{"id":154783,"date":"2007-01-01T00:00:00","date_gmt":"2007-01-01T00:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/msr-research-item\/maximum-entropy-model-parameterization-with-tf-idf-weighted-vector-space-model\/"},"modified":"2018-10-16T21:34:23","modified_gmt":"2018-10-17T04:34:23","slug":"maximum-entropy-model-parameterization-with-tf-idf-weighted-vector-space-model","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/maximum-entropy-model-parameterization-with-tf-idf-weighted-vector-space-model\/","title":{"rendered":"Maximum Entropy Model Parameterization with Tf-Idf Weighted Vector Space Model"},"content":{"rendered":"<div class=\"asset-content\">\n<p>Maximum entropy (MaxEnt) models have been used in many spoken language tasks. The training of a MaxEnt model often involves an iterative procedure that starts from an initial parameterization and gradually updates it towards the optimum. Due to the convexity of its objective function (hence a global optimum on a training set), little attention has been paid to model initialization in MaxEnt training. However, MaxEnt model training often ends early before convergence to the global optimum, and prior distributions with hyper-parameters are often added to the objective function to prevent over-fitting. This paper shows that the initialization and regularization hyper-parameter setting may significantly affect the test set accuracy. It investigates the MaxEnt initialization\/regularization based on an n-gram classifier and a TF*IDF weighted vector space model. The theoretically motivated TF*IDF initialization\/regularization has achieved significant improvements over the baseline flat initialization\/regularization. In contrast, the n-gram based initialization\/regularization often does not exhibit significant improvements.<\/p>\n<\/div>\n<p><!-- .asset-content --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Maximum entropy (MaxEnt) models have been used in many spoken language tasks. The training of a MaxEnt model often involves an iterative procedure that starts from an initial parameterization and gradually updates it towards the optimum. Due to the convexity of its objective function (hence a global optimum on a training set), little attention has [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":[{"type":"user_nicename","value":"yeyiwang","user_id":"34993"},{"type":"user_nicename","value":"alexac","user_id":"30932"}],"msr_publishername":"Institute of Electrical and Electronics Engineers, Inc.","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"IEEE Automatic Speech Recognition and Understanding Workshop","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"213-218","msr_page_range_start":"213","msr_page_range_end":"218","msr_series":"","msr_volume":"","msr_copyright":"\u00a9 2007 IEEE. Personal use of this material is permitted. However, permission to reprint\/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.","msr_conference_name":"IEEE Automatic Speech Recognition and Understanding Workshop","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":"","msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2007-01-01","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":2007,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13545],"msr-publication-type":[193716],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-154783","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-human-language-technologies","msr-locale-en_us"],"msr_publishername":"Institute of Electrical and Electronics Engineers, Inc.","msr_edition":"IEEE Automatic Speech Recognition and Understanding Workshop","msr_affiliation":"","msr_published_date":"2007-01-01","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"213-218","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"227353","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"2007-yeyiwang-asru.pdf","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2007\/01\/2007-yeyiwang-asru.pdf","id":227353,"label_id":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[{"id":227353,"url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2007\/01\/2007-yeyiwang-asru.pdf"}],"msr-author-ordering":[{"type":"user_nicename","value":"yeyiwang","user_id":34993,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=yeyiwang"},{"type":"user_nicename","value":"alexac","user_id":30932,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=alexac"}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[171150,170147],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"inproceedings","related_content":{"projects":[{"ID":171150,"post_title":"Spoken Language Understanding","post_name":"spoken-language-understanding","post_type":"msr-project","post_date":"2013-05-01 11:46:32","post_modified":"2019-08-19 14:48:51","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/spoken-language-understanding\/","post_excerpt":"Spoken language understanding (SLU) is an emerging field in between the areas of speech processing and natural language processing. The term spoken language understanding has largely been coined for targeted understanding of human speech directed at machines. This project covers our research on SLU tasks such as domain detection, intent determination, and slot filling, using data-driven methods. Projects Deeper Understanding: Moving\u00a0beyond shallow targeted understanding towards building domain independent SLU models. Scaling SLU: Quickly bootstrapping SLU&hellip;","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/171150"}]}},{"ID":170147,"post_title":"Understand User's Intent from Speech and Text","post_name":"understand-users-intent-from-speech-and-text","post_type":"msr-project","post_date":"2008-12-17 11:20:26","post_modified":"2019-08-19 15:33:37","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/understand-users-intent-from-speech-and-text\/","post_excerpt":"Understanding what users like to do\/need to get is critical in human computer interaction. When natural user interface like speech or natural language is used in human-computer interaction, such as in a spoken dialogue system or with an internet search engine, language understanding becomes an important issue. Intent understanding is about identifying the action a user wants a computer to take or the information she\/he would like to obtain, conveyed in a spoken utterance or&hellip;","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/170147"}]}}]},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/154783","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/154783\/revisions"}],"predecessor-version":[{"id":537037,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/154783\/revisions\/537037"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=154783"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=154783"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=154783"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=154783"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=154783"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=154783"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=154783"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=154783"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=154783"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=154783"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=154783"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=154783"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=154783"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}