Automatic recommendation of products using latent semantic indexing of content

A method for automatically recommending textual items stored in a database to a user of a computer-implemented service, the method comprising the steps of

• filtering the textual items based upon pre-selected textual content,
• applying a latent semantic algorithm to the filtered textual items to establish a conceptual similarity among the textual items,
• generating a list of items nearest to each of the textual items with reference to the conceptual similarity,
• filtering the list of nearest items based upon pre-determined criteria to produce a second list of filtered nearest items, and
• upon selection of one of the items by the user, outputting to the user a recommended set of nearest items to the selected item created from a subset of the second list based upon the conceptual similarity.

A method for automatically recommending textual items stored in a database to a user of a computer-implemented service, the user having selected a plurality of items, the method comprising the steps of

• filtering the textual items based upon pre-selected textual content,
• applying a latent semantic algorithm to the filtered textual items with reference to the plurality of selected items to establish a conceptual similarity among the textual items and the selected items,
• generating a list of items nearest to each of the textual items with reference to the conceptual similarity,
• filtering the list of nearest items based upon pre-determined criteria to produce a second list of filtered nearest items, and
• outputting to the user a recommended set of nearest items to the selected item created from a subset of the second list based upon the conceptual similarity.

A method for automatically recommending items stored in a database to a user of a computer-implemented service, the method comprising the steps of

• generating a text-based description of each of the items,
• filtering the items based upon pre-selected textual content to produce filtered items,
• applying a latent semantic indexing algorithm to the filtered items to establish a conceptual similarity among the filtered items,
• generating a first list of filtered items nearest to the selected item with reference to the conceptual similarity among the filtered items,
• filtering the first list of nearest items based upon pre-determined criteria to produce a second list of filtered nearest items, and
• upon selection of one of the items by the user, outputting to the user a recommended set of filtered nearest items to the selected item by consulting the second list.

A method for on-line merchandising of textual items to a user over the Internet during an on-line session, the textual items being stored remotely in a database of an Internet application server, the user having selected one of the items, the method comprising the steps of

• filtering the textual items based upon pre-selected textual content,
• executing in the server a latent semantic algorithm on the filtered textual items to establish a conceptual similarity among the textual items and the selected item,
• generating a first list of filtered items nearest to the selected item with reference to the conceptual similarity,
• filtering the first list of nearest items based upon pre-determined criteria to produce a second list of filtered nearest items, and
• outputting to the user a recommended set of nearest items to the selected item created from a subset of the second list based upon the conceptual similarity.