Browse Prior Art Database

Social events recommendation ranking mechanism

IP.com Disclosure Number: IPCOM000244287D
Publication Date: 2015-Nov-30
Document File: 3 page(s) / 34K

Publishing Venue

The IP.com Prior Art Database

Abstract

This invention enables a better ranking of the recommended for user events in a social network. The new ranking mechanism refers to the events which were earlier attended by the user and which maybe of higher probability to be attended by the user with respect to his interests, his social connections and different event categories like size, location, time etc.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 40% of the total text.

Page 01 of 3

Social events recommendation ranking mechanism

This invention relates to the field of social networks, and in particular to social events recommendations.

Social network provide a convenient platform for organizing and managing social events. The events are created by the event organizers, and contains time, location and description of the event. The event can include categories, tags and other classification information which can be provided explicitly, or rendered automatically from the event description and other metadata. In order to connect between the users and the relevant events, which are of interest to the users, recommendation engines are used. These engines usually take into consideration the users' interests, users' actions history and the event time and location.

These recommendation engines can provide a list of recommended users for an event, a list of recommended events for a user, or both the events and the users.

The problem in the approach taken by the existing engines is that they do not verify the probability of likelihood of each user to attend a specific event, based on factors other than interests and explicitly predefined distance between the event location and the user location .

    The history of other event invitations is analyzed. For each event we look at various factors such as the categories, location, time, list of participants, size of event and other relevant information. This information is used to create a decision profile for a user, based on event categories.

We look only at events which are of interest to the user (based on factors already used by existing engines - tags, topics in the event text, categories etc.).

Based on the user's decision whether to attend these events, we analyze the factors that influence the likelihood or probability that the user joins the events, or on the contrary declines the invitation.

Some of such factors can be:

Event size (some users may not like large or very small events)

Time to get to the event (some users may not like events which are far away from them)

The list of event participants (there maybe negative correlation between some users and the others)

Specific time of event - whether it is during the weekend, or working days, and what time of day. Based on statistics of events that the user attended (i.e. marked "attending" on the event, or was tagged at the time and location of the event) and events that the user was invited to but did not attend, we build a profile for the user.

In the profile we calculate probabilities of the user attending an event based on various factors , assuming that the event is of interest to the user (in terms of category/topic).

For example, for a user u:

p(u, part_of_week) is the probability of the user to attend an event in part_of_week, where part_of_week is weekend or weekday.

p(u, size_of_event) is the probability of the user to attend an event of size size_of_event, where size_of_event is small/medium/large.

p(u, t) i...