Method of finding correlations between news events and user feedback based on Wikipedia data
Jazyk: | ruština |
---|---|
Rok vydání: | 2022 |
Předmět: | |
DOI: | 10.18720/spbpu/3/2023/vr/vr23-397 |
Popis: | Тема вÑпÑÑкной квалиÑикаÑионной ÑабоÑÑ: «ÐеÑод поиÑка коÑÑелÑÑий Ð¼ÐµÐ¶Ð´Ñ Ð½Ð¾Ð²Ð¾ÑÑнÑми ÑобÑÑиÑми и оÑкликами полÑзоваÑелей на оÑнове даннÑÑ Â Wikipedia».ÐÐ°Ð½Ð½Ð°Ñ ÑабоÑа поÑвÑÑена иÑÑÐ»ÐµÐ´Ð¾Ð²Ð°Ð½Ð¸Ñ Ð¼ÐµÑодов Ð´Ð»Ñ Ð°Ð½Ð°Ð»Ð¸Ð·Ð° ÑеакÑии лÑдей на ÑазлиÑнÑе ÑобÑÑиÑ. ÐÐ»Ñ Ð¸ÑÑÐ»ÐµÐ´Ð¾Ð²Ð°Ð½Ð¸Ñ Ð±Ñли взÑÑÑ Ð´Ð°Ð½Ð½Ñе колиÑеÑÑва пÑоÑмоÑÑов ÑÑÑаниÑ Wikipedia. ÐадаÑи, коÑоÑÑе ÑеÑалиÑÑ Ð² Ñ Ð¾Ð´Ðµ иÑÑледованиÑ:ÐзÑÑение меÑодов Ð´Ð»Ñ Ð²ÑÑÐ²Ð»ÐµÐ½Ð¸Ñ ÑобÑÑий на оÑнове ÑекÑÑовÑÑ Ð´Ð°Ð½Ð½ÑÑ Ð¸Ð»Ð¸ ÑиÑленнÑÑ Ñ Ð°ÑакÑеÑиÑÑик.РазÑабоÑка меÑода, коÑоÑÑй Ð¿Ð¾Ð·Ð²Ð¾Ð»Ð¸Ñ Ð²ÑÑвиÑÑ ÑобÑÑÐ¸Ñ Ð² ÑеÑии даннÑÑ Ð¾ колиÑеÑÑве пÑоÑмоÑÑов и обнаÑÑжиÑÑ ÐºÐ¾ÑÑелÑÑÐ¸Ñ ÑобÑÑий Ñ Ð½Ð¾Ð²Ð¾ÑÑнÑми ÑÑаÑÑÑми.РазÑабоÑка пÑогÑаммного ÑеÑениÑ, коÑоÑое бÑÐ´ÐµÑ ÑеализовÑваÑÑ Ð´Ð°Ð½Ð½Ñй меÑод.ÐÑÑем анализа ÑÑÑеÑÑвÑÑÑÐ¸Ñ ÑÐ°Ð±Ð¾Ñ Ð² данной пÑедмеÑной облаÑÑи бÑли вÑÐ´ÐµÐ»ÐµÐ½Ñ Ð¾ÑновнÑе меÑÐ¾Ð´Ñ Ð´Ð»Ñ Ð²ÑÑÐ²Ð»ÐµÐ½Ð¸Ñ ÑобÑÑий. ÐÑл пÑоведен ÑÑавниÑелÑнÑй анализ инÑÑÑÑменÑов, коÑоÑÑе анализиÑÑÑÑ ÑÐ¾Ñ Ð¶Ðµ Ð½Ð°Ð±Ð¾Ñ Ð´Ð°Ð½Ð½ÑÑ . ÐÑла вÑделена ÑÑнкÑионалÑноÑÑÑ, коÑоÑÐ°Ñ Ð¾ÑÑÑÑÑÑвÑÐµÑ Ð²Ð¾ вÑÐµÑ ÑаÑÑмаÑÑиваемÑÑ Ð¸Ð½ÑÑÑÑменÑÐ°Ñ . ÐÑÑем обобÑÐµÐ½Ð¸Ñ Ð¸Ð½ÑоÑмаÑии о ÑÑÑеÑÑвÑÑÑÐ¸Ñ Ð¼ÐµÑÐ¾Ð´Ð°Ñ Ð´Ð»Ñ Ð¾Ð¿ÑÐµÐ´ÐµÐ»ÐµÐ½Ð¸Ñ ÑобÑÑий, бÑл ÑазÑабоÑан и пÑедÑÑавлен меÑод, коÑоÑÑй позволÑÐµÑ Ð°Ð²ÑомаÑизиÑоваÑÑ Ð¿ÑоÑеÑÑ Ð¾Ð±Ð½Ð°ÑÑÐ¶ÐµÐ½Ð¸Ñ ÑобÑÑÐ¸Ñ Ð² ÑеÑии даннÑÑ ÐºÐ¾Ð»Ð¸ÑеÑÑва пÑоÑмоÑÑов ÑазлиÑнÑÑ ÑÑÑаниÑ Wikipedia и найÑи коÑÑелÑÑÐ¸Ñ Ð¼ÐµÐ¶Ð´Ñ Ð²ÑÑвленнÑми ÑобÑÑиÑми и новоÑÑнÑми ÑÑаÑÑÑми, коÑоÑÑе могли повлиÑÑÑ Ð½Ð° возникновение ÑÑÐ¸Ñ ÑобÑÑий. Topic of the graduation thesis: "Method of searching for correlations between news events and user responses based on Wikipedia data".This work is devoted to the study of methods for analyzing people's reactions to various events. For the study, data on the number of page views of Wikipedia were taken. Tasks that were solved in the course of the study:1. Studying methods for identifying events based on textual data or numerical characteristics.2. Development of a method that will allow to identify events in a series of data on the number of views and to detect the correlation of events with news articles.3. Development of a software solution that will implement this method.By analyzing existing works in this subject area, the main methods for detecting events were identified. A comparative analysis of tools that analyze the same data set was carried out. Functionality that is missing in all the considered tools was highlighted. By summarizing information about existing methods for determining events, a method was developed and presented that allows you to automate the process of detecting an event in a series of data on the number of views of various Wikipedia pages and find a correlation between detected events and news articles that could influence the occurrence of these events. |
Databáze: | OpenAIRE |
Externí odkaz: |