{"id":925,"date":"2021-04-09T19:02:00","date_gmt":"2021-04-09T16:02:00","guid":{"rendered":"https:\/\/dialexity.com\/blog\/?p=925"},"modified":"2023-12-07T09:28:51","modified_gmt":"2023-12-07T07:28:51","slug":"development-of-semantic-sentimental-similarity-algorithms","status":"publish","type":"post","link":"https:\/\/dialexity.com\/blog\/development-of-semantic-sentimental-similarity-algorithms\/","title":{"rendered":"Development of Semantic \/ Sentimental Similarity Algorithms"},"content":{"rendered":"\n

Semantic similarity is often estimated using word2vec<\/a>, but the freely available algorithms are not good enough [1<\/a>, 2<\/a>, 3<\/a>]. Other word embeddings seem to be more successful [1<\/a>, 4<\/a>, 5<\/a>]. Another (more rigorous) approach – through enriching wordnet with semantic relations from other thesauruses with subsequent clustering [6<\/a>]. The third approach is to measure Sentimental similarity [7<\/a>], possibly empowering by various word-emotion associations [8<\/a>, 9<\/a>, 10<\/a>, 11<\/a>]. Our goal is to find the winning combination of all of these approaches, that could be used for identifying our moods, opinions, values, cause-effect sequences, and automated decision making in Global Wisdom Network<\/a>. Note that similarity may differ from association (e.g., coffee and cup are not similar, but associated) [12<\/a>], and we may need both of them<\/p>\n\n\n\n

\n\n\n\n

[1] How can I get a measure of the semantic similarity of words?<\/a><\/p>\n\n\n\n

[2] Free algorithm(s) for semantic similarity<\/a>, but the results are disappointing, e.g. prudent + thoughtful (synonyms!) = 0.193, but prudent + careless (antonyms!) = 0.43; eccentric + odd (synonyms!) = 0.35, but eccentric + normal (antonyms!) = 0.275<\/p>\n\n\n\n

[3] Another compilation of free algorithms is here<\/a><\/p>\n\n\n\n

[4] Integrating Distributional Lexical Contrast into Word Embeddings for Antonym-Synonym Distinction<\/a><\/p>\n\n\n\n

[5] ConceptNet is used to create word embeddings<\/a> — representations of word meanings as vectors, similar to word2vec, GloVe, or fastText, but better<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

[6] Enriching wordnet with synonyms, near-synonyms, antonyms, near-antonyms from Merriam Webster<\/a> ,\u00a0Word Hippo<\/a> ,\u00a0Thesaurus Plus<\/a> ,\u00a0Thesaurus.com<\/a> ,\u00a0The Free Dictionary<\/a>, and other thesauri, with subsequent semantic clustering and counting the chain lengths<\/a><\/p>\n\n\n\n

[7] Mohtarami et al (2012) Sense Sentiment Similarity<\/a><\/p>\n\n\n\n

[8] Mohammad, Sentiment and Emotion Lexicons<\/a><\/p>\n\n\n\n

[9] Mohammad, Valence, Arousal, and Dominance for 20,000 English Words<\/a><\/p>\n\n\n\n

[10] ScenticNet: 200,000 natural language concepts (in 40 languages, indexed by 8 continuous sentiment scales)<\/a><\/p>\n\n\n\n

[11] Sungjoon et al (2019) Toward Dimensional Emotion Detection from Categorical Emotion Annotations<\/a><\/p>\n\n\n\n

[12] SimLex-999: Evaluating Semantic Models With (Genuine) Similarity Estimation<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Semantic similarity is often estimated using word2vec, but the freely available algorithms are not good enough [1, 2, 3]. Other […]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[27],"class_list":["post-925","post","type-post","status-publish","format-standard","hentry","category-uncategorized","tag-analysis"],"_links":{"self":[{"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/posts\/925","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/comments?post=925"}],"version-history":[{"count":5,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/posts\/925\/revisions"}],"predecessor-version":[{"id":932,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/posts\/925\/revisions\/932"}],"wp:attachment":[{"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/media?parent=925"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/categories?post=925"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dialexity.com\/blog\/wp-json\/wp\/v2\/tags?post=925"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}