Explaining Topical Distances Using Word Embeddings

Nils Witt; Christin Seifert; Michael Granitzer

doi:10.1109/DEXA.2016.052

Explaining Topical Distances Using Word Embeddings

Nils Witt, Christin Seifert, Michael Granitzer

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

2 Citations (Scopus)

Abstract

Word and document embeddings have gained a lot of attention recently, because they tend to work well in text mining tasks. Yet, they elude humans intuition. In this paper we are making the attempt to explain the arithmetic difference between two document embeddings by a series of word embeddings. We present an algorithm that iteratively picks words from a vocabulary that closes the topical gap between the documents. Moreover, we present the Econstor16 corpus that was used for the experiments. Although not all words that are found are great matches, the algorithm is able to find sets of words that are reasonable to a human that reads both documents. Remarkably, some of the well-explaining words are mentioned in neither documents.

Original language	English
Title of host publication	2016 27th International Workshop on Database and Expert Systems Applications (DEXA)
ISBN (Electronic)	978-1-5090-3635-6
DOIs	https://doi.org/10.1109/DEXA.2016.052
Publication status	Published - 1 Sep 2016
Event	27th International Conference on Database and Expert Systems Applications 2016 - Instituto Superior de Engenharia do Porto, Porto, Portugal Duration: 5 Sep 2016 → 8 Sep 2016 Conference number: 27 http://www.dexa.org/previous/dexa2016/dexa2016.html

Conference

Conference	27th International Conference on Database and Expert Systems Applications 2016
Abbreviated title	DEXA 2016
Country	Portugal
City	Porto
Period	5/09/16 → 8/09/16
Internet address	http://www.dexa.org/previous/dexa2016/dexa2016.html

Fingerprint

Cite this

Witt, N., Seifert, C., & Granitzer, M. (2016). Explaining Topical Distances Using Word Embeddings. In 2016 27th International Workshop on Database and Expert Systems Applications (DEXA) https://doi.org/10.1109/DEXA.2016.052

@inproceedings{8c3f32fa9fe745538f349c5f2284a53f,

title = "Explaining Topical Distances Using Word Embeddings",

abstract = "Word and document embeddings have gained a lot of attention recently, because they tend to work well in text mining tasks. Yet, they elude humans intuition. In this paper we are making the attempt to explain the arithmetic difference between two document embeddings by a series of word embeddings. We present an algorithm that iteratively picks words from a vocabulary that closes the topical gap between the documents. Moreover, we present the Econstor16 corpus that was used for the experiments. Although not all words that are found are great matches, the algorithm is able to find sets of words that are reasonable to a human that reads both documents. Remarkably, some of the well-explaining words are mentioned in neither documents.",

author = "Nils Witt and Christin Seifert and Michael Granitzer",

year = "2016",

month = "9",

day = "1",

doi = "10.1109/DEXA.2016.052",

language = "English",

booktitle = "2016 27th International Workshop on Database and Expert Systems Applications (DEXA)",

}

TY - GEN

T1 - Explaining Topical Distances Using Word Embeddings

AU - Witt, Nils

AU - Seifert, Christin

AU - Granitzer, Michael

PY - 2016/9/1

Y1 - 2016/9/1

N2 - Word and document embeddings have gained a lot of attention recently, because they tend to work well in text mining tasks. Yet, they elude humans intuition. In this paper we are making the attempt to explain the arithmetic difference between two document embeddings by a series of word embeddings. We present an algorithm that iteratively picks words from a vocabulary that closes the topical gap between the documents. Moreover, we present the Econstor16 corpus that was used for the experiments. Although not all words that are found are great matches, the algorithm is able to find sets of words that are reasonable to a human that reads both documents. Remarkably, some of the well-explaining words are mentioned in neither documents.

AB - Word and document embeddings have gained a lot of attention recently, because they tend to work well in text mining tasks. Yet, they elude humans intuition. In this paper we are making the attempt to explain the arithmetic difference between two document embeddings by a series of word embeddings. We present an algorithm that iteratively picks words from a vocabulary that closes the topical gap between the documents. Moreover, we present the Econstor16 corpus that was used for the experiments. Although not all words that are found are great matches, the algorithm is able to find sets of words that are reasonable to a human that reads both documents. Remarkably, some of the well-explaining words are mentioned in neither documents.

U2 - 10.1109/DEXA.2016.052

DO - 10.1109/DEXA.2016.052

M3 - Conference contribution

BT - 2016 27th International Workshop on Database and Expert Systems Applications (DEXA)

ER -

Access to Document

10.1109/DEXA.2016.052