Method for predicting citation counts
OwnerLawrence Fu, Constantin Aliferisinventors
Life of this patent
Estimating the term…
Abstract
A computerized process to predict citation counts of articles using only information available before or at the time of publication. The process involves obtaining a set of articles and extracting a set of features containing information about the article, author, and bibliometric data. The extracted features are converted into a format suitable for analysis, and models are constructed using a pattern recognition process. The constructed models are applied to a related article that was not included in the original article set for model construction. Features are extracted from the article of interest, and the models provide a prediction of whether a given number of citations will be received by the article.
Claims28
A computer-based process utilizing a specifically programmed computer to predict before, at, or after the time of publication the likely number of citations articles will receive over time comprising the steps of:
- a) receiving a selected article through an input along with relevant article information related to the selected article;
- b) obtaining, through the input, a set of articles along with relevant article information related to the set of articles related to the same general field as the selected article but exclusive of the selected article;
- c) storing in a memory the set of articles and the selected article;
- d) extracting and formatting through a computer processor, features from each article in the set of articles, the selected article, relevant article information related to the set of articles, and relevant article information related to the selected article, to enable a pattern recognition process to use the articles and relevant article information as inputs, wherein the relevant article information related to the selected article uses only features that are available before or at the time of publication of the selected article;
- e) from each article in the stored set of articles;
- 1) constructing models through the computer processor employing feature reduction and a pattern recognition process and the extracted article features;
- and
- 2) selecting, through the processor, a best model;
- f) predicting by application of the best model to the selected article by the processor future citation counts for the selected article;
- and
- g) outputting, the selected article along with its predicted future citation counts.
A computer system for predicting a citation count of an article before or at the time of publication and controlling publication based on said prediction comprising:
- a) an input device for receiving information regarding articles;
- b) a memory for storing the information regarding articles;
- c) a processor configured to:
- 1) receive a selected article through an input along with relevant article information related to the selected article;
- 2) obtain, through the input, a set of articles along with relevant article information related to the set of articles related to the same general field as the selected article but exclusive of the selected article;
- 3) store in a memory the set of articles and the selected article;
- 4) extract and format features from each article in the stored set of articles, the selected article, relevant article information related to the selected article, and relevant article information related to the set of articles, to enable a pattern recognition process to use the articles and relevant article information as inputs, wherein the relevant article information related to the selected article uses only features that are available before or at the time of publication of the selected article;
- 5) use each article in the stored set of articles to construct models employing feature reduction and a pattern recognition process and the extracted article features;
- and to select a best model;
- 6) predict by application of the best model to the selected article future citation counts for the selected article;
- d) an output device for outputting the selected article along with its predicted citation counts;
- e) a publication controller to control the distribution of the selected article.
A non-transitory computer readable storage medium containing a program for operating a computer for predicting before, at, or after the time of publication the likely number of citations articles will receive over time, the program causing the computer to perform acts comprising:
- a) receiving a selected article through an input along with relevant article information related to the selected article;
- b) obtaining, through an input, a set of articles along with relevant article information related to the set of articles related to the same general field as the selected article but exclusive of the selected article;
- c) storing in a memory the set of articles and the selected article;
- d) extracting and formatting through a computer processor features from each article in the stored set of articles, the selected article, relevant article information related to the set of articles, and relevant article information related to the selected article, to enable a pattern recognition process to use the articles and relevant article information as inputs, wherein the relevant article information related to the selected article uses only features that are available before or at the time of publication of the selected article;
- e) from each article in the stored set of articles;
- 1) constructing models through a computer processor employing feature reduction and a pattern recognition process and the extracted article features;
- and
- 2) selecting, through the processor, a best model;
- f) predicting by application of the best model to the selected article by the processor future citation counts for the selected article;
- and
- g) outputting the selected article along with its predicted future citation counts.
Dependencies are inferred from the claim text.