學術產出-Theses
Article View/Open
Publication Export
-
題名 基於人力銀行之台灣地區薪資預測模型
Web-Recruitment Data for Salary Prediction in Taiwan作者 廖宜川
Liao, Yi-Chuan貢獻者 陳樹衡
Chen, Shu-Heng
廖宜川
Liao, Yi-Chuan關鍵詞 薪資預測
機器學習
卷積神經網路
自然語言處理
Word2Vec
詞向量
高維數據
Salary prediction
Machine learning
Convolutional neural network
Natural language processing
Word2Vec
Word vector
High dimension data日期 2020 上傳時間 2-Sep-2020 12:45:28 (UTC+8) 摘要 本文的研究目的在於建構一個薪資預測模型,在此特別針對資訊軟體系統類相關職缺。此薪資預測模型可作為求職者與企業方的參考依據,根據結構化變數,包括個人資料與職位相關技能等等,以及工作內容的文字描述,可以讓他們了解該職位的大略薪資,減少雙方對於薪資的歧見。同時,從迴歸模型輸出的係數也可以知道各種變數所反映的市場價值,例如熟悉某項工作技能會對於薪資水準有甚麼樣的影響,提供求職者自我精進的方向與參考。本研究從資料的探索性分析開始,了解各個變數的基本特徵,並嘗試整合結構化變數(職位需求的條件等等)以及非結構化的變數(工作內容的文字描述),藉由許多的機器學習演算法建立薪資預測模型。另外,也嘗試使用詞向量轉換的神經網路模型,針對工作內容的文字描述建立薪資預測模型,其評估結果並不亞於使用結構化變數的薪資預測模型,這顯示了中文的自然語言處理,應用於網路人力銀行資料集的薪資預測模型之建構是可行的。
The purpose of this thesis is to construct a salary prediction model, especially for information software system related positions using web-recruitment data. Based on structured data, including personal information and job-related skills, as well as unstructured text describing job content, the established models can be used as a reference for job seekers and companies to estimate the salary level of a certain job. Meanwhile, the variable coefficients from the regression models provide information about the market value reflected by those variables. The identified high-pay skills and expertise could guide the job seekers in which areas they can improve themselves. This research starts with an exploratory data analysis which helps us to understand the basic characteristics of each variable. Next, we apply various machine learning algorithms to the integrated structured and unstructured data to establish salary prediction models. The results show Random Forest, Ridge and Lasso perform well on the sparse high-dimension dataset. After that, we adopt a natural language processing approach by employing a convolutional neural network on the word vector data transformed from job content text. The result shows that the created salary prediction model is on a par with the models constructed using integrated structured and unstructured data. This endorses natural language processing as a viable approach to construct salary prediction models using online recruitment data.參考文獻 [1] 104人力銀行,AI大浪捲動企業搶才職缺是5年前的3.2倍,上網日期2020年06月20日,檢自:https://corp.104.com.tw/archive/files/news/20200121.pdf[2] 104人力銀行,上網日期2020年06月20日,檢自:https://www.104.com.tw/jobs/main/https://www.cnbc.com/2019/12/30/5-hig[3] Altman, N. S. (1992). An introduction to kernel and nearest-neighbor nonparametric regression. The American Statistician, 46(3), 175-185.[4] Breiman, L. (2001). Random forests. Machine learning, 45(1), 5-32.[5] Breiman, L., J. Friedman, R. Olshen, and C. Stone, (1984). Classification and Regression Trees. Belmont, California : Wadsworth International Group.[6] Collobert, R., Weston, J., Bottou, L., Karlen, M., Kavukcuoglu, K., & Kuksa, P. (2011). Natural language processing (almost) from scratch. Journal of machine learning research, 12(ARTICLE), 2493-2537.[7] Drucker, H., Burges, C.J.C., Kaufman, L., Smola, A. and Vapnik, V, (1997). “Support vector regression machines”, Advances in Neural Information Processing Systems, 9:155–161.[8] Freund, Y., & Schapire, R. E. (1997). A decision-theoretic generalization of on-line learning and an application to boosting. Journal of computer and system sciences, 55(1), 119-139.[9] Friedman, J., Hastie, T., & Tibshirani, R. (2010). Regularization paths for generalized linear models via coordinate descent. Journal of statistical software, 33(1), 1.[10] Hinton, G. E. (1990). Connectionist learning procedures. In Machine learning (pp. 555-610). Morgan Kaufmann.[11] Jones, K. S. (1972). A statistical interpretation of term specificity and its application in retrieval. Journal of documentation.[12] Keras, Retrieved June 20 2020, from: https://keras.io/[13] Kim, Y. (2014). Convolutional neural networks for sentence classification. arXiv preprint arXiv:1408.5882.[14] Martín, I., Mariello, A., Battiti, R., & Hernández, J. A. (2018). Salary Prediction in the IT Job Market with Few High-Dimensional Samples: A Spanish Case Study. International Journal of Computational Intelligence Systems, 11(1), 1192-1209.[15] Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781.[16] Pawha, A., & Kamthania, D. (2019). Quantitative analysis of historical data for prediction of job salary in India-A case study. Journal of Statistics and Management Systems, 22(2), 187-198.[17] Scikit-learn, Retrieved June 20 2020, from: https://scikit-learn.org/stable/[18] Selenium with Python, Retrieved June 20 2020, from: https://selenium-python.readthedocs.io/[19] Singh, R. (2016). A Regression Study of Salary Determinants in Indian Job Markets for Entry Level Engineering Graduates.[20] Sun Junyi,结巴中文分词,上網日期2020年06月20日,檢自https://github.com/fxsjy/jieba[21] Support Vector Machine - Regression(SVR), Retrieved June 20 2020, from: http://www.saedsayad.com/support_vector_machine_reg.htm[22] These 5 high-paying, growing jobs didn’t exist a decade ago—but they’ll be booming through the 2020s, Retrieved June 20 2020, from: https://www.cnbc.com/2019/12/30/5-high-paying-growing-jobs-that-will-be-booming-through-the-2020s.html?fbclid=IwAR1mOcFVDUNxaGk5EAsbkxLU2wP40yxLb8cBqNGjrccXgXoCoiuR4_LxTTQ[23] Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B (Methodological), 58(1), 267-288.[24] Vapnik, V. N. (1995). Constructing learning algorithms. In The nature of statistical learning theory (pp. 119-166). Springer, New York, NY.[25] Zou, H., & Hastie, T. (2005). Regularization and variable selection via the elastic net. Journal of the royal statistical society: series B (statistical methodology), 67(2), 301-320.[26] 中央研究院詞庫小組,中文斷詞系統,上網日期2020年06月20日,檢自:http://ckipsvr.iis.sinica.edu.tw/[27] 江易麇,(2018)。應用雙向長短期記憶神經網路於新聞分類。未出版之碩士論文,國立雲林科技大學,資訊管理系,雲林縣。[28] 周宜滿,(2004)。高等教育薪資所得差異之經濟分析-臺灣實證研究。未出版之碩士論文,佛光大學,經濟學研究所,宜蘭縣。[29] 林鼎晃,(2012)。大學科系別薪資決定因素分析- 熱門科系是否代表「錢」景看好?。未出版之碩士論文,國立東華大學,經濟學系,花蓮縣。[30] 徐豪,(2019)。使用深度學習進行基於社群網路評論的產品評價系統。未出版之碩士論文,淡江大學,資訊工程學系碩士在職專班,新北市。[31] 莊惠婉,(2010)。影響我國產業別員工薪資之因素-應用最大概似法及兩階段有序機率選擇模型。未出版之碩士論文,國立中正大學,國際經濟研究所,嘉義縣。[32] 創市際市場研究顧問公司,就業調查與就業服務/職涯類別網域使用概況,上網日期2020年06月20日,檢自:https://www.ixresearch.com/wp-content/uploads/report/InsightXplorer%20Biweekly%20Report_20160815.pdf[33] 曾厚強、洪孝宗、宋曜廷、陳柏琳,(2016)。基於深層類神經網路及表示學習技術之文件可讀性分類。The 2016 Conference on Computational Linguistics and Speech Processing ROCLING, pp. 255-270。[34] 劉姿君,(1993)。教育投資與薪資報酬─人力資本理論之應用。未出版之碩士論文,國立政治大學,教育學研究所,台北市。 描述 碩士
國立政治大學
經濟學系
107258007資料來源 http://thesis.lib.nccu.edu.tw/record/#G0107258007 資料類型 thesis dc.contributor.advisor 陳樹衡 zh_TW dc.contributor.advisor Chen, Shu-Heng en_US dc.contributor.author (Authors) 廖宜川 zh_TW dc.contributor.author (Authors) Liao, Yi-Chuan en_US dc.creator (作者) 廖宜川 zh_TW dc.creator (作者) Liao, Yi-Chuan en_US dc.date (日期) 2020 en_US dc.date.accessioned 2-Sep-2020 12:45:28 (UTC+8) - dc.date.available 2-Sep-2020 12:45:28 (UTC+8) - dc.date.issued (上傳時間) 2-Sep-2020 12:45:28 (UTC+8) - dc.identifier (Other Identifiers) G0107258007 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/131783 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 經濟學系 zh_TW dc.description (描述) 107258007 zh_TW dc.description.abstract (摘要) 本文的研究目的在於建構一個薪資預測模型,在此特別針對資訊軟體系統類相關職缺。此薪資預測模型可作為求職者與企業方的參考依據,根據結構化變數,包括個人資料與職位相關技能等等,以及工作內容的文字描述,可以讓他們了解該職位的大略薪資,減少雙方對於薪資的歧見。同時,從迴歸模型輸出的係數也可以知道各種變數所反映的市場價值,例如熟悉某項工作技能會對於薪資水準有甚麼樣的影響,提供求職者自我精進的方向與參考。本研究從資料的探索性分析開始,了解各個變數的基本特徵,並嘗試整合結構化變數(職位需求的條件等等)以及非結構化的變數(工作內容的文字描述),藉由許多的機器學習演算法建立薪資預測模型。另外,也嘗試使用詞向量轉換的神經網路模型,針對工作內容的文字描述建立薪資預測模型,其評估結果並不亞於使用結構化變數的薪資預測模型,這顯示了中文的自然語言處理,應用於網路人力銀行資料集的薪資預測模型之建構是可行的。 zh_TW dc.description.abstract (摘要) The purpose of this thesis is to construct a salary prediction model, especially for information software system related positions using web-recruitment data. Based on structured data, including personal information and job-related skills, as well as unstructured text describing job content, the established models can be used as a reference for job seekers and companies to estimate the salary level of a certain job. Meanwhile, the variable coefficients from the regression models provide information about the market value reflected by those variables. The identified high-pay skills and expertise could guide the job seekers in which areas they can improve themselves. This research starts with an exploratory data analysis which helps us to understand the basic characteristics of each variable. Next, we apply various machine learning algorithms to the integrated structured and unstructured data to establish salary prediction models. The results show Random Forest, Ridge and Lasso perform well on the sparse high-dimension dataset. After that, we adopt a natural language processing approach by employing a convolutional neural network on the word vector data transformed from job content text. The result shows that the created salary prediction model is on a par with the models constructed using integrated structured and unstructured data. This endorses natural language processing as a viable approach to construct salary prediction models using online recruitment data. en_US dc.description.tableofcontents 第一章 緒論 1第一節 研究緣起與目的 1第二節 研究貢獻 2第三節 論文架構 3第二章 文獻回顧 4第一節 台灣地區薪資模型 4第二節 國外地區薪資模型 5第三節 中文自然語言處理用於預測模型變數之相關研究 6第三章 研究方法 8第一節 Selenium-WebDriver in Python 9第二節 統計檢定 10第三節 迴歸模型 11第四節 中文自然語言處理 19第四章 資料前處理與探索性分析 27第一節 資料取得 27第二節 資料前處理與探索性分析 29第三節 工作內容文字前處理 61第四節 資料總結 63第五章 迴歸模型與實證結果 65第一節 基準模型 65第二節 變數組合生成 65第三節 變數組合與迴歸模型篩選 66第四節 變數篩選與顯著性 68第五節 詞向量轉換薪資預測模型建構 74第六章 結論與建議 76第一節 結論 76第二節 建議 76參考文獻 79附錄 83 zh_TW dc.format.extent 2953577 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0107258007 en_US dc.subject (關鍵詞) 薪資預測 zh_TW dc.subject (關鍵詞) 機器學習 zh_TW dc.subject (關鍵詞) 卷積神經網路 zh_TW dc.subject (關鍵詞) 自然語言處理 zh_TW dc.subject (關鍵詞) Word2Vec zh_TW dc.subject (關鍵詞) 詞向量 zh_TW dc.subject (關鍵詞) 高維數據 zh_TW dc.subject (關鍵詞) Salary prediction en_US dc.subject (關鍵詞) Machine learning en_US dc.subject (關鍵詞) Convolutional neural network en_US dc.subject (關鍵詞) Natural language processing en_US dc.subject (關鍵詞) Word2Vec en_US dc.subject (關鍵詞) Word vector en_US dc.subject (關鍵詞) High dimension data en_US dc.title (題名) 基於人力銀行之台灣地區薪資預測模型 zh_TW dc.title (題名) Web-Recruitment Data for Salary Prediction in Taiwan en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1] 104人力銀行,AI大浪捲動企業搶才職缺是5年前的3.2倍,上網日期2020年06月20日,檢自:https://corp.104.com.tw/archive/files/news/20200121.pdf[2] 104人力銀行,上網日期2020年06月20日,檢自:https://www.104.com.tw/jobs/main/https://www.cnbc.com/2019/12/30/5-hig[3] Altman, N. S. (1992). An introduction to kernel and nearest-neighbor nonparametric regression. The American Statistician, 46(3), 175-185.[4] Breiman, L. (2001). Random forests. Machine learning, 45(1), 5-32.[5] Breiman, L., J. Friedman, R. Olshen, and C. Stone, (1984). Classification and Regression Trees. Belmont, California : Wadsworth International Group.[6] Collobert, R., Weston, J., Bottou, L., Karlen, M., Kavukcuoglu, K., & Kuksa, P. (2011). Natural language processing (almost) from scratch. Journal of machine learning research, 12(ARTICLE), 2493-2537.[7] Drucker, H., Burges, C.J.C., Kaufman, L., Smola, A. and Vapnik, V, (1997). “Support vector regression machines”, Advances in Neural Information Processing Systems, 9:155–161.[8] Freund, Y., & Schapire, R. E. (1997). A decision-theoretic generalization of on-line learning and an application to boosting. Journal of computer and system sciences, 55(1), 119-139.[9] Friedman, J., Hastie, T., & Tibshirani, R. (2010). Regularization paths for generalized linear models via coordinate descent. Journal of statistical software, 33(1), 1.[10] Hinton, G. E. (1990). Connectionist learning procedures. In Machine learning (pp. 555-610). Morgan Kaufmann.[11] Jones, K. S. (1972). A statistical interpretation of term specificity and its application in retrieval. Journal of documentation.[12] Keras, Retrieved June 20 2020, from: https://keras.io/[13] Kim, Y. (2014). Convolutional neural networks for sentence classification. arXiv preprint arXiv:1408.5882.[14] Martín, I., Mariello, A., Battiti, R., & Hernández, J. A. (2018). Salary Prediction in the IT Job Market with Few High-Dimensional Samples: A Spanish Case Study. International Journal of Computational Intelligence Systems, 11(1), 1192-1209.[15] Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781.[16] Pawha, A., & Kamthania, D. (2019). Quantitative analysis of historical data for prediction of job salary in India-A case study. Journal of Statistics and Management Systems, 22(2), 187-198.[17] Scikit-learn, Retrieved June 20 2020, from: https://scikit-learn.org/stable/[18] Selenium with Python, Retrieved June 20 2020, from: https://selenium-python.readthedocs.io/[19] Singh, R. (2016). A Regression Study of Salary Determinants in Indian Job Markets for Entry Level Engineering Graduates.[20] Sun Junyi,结巴中文分词,上網日期2020年06月20日,檢自https://github.com/fxsjy/jieba[21] Support Vector Machine - Regression(SVR), Retrieved June 20 2020, from: http://www.saedsayad.com/support_vector_machine_reg.htm[22] These 5 high-paying, growing jobs didn’t exist a decade ago—but they’ll be booming through the 2020s, Retrieved June 20 2020, from: https://www.cnbc.com/2019/12/30/5-high-paying-growing-jobs-that-will-be-booming-through-the-2020s.html?fbclid=IwAR1mOcFVDUNxaGk5EAsbkxLU2wP40yxLb8cBqNGjrccXgXoCoiuR4_LxTTQ[23] Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B (Methodological), 58(1), 267-288.[24] Vapnik, V. N. (1995). Constructing learning algorithms. In The nature of statistical learning theory (pp. 119-166). Springer, New York, NY.[25] Zou, H., & Hastie, T. (2005). Regularization and variable selection via the elastic net. Journal of the royal statistical society: series B (statistical methodology), 67(2), 301-320.[26] 中央研究院詞庫小組,中文斷詞系統,上網日期2020年06月20日,檢自:http://ckipsvr.iis.sinica.edu.tw/[27] 江易麇,(2018)。應用雙向長短期記憶神經網路於新聞分類。未出版之碩士論文,國立雲林科技大學,資訊管理系,雲林縣。[28] 周宜滿,(2004)。高等教育薪資所得差異之經濟分析-臺灣實證研究。未出版之碩士論文,佛光大學,經濟學研究所,宜蘭縣。[29] 林鼎晃,(2012)。大學科系別薪資決定因素分析- 熱門科系是否代表「錢」景看好?。未出版之碩士論文,國立東華大學,經濟學系,花蓮縣。[30] 徐豪,(2019)。使用深度學習進行基於社群網路評論的產品評價系統。未出版之碩士論文,淡江大學,資訊工程學系碩士在職專班,新北市。[31] 莊惠婉,(2010)。影響我國產業別員工薪資之因素-應用最大概似法及兩階段有序機率選擇模型。未出版之碩士論文,國立中正大學,國際經濟研究所,嘉義縣。[32] 創市際市場研究顧問公司,就業調查與就業服務/職涯類別網域使用概況,上網日期2020年06月20日,檢自:https://www.ixresearch.com/wp-content/uploads/report/InsightXplorer%20Biweekly%20Report_20160815.pdf[33] 曾厚強、洪孝宗、宋曜廷、陳柏琳,(2016)。基於深層類神經網路及表示學習技術之文件可讀性分類。The 2016 Conference on Computational Linguistics and Speech Processing ROCLING, pp. 255-270。[34] 劉姿君,(1993)。教育投資與薪資報酬─人力資本理論之應用。未出版之碩士論文,國立政治大學,教育學研究所,台北市。 zh_TW dc.identifier.doi (DOI) 10.6814/NCCU202001406 en_US