02-14 | 2015-08-25 | Anshuman Bapna | Method and system for dynamic travel plan management |
US9082134B2 (en) | 2013-03-08 | 2015-07-14 | Zzzoom, LLC | Displaying advertising using transit time data |
US20140304014A1 (en) | 2013-04-09 | 2014-10-09 | Staccatotrip | Systems and methods associated with travel planning |
US20140330621A1 (en) | 2013-05-02 | 2014-11-06 | Optum, Inc. | Cms stars rating data management |
US20140330605A1 (en) | 2013-05-03 | 2014-11-06 | General Electric Company | System and method for monitoring and scheduling a workforce |
US20140330606A1 (en) | 2013-05-03 | 2014-11-06 | General Electric Company | System and method for scheduling |
US10909475B2 (en) | 2013-05-09 | 2021-02-02 | TravelPass, Group, LLC | Systems and methods for minimizing travel costs for multi-night stays |
WO2015021180A1 (en) | 2013-08-06 | 2015-02-12 | Amgine Technologies Limited | Travel booking platform |
US20150066594A1 (en) | 2013-08-27 | 2015-03-05 | New York University | System, method and computer accessible medium for determining one or more effects of rankings on consumer behavior |
US9946839B1 (en) | 2013-09-10 | 2018-04-17 | MD Insider, Inc. | Search engine systems for matching medical providers and patients |
US9043151B2 (en) | 2013-10-03 | 2015-05-26 | Sap Se | Large scale demand responsive transit framework |
US20150242927A1 (en) | 2013-10-03 | 2015-08-27 | Jason Will | Method and system of an online travel website |
US10191987B2 (en) | 2013-11-22 | 2019-01-29 | Capital One Services, Llc | Systems and methods for searching financial data |
US9709413B2 (en) | 2013-12-12 | 2017-07-18 | Cellco Partnership | Directions based on predicted future travel conditions |
US20150193583A1 (en) | 2014-01-06 | 2015-07-09 | Cerner Innovation, Inc. | Decision Support From Disparate Clinical Sources |
WO2015123390A1 (en) | 2014-02-12 | 2015-08-20 | Quixey, Inc. | Query cards |
US20150235478A1 (en) | 2014-02-14 | 2015-08-20 | International Business Machines Corporation | Global positioning system based toll road pricing |
CA2944652C (en) | 2014-04-01 | 2024-05-21 | Amgine Technologies (Us), Inc. | Inference model for traveler classification |
US9679078B2 (en) | 2014-05-21 | 2017-06-13 | Facebook, Inc. | Search client context on online social networks |
US9904769B2 (en) | 2014-06-06 | 2018-02-27 | Northwestern University | Systems and methods for providing a user engagement platform to support clinical decisions |
US9830647B2 (en) | 2014-08-15 | 2017-11-28 | International Business Machines Corporation | Adaptive dynamic budgeting |
US20160055162A1 (en) | 2014-08-20 | 2016-02-25 | Coalesce, Inc. | Systems and Methods for Information Search, Retrieval, Summarization and Interpretation using Related-Concept Analysis |
US10740412B2 (en) | 2014-09-05 | 2020-08-11 | Facebook, Inc. | Pivoting search results on online social networks |
US20160125559A1 (en) | 2014-11-02 | 2016-05-05 | Trippd Inc. | Trip planning platform |
US20160132791A1 (en) * | 2014-11-07 | 2016-05-12 | Graph SQL, Inc. | Methods and systems for distributed graphical flight search |
US11054266B2 (en) | 2015-01-08 | 2021-07-06 | International Business Machines Corporation | Confidential route monitoring with traveler-configured traveler safety alerts |
US20160203422A1 (en) | 2015-01-14 | 2016-07-14 | Nextop Italia Srl Semplificata | Method and electronic travel route building system, based on an intermodal electronic platform |
US20160232626A1 (en) | 2015-02-11 | 2016-08-11 | Amadeus S.A.S. | Travel activity tracking system |
US10281289B2 (en) | 2015-03-08 | 2019-05-07 | Microsoft Technology Licnensing, LLC | Search along the route |
US10169467B2 (en) | 2015-03-18 | 2019-01-01 | Microsoft Technology Licensing, Llc | Query formulation via task continuum |
US11049047B2 (en) | 2015-06-25 | 2021-06-29 | Amgine Technologies (Us), Inc. | Multiattribute travel booking platform |
US20160364815A1 (en) | 2015-06-11 | 2016-12-15 | Amgine Technologies (Us), Inc. | Multi-Passenger Travel Booking Platform |
CA2989325A1 (en) | 2015-06-17 | 2016-12-22 | Amgine Technologies (Us), Inc. | Travel booking with automatic consumption of loyalty reward points |
CA2988975C (en) | 2015-06-18 | 2022-09-27 | Amgine Technologies (Us), Inc. | Scoring system for travel planning |
US20190122315A1 (en) | 2017-10-19 | 2019-04-25 | Amgine Technologies (Us), Inc. | Itinerary Portfolio Evaluation Tool |
US20170154353A1 (en) | 2015-10-22 | 2017-06-01 | TripStreak Holdings, LLC | Systems and methods for generating a tripscore |
US10586300B2 (en) | 2015-11-10 | 2020-03-10 | Gt Gettaxi Limited | Graphical user interface (GUI) for implementing controls for geographic conveyance |
WO2017180483A1 (en) | 2016-04-11 | 2017-10-19 | Amgine Technologies (Us), Inc. | Insurance evaluation engine |
US20180336642A1 (en) | 2017-05-17 | 2018-11-22 | Amgine Technologies (Us), Inc. | Normalizing Allowable Budget of Travelers |
FR3078806A1 (en) | 2018-03-12 | 2019-09-13 | Amadeus S.A.S. | DETECTION, MONITORING AND MANAGEMENT OF TRAVEL DISTURBANCES |
US20190332921A1 (en) | 2018-04-13 | 2019-10-31 | Vosai, Inc. | Decentralized storage structures and methods for artificial intelligence systems |
-
2019
- 2019-04-26 US US16/396,487 patent/US11763212B2/en active Active
-
2023
- 2023-08-04 US US18/365,819 patent/US20230385718A1/en active Pending
Also Published As
Publication number | Publication date |
---|---|
US20200027039A1 (en) | 2020-01-23 |
US11763212B2 (en) | 2023-09-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20230385718A1 (en) | Artificially Intelligent Computing Engine for Travel Itinerary Resolutions | |
US10275810B2 (en) | Processing and fulfilling natural language travel requests | |
US11222269B2 (en) | Cognitive media content | |
US11488601B2 (en) | Dependency graph conversation modeling for use in conducting human-to-computer dialog sessions with a computer-implemented automated assistant | |
US10679622B2 (en) | Dependency graph generation in a networked system | |
US11010700B2 (en) | Identifying task and personality traits | |
US10824806B2 (en) | Counterintuitive recommendations based upon temporary conditions | |
US20190019233A1 (en) | Real time recommendation engine | |
US11443215B2 (en) | Intelligent recommendation of convenient event opportunities | |
US20190019217A1 (en) | Group formation and recommendations based on trigger events | |
CA3059005C (en) | Artificially intelligent computing engine for travel itinerary resolutions | |
US20230101339A1 (en) | Automatic response prediction | |
CN116235163A (en) | Inference-based natural language interpretation | |
US11100535B2 (en) | Group recommendations based on external factors | |
US20230153060A1 (en) | Dynamic display accommodations for multiple voice commands |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: AMGINE TECHNOLOGIES (US), INC., DELAWARE Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:MILLER, JONATHAN DAVID;MILLER, HAROLD ROY;REEL/FRAME:064506/0231 Effective date: 20190606 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |