×

Online taxonomy for constructing customer service queries

  • US 7,979,301 B1
  • Filed: 07/16/2007
  • Issued: 07/12/2011
  • Est. Priority Date: 09/03/2002
  • Status: Expired due to Fees
First Claim
Patent Images

1. An Internet-based computer-assisted method for generating a static taxonomy of classes of data objects representing a knowledge base that provides on a server a navigable network structure configured to construct a query, wherein a user communicates with the server via a web browser and the construction of the query is based at least upon knowledge captured by the knowledge base as a result of a sequence of selection steps executed by the user in connection with a navigation of the network structure, wherein at each selection step a choice of classes of data objects is presented to the user, the method comprising:

  • providing on a server a navigable network structure comprising a plurality of unique nodes and links configured to present to a user, in one or more web pages of a web browser, a static taxonomy of classes of data objects representing a knowledge base, wherein;

    the static taxonomy starts with a static taxonomy setup state based upon existing available knowledge, wherein the static taxonomy setup state provides a seed configuration for a subsequent evolution of the static taxonomy;

    the static taxonomy setup state is followed by a learning phase, comprising a finite sequence of additional static taxonomy states;

    each static taxonomy state in the sequence reflects an improved level of knowledge and performance of the static taxonomy over previous static taxonomy states;

    each static taxonomy state in the sequence is associated with a life cycle during which navigational data is archived, wherein the navigational data provides the foundation to generate upgrades for a subsequent static taxonomy state and wherein the length of the life cycle of each static taxonomy state is based at least upon the time required to collect a sufficient amount of navigational data consistent with sample size requirements for statistical analysis;

    each static taxonomy state in the sequence is associated with the subsequent static taxonomy state, which is configured off-line during the life cycle of the static taxonomy state, based at least upon the generated upgrades;

    each static taxonomy state in the sequence is associated with a selected date and time for a transition from the static taxonomy state to the subsequent static taxonomy state;

    a last static taxonomy state of the learning phase occurs when the static taxonomy has evolved to a point wherein no further knowledge needs to be added for an efficient operation of the static taxonomy;

    the last static taxonomy state of the learning phase is followed by a maintenance taxonomy state limited to edits and updates, which may be required to reflect any changes in the environment in which the static taxonomy operates including changes in user behavior; and

    each node of the plurality of nodes in each of the static taxonomy states of the static taxonomy;

    is uniquely identified;

    is associated with a class of data objects;

    provides in the one or more web browser pages a presentation of the class of data objects associated with the node, wherein the presentation comprises characteristics of the class of data objects associated with the node;

    provides access to a user-exercisable option for the user to indicate acceptance of the class of data objects associated with the node; and

    provides access to a user-exercisable option for the user to request help from a live agent;

    for each state of the finite sequence of static taxonomy states, accumulating navigational data for each navigation of a plurality of navigations performed over the life cycle of the static taxonomy state, wherein the navigational data comprise, for each node visited sequentially in the course of each navigation;

    a node identifier;

    a user identifier;

    a time stamp identifying an arrival time of the user at the node; and

    help request data, if the user exercisable option to request help from a live agent is exercised, wherein the help request data comprise;

    a time stamp identifying a time when the help request is exercised;

    a session time, wherein the session time comprises a length of time of a communication session between the user and the live agent; and

    at least one reason why the help request was exercised;

    for each state of the finite sequence of static taxonomy states, based at least on the accumulated navigational data collected over the life cycles of previous static taxonomy states, predictively scheduling live agents to respond to help requests within a programmable maximum acceptable response time after the help request is exercised;

    in response to receiving from a user visiting a node an indication of acceptance of the class of data objects associated with the node, constructing a query based at least on a sequence of selection steps executed by the user;

    for each static taxonomy state of the finite sequence of static taxonomy states, based at least on the accumulated navigational data, and the selected date and time for a transition from the static taxonomy state to a subsequent static taxonomy state, executing the transition to the subsequent static taxonomy state; and

    repeating the steps of accumulating navigational data and transitioning to the next static taxonomy state until the maintenance state is reached.

View all claims
  • 0 Assignments
Timeline View
Assignment View
    ×
    ×