Determining the average reading speed of a user
AssigneeAmazon Technologies, Inc.
InventorsRajasimman Madhivanan, Zaka Ur Rehman Ashraf, Edward J. Gayles, Martin D. Hock
Life of this patent
Estimating the term…
Abstract
A user device determines a reading speed value for each of a plurality of measurable units of an electronic publication. The user device adds each reading speed value into one of a plurality of data sets or one of a plurality of outlier sets. The user device calculates the average reading speed of the user based on one or more of the plurality of data sets.
Claims24
A method comprising:
- detecting, using an optical sensor of a user device, that a user is viewing an electronic publication presented on a display of the user device, the detecting including tracking eye movement of the user;
- determining a plurality of reading speed values, wherein each reading speed value, of the plurality of reading speed values, is associated with a respective portion of the electronic publication and is determined based on tracking the eye movement of the user;
- determining a first normal probability distribution from the plurality of reading speed values;
- determining an outlier distribution from the plurality of reading speed values, wherein reading speed values in the outlier distribution do not follow a normal distribution;
- determining a second normal probability distribution using a subset of the reading speed values in the outlier distribution;
- calculating, by a processing device of the user device, a weighted average reading speed using reading speed values in one or more of the first normal probability distribution or the second normal probability distribution;
- estimating an amount of time remaining for a user to finish reading an unread portion of the electronic publication based on the weighted average reading speed;
- and
- causing presentation of the amount of time remaining on the display of the user device.
A device comprising:
- an optical sensor;
- a memory configured to store a plurality of reading speed values;
- and
- a processing device, coupled to the memory, the processing device configured to:
- determine, using the optical sensor configured to track eye movement of a user viewing an electronic publication, a first reading speed value of a plurality of reading speed values, wherein the first reading speed value corresponds to a respective portion of the electronic publication;
- determine a plurality of probability distributions from the plurality of reading speed values;
- determine a plurality of outlier distributions from the plurality of reading speed values;
- determine an additional probability distribution using a subset of reading speed values in one of the plurality of outlier distributions when a number of reading speed values in the subset is greater than a threshold size;
- and
- calculate a weighted average reading speed using one or more of the plurality of probability distributions and the additional probability distribution.
A non-transitory computer-readable storage medium storing instructions which, when executed, cause a processing device to perform operations comprising:
- determining, using an optical sensor configured to track eye movement of a user viewing an electronic publication, a plurality of reading speed values, wherein a first reading speed value corresponds to a respective portion of the electronic publication;
- adding each reading speed value into one of a plurality of data sets or into one of a plurality of outlier sets;
- adding a subset of reading speed values to a new data set when the subset of reading speed values from one of the plurality of outlier sets follows a normal probability distribution and when a number of reading speed values in the subset is greater than a threshold size;
- estimating, by the processing device, an amount of time remaining for the user to finish reading the electronic publication, based on one or more of the plurality of data sets and the new data set;
- and
- causing display of data indicative of the amount of time remaining.
Dependencies are inferred from the claim text.