【Purpose/significance】This paper comes up with an improved algorithm of TF*IDF to calculate the weight of words from papers after the word segmentation, extract high weight words and determine the research hot spots.【Method/process】Taking the articles published in Journal of the China Society for Scientific and Technical Information which from Wan Fang data platform in 2015 as an example, segment the word of every article, then calculate the weight of words by the improved algorithm of TF*IDF.【Result/conclusion】This improved algorithm is accurate and feasible. And through analyzing, it is found that the research hot spots of information science in 2015 are user researches,big data,informatics,social network,technical field,the author,emergency and non-citation and so on.